diff --git a/data/adapter_stats.json b/data/adapter_stats.json index f7d5ce5ad817855433b132a8c811709c20a20d42..0da3c8aca4b460a56ec6f63ee4b126db13a986a6 100644 --- a/data/adapter_stats.json +++ b/data/adapter_stats.json @@ -70,12 +70,260 @@ }, "hal": { "data_fingerprint": "7bb18f848bce3afee9bf1be8268e2dc115892db8dc93bd8d3daa5cbf21b60ead", - "last_check_ts": 1783518913.280785, + "entry_hashes": [ + "0d0ed62fb881902ec1be27b66529caee79a8028c282a05f44636e2c58d3b731c", + "d75c5be0f246fa12f14edf98a63a004b5fb17dd37fcfc5e98f1ee0cb9fb712e3", + "6bc415f067f0c6fab61db309299cd2ad65c773941481354b44e2a7b84021eccc", + "0e54d1b6886e100a6618e4c62b5f027ce8bf3bab32a979ca8af2fada6545a74b", + "dd0a60ca93dda9758bbd7827288b5c7a03e69f58084093982b3aae25adc76774", + "3049b037a2df1dfcc71bda2d9c7ae6f4e7968493d96b87c110739e752147a82e", + "7425547c7d1e1a0b0d5852636e920adac9bd3e9a1aaf56a3f7bdb3638fe818d4", + "a8f2c017686674e2005ab4bf4045548c97ebe67d1a787b7c76742891d14cdf9a", + "7461abcb499c28f21e005de853617b4de44d0509f199198f6f52266f8b800b00", + "5d60c4f627370dd554d72d5201ed21949794bf22b2b803fcccc82250df0eff47", + "699b61d1763c4d0e3316a522a317fea28790130c9e0c870c718cacd0b4689e58", + "ef708e1aa4f3b61da3c8e2392b59162cae6c9cdb97e504a96114ee1eef7b660a", + "f1f973982c014ddf1b56ccdf56977ecb09d7f8fb2639415af57928f0962b921d", + "9441e40fc6e25f00da4d7c83724035232d52022a1b967cb3056594108a6c2af9", + "c196fee89a8b475b1c89fe15aa803086f1fff075f6e8eaa2a3b6f38a0eab0668", + "cab478e701dcb80e79f4c7238129db7c15cdf50ed172b3f18745d47dde42d807", + "43ff7617b9990cc678abbd3113532325f03c606b4fbb74601762572fd36a82d1", + "de0f71c9281017a4689eecd27b3e745250ad112e61ee9232fdb56c815618027d", + "30f5ca60494009a3db41d476e1861a4ca9accfd86630d4fcf8abdef5184a7b0f", + "e373ced4e7e54e70063bbf4c28853f82a2af425a1b52c4b984803d0a3cb29ccd", + "af907043fc0f60702f25f0db1d3c6f1875b4ebcaecb35964257d519e02ed2b54", + "600e835ae77d24e35770fcf37a6f112c476114a3f0ddf9f83ff0677f38a9d866", + "c89ce5ed60bd0fed12ef10b4556662522b34c7e40b08e5c5aace74945bb2f2dc", + "8629d50b497cebe6b97ed0c6a235d6509088ccc6f279d57059164baa98c3bb29", + "2d424445e486543fd7880537c03e378fb0d6ed8de27627ade7a597e0cc90002a", + "52b279989b8272e2f130629a5a05f09d22956ffca5ce77e996b546c3a51ba560", + "5ad80b85841ae383a44ca003f95fd72d50c64831ed055991fdeedb76ddc5847a", + "18f4a15a60014369d48cad320147e1d90ac36de276c45d7b773ab3341ba8e2bd", + "06fef5eecac0796890b88ca5979073a05f3e226e46d0a65cf4d15c27e8666d95", + "5f69c3856d66da3a16667557c53844eca50949f9759aa7199a02a3bb1bbb159f", + "ed3a05029eb6c5480dff45d308e050f80738df293c2fc0399b98fc05bd970b91", + "08d5a63996ffb95adddccc18e300ba4c0cb04ae509a80dbe6fe6087193646e58", + "bee380539354e9b80408c90ae3ec48432db15874882309c137f8ff5e56e62736", + "04e1ad4f859a0af33950312bae627a3e8a5a08290d46b97aecacec50d70dbe43", + "0a00754ce61a8d0db4ae68f46bb6f5edce7023e06c9f0e217e3f989d61e7e5e9", + "24dc9ca203150f9f67e39e315766f71773a4a6650cc6546d2148489fd9a7e3ce", + "178265025d04b695fde44dba4ac4be96f5cc8e0dcc597330ba58a8abd07f4d1f", + "2a02d74c7e11458fe103c3a2b0ecdf9c8fb7070c51f7ea5b60302193c0b2acc5", + "cc8b9675df6624d14dcd46801c87de7b54ca118b8217811a02d58d5a2a02811e", + "16ebf6db452b22db2ffd995159203d8e849f6985d875243344b32d63b3edf0d2", + "a4818dd2607a4294222be16e5b39fdabed823eb3ecebdc0ee4a8d2b2f4f91dc5", + "aa47338e042d11e23dcc4736966186bddbab8471ff35e7ef7dd833842344c124", + "8cb9127cba0733df9931ab59bd490360b2f5a8fc7e154a540bb8f9fc3184866f", + "ebaa49f2a3e8934a52138f4d10dbf60f499ec747fa480d16f233a2c1105f8f82", + "901a5fb540c46594e6eee4ffdf69658874fa65284db5a4778f64833fac532112", + "cf7ec24dcbc3c0afb0a154bb30a20dd87beeb66ea615176b16e49a17547f8c30", + "4cada911434ae0c3818d665db7d7708c95f5f4f239c816ed48ec1ea7bddbc2ad", + "4858679ac8529d9d99725cb71b46bddbf89c4ea67d12c1aea448f4baac44ae78", + "43a3e8541034ae7b130f378b2cf1916fa4d1ea70ab97318e57779d881faaa749", + "876a93b7937fa5472d2f9d523cbb003d07c80c4063167288f544c38014bb3dbb", + "a663e3edbbf2e70fd93c56680a37ad914d4b5fa9936adfe1fdadfc5039f515e3", + "eb460323ee36ec53f59ea956ca8f3e1e42f5edafb79cb85e457238da745c7c3a", + "44c030677713649b0f6286873b3158c2e7c934f9b9f40f2e03f6abd99fdf36a8", + "2b5af123f328017432156bc7cdef06ec8a6ff0fc02d130e6f8597e0e91d61bb9", + "04518330e0b93d5f1171ab53077a411bf09c35867b0cc1a83c57b126a209cadd", + "19cc41cefc503935d64dc4cb9b33b8c497061cd9626ac373480e46e8c844d5b4", + "e8c80cca7c7adc0e4734f920082c65cdbcbffb13b07e90abe04772324dabedfd", + "1efbca0c5c6710d622768a883d41ecbeab83608c303bea788914e2a84bf0e3eb", + "11057b14468bd4dc7a282b481d17e8cf683fba5f642dc729695ba86c9b7a4e83", + "1a3f1593818582ac936ba7f610246acbff392fbf287df146090c6c7a58d30fae", + "3afb746916eafc54a21331893a48511d2968702abafcd8edcc1ac1da80626900", + "1246ef7983a3a7f0ff993b2868c1fa99be82e8cf21b2d05cad957d5ebb447802", + "78afb56c6ecd8b10776a16bbe20c4381bd6dc69c812e2a04e7b879f533a9ff98", + "fce2adfb7be10637ab16ec9b6d0ff937621e6bc614f78d30182296940e13fc39", + "05bab1fd3dd944b4222d93a84225a7824a7fe6c1a59adf952cad849e88774fba", + "354f8ffdafa3f9d9655bd6682d4fb68a56cdfcd5b0d0fcad3490260e557f31e8", + "c4b05479f6db83ef82e8bb4ec1106acb7a3b019b5207f47eddf40eed6c0b8dc1", + "3622cd88e427e3573660a528b4f3531089b185da8f2a6b71f2d9591529d6c459", + "5df59f371cdf3bc4a1cce8850d2e261522dbe73cb56f4c61bffc55c1b67b92b0", + "27237ae8a9fd697bd028d763e0c66d4be471655a502d0924190f30f03f731075", + "685f1985e3c1e4c2eb9332bdcda952e1c4a6bb266da31ab74aa8ffb10aafbc94", + "8c72cd29fe223b9b2898b068d8c3f57f9f852d48acc139f76be6a090021fe81f", + "24c6b7a90253ddf3e1d3c79ebdc56db09462b9f0a1add5638716cb0e24933eef", + "ad0453277abb300dc14004a2f8e423fb0a4b744d4cb189ca8da23d1d3fd230e9", + "3470521732b52125a623183cbaba204893c5fad9982b4bb42602f94e5cf48b84", + "8f910e6ea442d436cfc03cbe9ff79094c25e44382cc6ba8da4de00ce0255aa09", + "e80be8fe43a7a128c20ea4473cd392a5623b9551e9b32f9fec440526aecf8fe0", + "0fbd3f5d5700527e1aec5be4247b9ed16ade81f4c438e51994147048198078eb", + "025a591938026552d98ffcf35bdd846704402498be7d1283f01eff3193c2a5d1", + "61ae36f871cd234ab96146ec3f5ccbb7fa3bbd5f1fbfcc2e1990d665923085f8", + "d62c5feeb8a4f6dc4ad767cb71b3c3e81f9c8617ddd50ef52440ca25edac51f5", + "b78a92c740983944265a899bd3108b59f3d2e48e1bd691b7b4237a9fed916c8f", + "39bed40779f8d7bc02164a467c1442035336d922bc54228b340adc52a6886ea4", + "03fcbf75624193c45041286c8c32f1213ede8d9004aeb8344ec614c19adbad6d", + "5f9e9d97d8eceb8eddd47decbe335aa19c71f663228b802a7be20eb6e8b6fe3c", + "67430d8b37e43fff6754a2dad943bdf0cfbedf1570162359d73f46ae930749fd", + "693157c02a012c41f51660d1b351fcb94dede101715bc04708398df84c27186b", + "4d1ec0c4fcdbc8d3dc060b35828226e399eb129f9fe3a60eb66a16f3ef1b0830", + "926a3b75f772c52c708b85ce945119bfd5d102b681685f1396aa6ade8d8a82f5", + "69668d8ddd08a970aee5c20dfa56481e481e8ec5f72afcd7fa8f28af9cbcfc7d", + "cf6b1554492bdb2cce2bb697c6f0c197f551cf88c17b6793730c5204f75a2bb7", + "0755a67c3a2e87f940e324def739683f2aab031fe912a309a29f0f8c411f70db", + "142941b1bae4f33046747595f8382d8d692401b24ad33a6c7c7b8d70066af67a", + "6a3fbb140f2bad5d86edc6af5d40e132acc969b3f7c91c06088a46002bbf1b1f", + "0c181f162b9cf6066139758df75b68cecc94d482bf6726242d68fc1577ae9861", + "6459fefd14130f39b4757124be0ee78831ef43dadd884cd3327dd5672da520cb", + "f0827ccc65d0f8fd00855862dfd7b9d374895f520bb14bfda475873c858d9d73", + "b5b1633318145f1558675c253784f964362414708878e05ae41d863ec714cd85", + "74227db8424090251ff0b1aa0d4dcaba3877563a42fc2e3dba08edf2cdf90fcf", + "cee5680f7b29bb737083104218d7ff52dd197b4ea13c0fe82be558c3b638738a", + "790c9ce0d908c67a484cf0c665ca6321dbbb92a953cb15344fe0128963c0c998", + "87cf52c00394ce0f8a2f9d56ca44cffb50b236fa758e72cbaf43ff97d5add6e5", + "6a203c2faca646ecefa1dd120924162669952b6fe1dbae27ebc95714c77131c0", + "117ffc8e462e7ca924606fb42121101291c48911135181c0d32ff7d8155b8472", + "c34042a10e2d5fa4a958621fbac2ad74c880fb8ce5f457a7fb3437bb8018e2c6", + "40b51b6f3b4a6908ccad9d49e1fc827d4d1a41dcda412a97c194074826533077", + "9876bf41d058bb035d310b21beece22eb1f4c3657bc854de869d27b039f88b41", + "f247cfe95d3a77c4b3e02c9eab7466b2a2b7bf7bdbef79deec905674d9a7bff3", + "21ea6a77acc75d68dd836b28230072bf835a6029b8f7179ed7c4b2b54dce6e12", + "6c9b1c67ecdab87405fc73ad9038563073b1fe022c1583e763ad19586fc146fc", + "d1671e18e91baefe443330686fe20ad2b69bcfc095f2a83f55ffbb48b13635c6", + "c16c33812c2f8594e4695a3b118676d2ab9f2ce852b38af9fec379335937cc91", + "805054c8d9ff09f2a26ee86e8336940a91b36cbe9a56369cd8a0acc7ed177412", + "40eef95553f8303bd0f7b439de5475747bcb5f5eb563746f0b9ef54610df9c7f", + "387bf6eefda003d3de7c06a7b85dc7ac575f04ff25b543a5f030ad9095853500", + "8e4a05b3f6fc1b4976d5149dea31d33e6e8a8544ce03b043bf2115116aed8058", + "997cbed3d21194c96886bf748f2433f6b3b97308333ed041d9458024d75ff744", + "a9b6132c50564123514725896b630d09a7d557dc6e8e3d36c2e1570e7ba84754", + "5ecc132a21509b2284e1810b92ad2e42fa1bae9005cddae570fa198a40340e7c", + "89b9f93f191be111c285bdb5d63eb0f2336c54c021b21a2b454e0afe397bfaa1", + "a57f20b20cc9cc609b2f2c21aa7be1e5f2499392a6dbdf8adfa079c832972298", + "34c8c3c308d92b5e04c6d120852734d836f04f68a47f62d13fcc79d9c2d85c0e", + "f3bbe7cc442a4099bcaba3382590423b58a7e8e9b7e63c9fdaca97958f255689", + "097824f78ce45ae78a29162914bd82e0d5808e861595f96ea1e34a9930373449", + "752be62489535ec2a7cceb098cbae00163d2ecb2dff9b185a21f46cd0b64aa59", + "78eac1160e8855cd9353bfb502f9db5f403a1527f4d74f21def0b1bbe04677d5", + "b27a35f28b3f703ec60c1ececdebb3137c53fab5854711f959681285a79cc7c8", + "58aca7015acca012da519758cc0ab7a3679f8769eae29c3830ae38bb39bbbc53", + "5fac32398de56000eb15c45b75af2ac4be96b2a249a6e95b82a5eefbd314c81a", + "5040953cdd122f8e50ca878bc670238e4e271602b5072491f004cbaffce0d400", + "e7bcc57a7a34f45e28fdc4125a88ac0e17b3886551611d3428d6f21dc44f2f49", + "65b5b3a49c000d9bedc74ae765571f10bf4516d7b06c662bd3d31f9dd90d2ddc", + "c3ec7d405ea25976c73a084080ce7891fc99c90ebff616e1e48d49e5a3746f33", + "160b808fb7f090274ebd869aef77d47d13d073a66392398b1ecd82df3bc3c8a1", + "3bb067fa4d7a8d4df0eac972478589595374c5a6e8f4e57e0e4ca28938a652a0", + "79f984ae14c6c9101994b5023b2a72275a9376c435a276528ace43c530287057", + "5412d4c272bfced8c949e24d2fa77d590871dfe480b6a1e0391468357979f772", + "62ccbe71e3f9008872e490deaa1bd557917da560e3f7831eafa4f17f5c0d00e1", + "fa4af0a536f3684113d77af894ea0a51086adb4704acb50017e23b4292fc7230", + "06211e7879dffbd8c10b54e5ac8c9ad15423d14a1449c7ea4b7ff5c405c4ecfe", + "f2310a563929bc29238910d2753660f2d8aadd7967f1ec75e8c9db2d393cee7d", + "00ce090485081eb91c2f4d4ce37872904710653dfbe5a8a4366c927c14525378", + "277ce6dfa98c680c881c5e3d07e665102c08bd6244a7ed980aa21e0f06a6d03f", + "b33f464671ee62b12d19bdacd82da151e48365ca2b6bb4d70eefc8e91ef7e3cc", + "fea4e3b4fe337472bf0d0e866d13babfae2b31d30ca703ffafb42ccf36392536", + "ea9df7aa06497b795604759285e27725a19153f800a8d713e4934a28cc64ccc9", + "66b05487b266562dd9915cd929d382334a44f36d15396f816536dde0ab8622ba", + "f2b54fbc1aa47cd709b7669395e3b35c75e568fc0a44a4639b96b850ba81b39b", + "5daaf3dbc1e086631047bc795d3a6ef8620357980555f6e029b214df131a4e97", + "544fb39925128c032c815ab65b51e259d8ce043dc19b96d0916657382695a802", + "03c3ae93232545b90ac1e083bfec7f84f521714d16cd5e5f56c215b857420807", + "cf3295110be2c343ee2d4da7aa8edcbe47f0ada9d1215700c19b6e198393116f", + "7009dbee94f5afe0c7846e128a180aa2cd86f9a91d258001bc1911b4ec902779", + "46ee6480262b190bb8ec73405332d5e01f0849900185183dd21d28d5d7b532a5", + "b04ee73e9a16dc74660fbb28dd145bfc8727e1c6f7977ad313a6e238cdd1736e", + "ccf9fbaf3307af80eee21153e9468d1ec50f187a844e4c2d4cfe4bb76673c8ae", + "caabee491d6c684ce5249da8aff66acbb711376370d1f164b8bc62f19992f76a", + "74e33dc9ed24c3da78b68f2aea206c7d5e449e7100917acd06b749482f90c51d", + "9b5b674e994f5502038cf63ab66d460cea2796880e9b4d6b2bf18b4fdd237ac7", + "00ef0a573689d910d006e9cf8b7e4e4af49f85a21587de844ce3c689f4a01012", + "370ef903a0c235717cc40982188453019de8dc01e8f0bd0d55c0a815dd216366", + "f1f5306e616a4ebd5556c487a0612cb7ac650d539226e046c39e1daa4b8410e7", + "95665deea8a1a2bc444836fa505f69752e2e3013ac104272a1478178416b149a", + "15927951cfb501352c7fa89c9ff6bf714807f61e295837123ee7d35eee6eb501", + "14b6f2dd65d059f8fd4d562c53e8caedc34c851f67ee2c17cf93ea04555f9ec1", + "f367b7ebb765c4e36940d7a7d5c68decb2438fbd2519e583ffc448edf82c27e7", + "96bb274355d19ae60ad71382b06402f42837b738718599679f1e3c5d60c92e1c", + "b0ad6760ba93cd423118103a33bcab29c8871ceda7b78c5d3f256feb7394fef5", + "aadfc7aab00d337412e2f4a3318276171b30cd12f4370071e87a933c1cf0360e", + "607d7f3199eb4667880f4ee8147dd95cccc1cca32fd62bcee0d472b5e801fb63", + "07a01d61d39892229d1ec78fb4fb27d605cc56b5d7a3209f68b2e5484bca8f41", + "d50013e6a690de259df9bf14b8acd695241671d940892b35458b00d496f2de68", + "54322d909f88b10bf79b2236bb894c766c566d6ae88fcebbda46338e0ba4d3ef", + "f9a6435cc6cc89dc7f100f9a287f0a3744d558b7b7b7ba5c4513cb88f0e73c74", + "c98acc559884bc72029169f1b9b707030d0f6e84b454e8b448c2a16626ac1eaa", + "d65b5114a874be26224c4ebf4dc3680243901c3ba88bd691c7224bb2d67a1afe", + "35233c2fd54f5b74dc56ca1ec73b95a6f58ee6b0cda1a64bb9499db8cd39e0c0", + "462a7222a9406e079805311bbc9211b48dcfaa47e9ecdb0df27bc406c31107c6", + "0f5d50f20beb6bcda249a6dcd211acd26b6e1645cb60f1139d561c40f78694a2", + "873830fdb2efb3777a25c5176e89064fedf70bb9296d92b9eda52dfecf03af56", + "8f9d04bace27f5dc447f8964b7f729f41e219894f2fd0cbbe888638e1f8aaeb6", + "156aca666f64b421b0420ed97c3985909e452569f490311016abe24d3aed86ca", + "8cf5a18cf5430727fd58c77d94f882efd2b49c2ce65c7f17015d3063cd182b3c", + "0a877648e0a32a107aa67351f40062e46751c0fdce144575e3196c9a2befa4ce", + "476e90bd0fab3f537276d9af02d982d55b83838746296b31abc104726ef764a7", + "2d831ac2856126cf5fb29834e33c886553635ef87851a086a87db1f6baa6041a", + "c222a46c1bf08ab640842f438952710ff5646482e50bc0eda022cd033fe1724b", + "82daa94dc90782d8114e359718bd58106537b1ce195c73126cb10ac7fb3b556b", + "69118e035eba687aa05e30bd07cccf185c476415ca2aaffa6df764ed7e02db9e", + "8f96b96a2a6457abb4bb5e5ef94cf95643bd8dcaed63688ba6d3b880732c8ae3", + "3ba01b2ff28060702e3af18c85c0ef4dd11beb465ab04a5a3eeaa6d593ca27f7", + "a7cdbf1f8c9291a3aa32eeac1ae807b9ec5bc7d6c9858ddfff2b3dd4461b61d7", + "f895f747f30be78d49cbc4782da4a22ad5578911a8aa497c516f16e94d1bc02a", + "a0b5e0b46bf0399a3cf33db02123bf5628cfee6d50c6040839b91904269db073", + "6fbb7903162c4459c0c1643b9cfb2e7cff4f0df588e3a28803ed85d81f2d4cd0", + "4e8325677550613c9b9ab4e62ab0ebc03c46b8e6c110067905a5ffd64e55cd92", + "37cec027c7f5a376843ea988bab767d5fc6d6f01f21a521a9af8aa950ab3e090", + "4b0d4b109702aa865110424afd6cd10d0c354a7c89101537610fe7b3bcd102c5", + "2ea49a7e6c8c854ae700e1b092595fabacb7ee23412733cb3ffa07234ba29702", + "8191aa8028ec685db39a906f712cb5a37406e5144cfa782ecf7e80e0f3659796", + "8695c0ed1d4155cffe722e74825082b40fc76c8842ef00fa19341bd5ce14731a", + "2204daf7975f1cf76d536aad09b2fa0eeb5c93d63d95238f421344f48fcab88f", + "13ae7a9d1eaaf250cc6992a6ff23742550617a01e7e2fec3ca4abab587435450", + "47c759f6d8d83eb12e8ab15d9484756ed0d44266d5b026405daf9d369cbc85a8", + "ad67d90856bdfc1e0b1efd434615fb9372297b72b5daf8814704d0923fab656b", + "45c60c7cf322ae98838070ae3dc5454c94a87d862edc524be39f63ef3856f7b0", + "98a5344a4a119a7c32a3edf2d0544964cc400a722a876a28bfb3bf8cb7761f43", + "cc854e6b217c194a30fde71726b9e4cd27fd24d00294007f8281170081534c85", + "a94dd065a5b0f41127295aeb3a92249513dfcc8e9593147a7e2e85a5540e6180", + "c52d2e978c2364ba882caa167d9bb2d1229447fd94389496b31e9166a34393dc", + "f8e48207990dbf121ca3529f726afbdfc7a28bdfe1988592799192d81a7c48df", + "894ee67103ae94a67682165e0670b34c398b49baa5b4b1802af8f7d58d38d2b3", + "fdea4139d5f1545a97b15983484770c00aa841b40d43ec9fe4d1942c48ed8ca9", + "df3f42fa9a671bbfcb065e269ebbdd66ed424703e1742c34a18efdb96caffdda", + "38cd7d9cddc532b4e9405e571ec5f08966b969561ff7f67c3155082ea978a200", + "d9162d1840d64dda2976a76bc002f9ce27e7b4b9ad2d8a30f8db6dc3c42baaaa", + "2034d37a2d03c31ceb6d5b820cc3fadabcc1ce1d09506661ef41d7c2bcf07c61", + "ca5b9c832487a6ea6f81829afcb6ad04f90ece3dd3944af9322ec1cafda859ed", + "07f5bc88388bd6ebfc1f003efa9bc1c16008419b4ae6594818ca637f697b8e2e", + "fc96622df7d067ae853db87068daa438a65d626b16b6731e04a084066b13bc2c", + "98d0ec8179c324f6d8ead3ef5692ace6a3f9b8490e62642e7792491ea22a1138", + "0ef76953ea5fa1fb27fa936ab93707874ea6906a42e06926b1315cd86ac61967", + "0896ab271c7a0fa409fe4593ffa3e66319dec793a161fcd7ac4e8b46971fedac", + "4d983fa653fdceef1cea7a5a8aca3384763c0342937f56213517602c9f7233d9", + "8950eaab11d31bca8577a9d27b301524b5e60dc2fca3dd9b354f8741d7df1949", + "0364cb78f8a9b1bebaed43afd037b237595073bc569e599de4e236ca5e0f3329", + "fcf7a7f0859ce2f21d4154c2bd2cc7c15bc0ddcee8bf3fb366402d507da27c43", + "a9eb0d3a41562cb6d2fa8fedd21e752b426a3f32e5718569db7e25362ee50c7e", + "1741a928f3bb53d72cde8ff4bd66873d2d778404bcdc18ec149bb101e74e9432", + "932abbb9113627d83c1ebfddca79dba32b92ac195f906c9255f11797f899f074", + "a07ed16def5bdf25c2917a55d4cc0110982789ff131440e19dbf1a071d15ed89", + "ca2da4af0f73ef623bf8c220ecaa61c250bd9540e2c70a1f710670398ea7d0e3", + "d1a102b5048c54bf164a25cd0f5d7a6581a18a73348a06639d2920beef2b7038", + "f18523972eadc648727f3ba2fa2c4d35d38930716c7b34f32c21912a833d1418", + "a7511ea34dec0b28e38ffb032f6cc368be0fffbf70a55f33f6f9114165f3c6c0", + "96a882fec2a7aed36b827de355b86649853a94c1ff5b3152455bd21913106d33", + "50d20b4b72b35fcb9030e3d9b40fb6ad018646500f9a8e559d6cfb9abf54e50f", + "722bf153123ec8a20f74b269769ffd3fd1940ee5dd9702cca038e389b10511cb", + "d81a2cdaf10714887ec16195d6a83535eaaff79093173507ec2f9643033af9ad", + "da5573787ad7b2daf47433acb127391f44ec1d661a19941ccf78bf672523715a", + "99bee1c045a2d00be06f91cfc8ac5955b1e8dafd38dbb8edcb744f772561719e", + "2587be042d4c4ac9b9584e149bac586db898d632ab651840049baad5de850abf", + "9966347f05b3cf3d37875254e7bea6676c1219370e52ebb981cd38c4fbd6977e", + "929a404c3968e0cf622283c76da74e40baef41f0152c20cc69542031b68752f7", + "be6ca8b8bccd67992ae589d9c1cb0d1da8aa1d22253746099372f286e184295f", + "fa07bb5fc21f0f34153fa0f04c90b83b7fafafcdd0c2f4f7b44dd7ac2323071c" + ], + "last_check_ts": 1783734284.204738, "last_data_change_ts": 1783518913.2807868, "last_failed": false, - "last_success_ts": 1783518913.2807846, - "size_mb": 0.7480096817016602, - "time_s": 18.590089321136475 + "last_success_ts": 1783734284.204737, + "size_mb": 0, + "time_s": 21.625577211380005 }, "helm": { "data_fingerprint": "7520de7be640828cdc5618faeae0492cd279b920a8afc4b62121c35ee89e2b00", @@ -4808,339 +5056,427 @@ "last_failed": true }, "llm_stats": { - "data_fingerprint": "b3f361b038059cc7d93ee49351206f5cb96587c21e13818a6cb093b4a6ba0b4c", + "data_fingerprint": "15035a6e1ab84e6d6c24579020e3f5f3ebd2cd6ab667b52d07d6ac03a226c463", "entry_hashes": [ - "37ec140467034946ae4494d97e562d888b3551a48c3613026ac747ba176781d8", + "104fc42198ace9e40ac5fa6a39f5ecd8d3e45fe1d28ba59605db361cf0007634", + "2644312fc811ad545a7909ce61c69c30e5eccbe9405e2698c85b98081c3215bd", + "3c2a65e6378db9325836cb4bb5d1b41629bd51fb1468e43396463aee0ee242d6", + "5abb0cc63c343717c4b888aa07c3e27a494cc66603c324c84e281bdf699e2493", + "d793f12f3196b6f62a9755ba51e917154bf605a614cf11ae3b21a4592faf14ab", + "526f34c2f3ee5be7c56b587bea87d2d359caf2be0b9a35ad10d5509b033056f3", + "2410ec7188a4dfc480796e45c7d87486d8037104d8da25fc193cdeba9b3353fc", + "6085ca91e1ae99b3c11c214ee590773da455ec915fa8f1b2913564b5e1e0cb02", + "44851a53630cbde6b2972a0564684a996e8f50f657a8f70dd6f6183cc1c5d70e", + "055c81fd0b51ef3d502579eceaaab352f46d315af4291075efa5c6618b0ff587", + "81f97d6db3262a0c976a0c0b906aff8e2d6f7b995ac3be30dbf0bd1d61ee7b7e", + "144c479641b987bcad9980a998ca75ae1c87454c3d9b4db11d9ad2276ee562d3", + "55478f13280c4e64ef382292d1321cc359c3acc780974748254d61bc7249afec", + "0e65bf0411adaf6e983d34743406483d1887887463787896aac310ea992477f7", + "ff1ac56824f6ab3cba428824f83eea2bd0cf6cf582a2a2acead8f58e7e9a332b", + "0502ac5de51c25c74aba902e24a76c7e5c4ace28de13ad3cc9f8c14323c9c2c1", + "9e09e5e59261df9508fe97b6d928ccc44bfcb6744065739c9488d816cb902a21", + "5b44da5a1418e7f8b6104d6b9dd4bd8a77644cb853e19038d2e07aad1663cd9e", + "bfcdc1ae5d4b4717818d32a93bffa446176d76cce25542da0057331381ab60de", + "adabf2b5acce4b8d6f56c3b2bc1a3edcd6cc235568c41e76cc92e9f3545a3fbb", + "2c3c412a6a3d14a2e5e83305a2833f1f44ffa53f8f2e7485cfcc4e41a20cb38d", + "83d5c28195fdaee0106b2fff660a7177c2b7b82bf3c86fb21ce8294cb7ab9c62", + "2750832bb6680e2ceabe653df70bde6bb78d620cdbe6095a622a43481cb153d5", + "f5178ded8ff5e65681fd7722b086143caa0da2290fa01db3b5c191c87e82320d", + "acde001479007c4a3d677002b308963821c668fda51fe5dc8316ac5920576078", + "7a16dcbf2f35c22cb00e4377e5c1133c2cce19d64647e663af63861cb9948f17", + "3ed6256ef9aa506201023470d899a32891368a9e2318dfd36292b58e7defaef9", + "0f8f58db710e3c3f7a6140d2918bba4677e6f39f0c25465b30d2890aa6a1fe2c", + "e40c511c9821d49d3d6404aa6fb65f519196d04a08fb248e8723ace7e8c3bad7", + "cace8f4e06b62075dc7d9571d82a9e9411f5fb06353f13c8bdef619c9dbf84f1", + "67a5b07cbbe4b5db13727a0cd160014c6903d51d95119d3fbb8f456318a17f04", + "06d66c66a8c313fca3a69f167c7f106d110cf63de52488c20b3639a5cf6dcdcd", + "d9ad70aa35c0035e940cd77b0b65e5937751bf95bfe5c3aaf8611fc6f60038f7", + "45a91a8271bcb7b2bc26267a5399f858426c8279be62964a814c7f582029703f", + "f074de330b5313389b5e1576e7ac5d6faa9c969b5c3e18d4eaeb3a01986e220d", + "669ef97910bec0b49906da900ec79eb63a31dac9ca1b2e9ecbfb687eaac229a5", + "3b70c21acbcedca1fbedfaa28cb0197ad39a10fafa87697d9727dfd320facf95", + "5d144b80daf421a7b516450994b82dd95923e679af47dd264b2836a0aa194083", + "58e097fff12ed3c5a71d83d2f7269b9718c95fe463b6baafa428f3e927b52d7d", + "69defb814b515fa41240c26f039750da87ee98ebf02eda27f6e1a99f701c6e46", + "3ab73ba9358ee8ba6d890298b377c7ff57d1dedfaa2a289393b44f92442b5775", + "e4a08ad2fdb152fe225ef710cb0253bd4a885e92eb38168dad9836b8d24399b2", + "d220e2fbfea15c1ea65373be2a79e67fa48915eed65478e831c9f0478158c8fc", + "308e9ce92c44cc4a703f53fede587fbd0dbda03d9c8e28e378ceee761788880c", + "4ae13f4e8efab1ff6a572c487c2256cb921ce1a6bf8dac34280c31c77b26791a", + "6d8509ea2b05a396a74f733f35aee83e8dadfec5e9984c4a79db026358191d63", + "25a7d86be9ebd83d933157b8a785aa555a771922fe871306921041c64c5fcab0", + "8c00ce93da31d69262ff453143355c1df5505f4fe458cc269a2a3ec9304973a9", + "be45b40ae288bac756d8877e2ddbff7337391f2f2b45ca7532bd999968d88c07", + "be78e7c7e2f933b94ac6448539ea741915ec146f12a1b0b5c85940d752806cc4", + "ca7e2c4dc03a37d5b3c89ac97309cc633b9df14ebcef542b7657b157972c17b3", + "ceb1da4530b20ef0a58cc7790af40376459629d33d5ac650bbe487adccd41d09", + "1d0a9fed02ed242a8c4fa86a08c5524ae8970ba91b3467602e6b3a9482a6e92a", + "229643ea98c02a79267ebf14b7fa0549cfb009119558c73d22711171f2a41925", + "64773f2881d255fe274d27757ef6e541d2fd7b9e8d26215d6bcd6f80e6641018", + "da5a3654e3830fcb2e1e25027bc74fba028366e8ba04b35ea30d733edda3b996", + "4a7bc8415234d1a6d707aee8795c2907a713768e23879dac5d1e4a8fdad2d26c", + "1ca6ead4b8959c80dc2df72dd851897e516a267e654b25ee543dc5b87e57a5c6", + "ef940bbc4f399edfb66ca83ff25e8231085c7c005bdc2f1840c84755253470e4", + "b143b5ec2edeec197b741dcb94f7e8f4aab0c141f9142633b90fdf192b9209f9", + "7cfea4ebedd0faa14d193e9ef9514fe40672d256522538d6a558c58f668bee2b", + "0174788d03a36cc41b824b01edf3233f5c65c3d0cb5cdabdbb2a5d73085aa900", + "1711107b207040b41d4ec0bbb6dae411462497f510130b266f1c0f0d9af883f3", + "fc6467bac47f96d49a1ea88d31836766f9faec54bdcee2394fa927ea12907037", + "0efe969055f2e039286e92c09f282b2bf2e2ed3afe692d5edd45de8962612c87", + "c8c96ec2a19a46360779516408e760513800153aa9f82daeb7904c0bbd6c536c", + "77f3aaf03a6182ec4808611712861b5831f7001b3acaeb821cca2277000a3226", + "efbe6cef7862e5a5e05ecdf878c9598dae96463b336c1f664d34bc3f2ba094df", + "3f28d0375be777b072d17bd15fcb6d8f3d229aeeae24a4caf1dbc8a1e8e6ebcc", + "5f1eb2233ef1bfd5ec0ad591553d7887f77a7ecde545fe473c20374e32cc2065", + "a587a5dd2de4d4fd8f01973fee8e39447a59048c8d0648007b7d8e946971720c", "c62c9297d5bf02fcc2a3d728e9ec8186d74c2f1f006853e55e9007f2c24dcd2d", - "d4edfea5ce889a964ca467e4b8fcc99151040e5966301e6954b7aabe881e947d", - "a5d1e6ae6769d7b2950f26628e80039893504bac8f0bad9beba7a028f641de6e", + "93fc2a5d80a2dc24b59a144f4201c196509285caa05d23bf0dd064c353f6e379", + "917a6013799d8334887cce9c02d6bc49113a0913c0065c480af5619fbe01057f", + "beadfad0756a95f260ee9761449f948b423621d3aa0292cadd6f0ac52ec02451", + "2005be2415728f9bb21ff08eec822a0c415b1bb93509dcacc47ad8294abbc2b5", + "7bfe98d629d3883ef4847a216aa3f510ac8662807d0049173fb83f8d6c522175", + "8314a32e2eae7989392b1b7c9bca3e5ab24f4832dd90ecfe861f54fcd59b85df", + "9f83f91a5a7225b4f2100747b708f727dfd349014b0f61f72a181bb583ab0a07", + "40332b1e26e618c5d8a8032038eb6e36925d77be346cd4af72b1623f0a411bca", + "5f967f10274707d72f95495576f92a0ae743de88ea11097d97fb9c700c4d64ef", + "962787e7cfc0bfe373a2d1a84ed9d9f852378211910cd17f9872b7f0aa6c9546", + "c1986305bc63a407a1e9536ed5759382f9c1c68f38b542fcd8a4605d5731ccfb", + "c2ff3b939c5a078b9b8d7a62696d8093219e90aa953f58731ef886465cdb7c1c", + "70bea43ea7f00a49e7f5aa8205427a7b4d494dee674f2b28736a9a6724e78031", + "eb4b5f06bf718a3009d81dd4b48e340c78b25d8c408818a4e24ccc7e4f01b588", + "3a16b5dc2aae2eb5b46e72f1a28a5206280c926ecc0b1984a385898d72c6948b", + "3cb27275c75ca6758b5f8cb661e11bf75565e7c405b557ea4a480925d8543d91", + "00947e1a8adf1d336f99be365e902e3c46faa383a5b297c4745ef1138b4c2e69", + "f1c862dd8ae20623c53ed1dc5499b3a70cba7c2227989239fd979e26d981cb21", + "85fc4ebc2aebcb62ad86c36c6bcebc06dbb4b5c6e23b7dc7ddb65cd8b80f04e7", + "ab8a6fb4fc0a3687b3edc2805560f10449e4145b7230146554957002f839427a", + "776f95f8ea0c224a88180238d1a3e6c391bcf0662ad7589cc3a24e6efc030878", + "02f79b2dd1e381d0e5bc85b8c28f438a60ca37d6ab65b8bf5577cc0dfd71bfa1", "edc3d65c1f6b61928b3fa20074c93dda76925f0d368c5934bf0827661b9e53be", - "06193195d847c6439c80511e8369f3df6d9622cd278e70023b6b53685352d7dc", - "435b5e43605476a55dfcf69ba748122c3d0c64fcd0ad9f4f5abe5b73041754b7", - "3addfe58610cff503feebcd253b1683311668adede2885f208d548d7bab4f568", + "19182796c8dffe9a637d2fba86f5658e2ca1a42473eae4d680ee7b06d5a8fa9d", + "c4f12db3c7ea9bf4cb1a02e5839f384c77cda688a8d5b9f533a9eaeb6cc2b3e4", + "3933d516b972d38318c3e390c32fae03d09985689246fbc21dbd8e2c4e389b15", + "2aaa3d7afda10e326fbe98b639009175072ad33b600469e538fc036dda97e679", + "d11aeed95f13d32835343065f87566140a9c52c0e3bbd64e9eccf1f4f3c1a0a8", "b96c340ee53ef23409e8c45b02da34d9db3a47cb6423c1038b43846bbe8d8875", - "0adb84cfab5b1e640f0383beac85d76b36ff37c6a1b35c09afad11ee270ccf61", - "5d4917445a8fa4db7d29a394c5c559d55d9a9a2cc4e46c3b49d7687b11d0a8db", - "2e092af8db369daa5c108b4ae3a8ab26e2fea0682d287887dba97aad7ae960e7", - "6d46731cabb8b6eb6481239fa48d0657711160e6e04f5dd6a74e273609a1f4e7", - "3ed6256ef9aa506201023470d899a32891368a9e2318dfd36292b58e7defaef9", - "64caec4a2db915012d162a3e71403409d85f13e85a55031f81a3ee88f60e9caa", - "e05d64e23e9c6f345ba316b086dec26f89baaea52d4cc4f8f5e5bea4d831965d", - "b3c44b04cf3d06726c0009807b5327747d2b619414a00075219451f715ff5108", - "8c00ce93da31d69262ff453143355c1df5505f4fe458cc269a2a3ec9304973a9", - "60fe1618c69590f51f0bf7c233b162ef4aca4cd686719438462681d867117d9a", - "1c29c23fe08ae7ee3aa5d0800984735f561487346735f559389a274a2884133a", - "c99c4ca2182e8c2af3c00e66af5f6712d16db9d1153436ba321699b9c8191138", - "7827ea6d7e72061f13bc13018a1a8d3e3fc9b202448dab4cf90a039ff76dd4a3", + "d129acc6b2a04c69882211369af3a88a9c7ce8c8a827c1ca9b0a5a25b8854145", + "ad352fdb134281f34dce64473154e7fdbfffe91294ff73c71a7d72d62b8790a0", + "a920b7f2c4bbc5906a9931e4c65b1db4447d33e3da5fa64d3268ab23e2e4676c", + "4f7a63dcba6af8bd95b5f3a56119515c47c6ab861ce4a00fa66fae9204704973", + "1a36e02efd1d6917c33c484ed0c432669dacea23521f348cf5541ee13aa84c2e", + "696c437d4b707021cc10d282c14b28de3d1a4b3d268a78cdeb6acbdd897f9ecf", + "5c44a3ec370d76330c3ed037cab1591d29a6e6f08623a28ee51088676d72ad49", + "bd6a4024820c9e62778bc05dfaa579bfb7758eac3f247c7ace052a6819b2d369", + "bf667b5d2ab22f975b5fe7d4d5b05dce62b999ac687958d74b1ea3dfe40aa4b4", + "6879c738f897d657eb2b6ef79b8a7e617e0490fd004541cdc14d06a4739e8d6a", "3c6aa5cd175799101062e98829208c9606624795507d09248c277c83d683ff4a", - "a587a5dd2de4d4fd8f01973fee8e39447a59048c8d0648007b7d8e946971720c", - "c88ab28b22f68d32b8c7844bc633957b1b6ef37c17b7ed2d070c7bb06959ce33", - "eb1f11f5d6e3e0e044527d294b58ed44acc8e4e150c1915c1c067d9334d7fee3", - "cbb884549963523a63790861272525223bf4fecff2555f82f8a7773cdf3630c0", - "69c3320bbe0d9afd6295e7408c03018ee0ab11ee99f687b9f768744415eb5f0a", - "75c06831b1b8504a58f541336a2c50688c6be0b0bc363f81d4ce4f98e7dd844e", - "21ea53794b0650ce1a40eb55a899b63845944a80ae436a98a24878a7d1a12925", - "d4786bdb82b223cd108ab909aaec54800f453e874820503ea262c797c5aeb440", - "9e09e5e59261df9508fe97b6d928ccc44bfcb6744065739c9488d816cb902a21", - "c49a549986fd8609ae2dd232998cf111a040f1710662f2634abddc309fcc8f33", - "d220e2fbfea15c1ea65373be2a79e67fa48915eed65478e831c9f0478158c8fc", - "e40c511c9821d49d3d6404aa6fb65f519196d04a08fb248e8723ace7e8c3bad7", + "c5f8f2ac3ac1f8dc1b32befbce5300382c075079a97a50bfe34bf81413837ed6", + "386f862bdf3c70fd3fa77ac830a1cd6d95b6989ae2f83eb1de0ce6653afad3d7", + "5e46d7b9dc04a377f5fb98e17e653c1291213e3bfd84bc8c97fdddcbb98e84df", + "92a153a6b6b0988245f6b025464a610bc09b9aae04f0bcc9c8dd392156bde9bf", + "b8f733e25d9f1324aebfb5fb4eef8183b9613cef205db92ef0bcf528b4511230", + "40cdc91549b2359e5af2a71793be3770eed05b8fa110bd37b0afecaecae1a782", + "301598e42c14fbfad17534618890846f62593ead552d2ca08bb23358df817340", + "ed6d1e4515cb8779ac1c1452f4a36b6cc386158a3b8f7a2b1f6f9361e6dd4b79", + "474cce92157003183710b0263e7ae711b51df7df2e44cd1379340cd455208e34", + "0cfc49a35824b6ccd92c62d90fa34903d6160f7a2db47baeb1cbdbbc73d85205", + "698a2d9fe3ccad20b21442c8d4e65d32349c3f686080c5c88a800329875c1475", + "aa202dcab28dac61f295c429016209061b899241dd4d8ab7ffeee1b4163318dd", "86ec35c870a7c7e7e2f54fa11bcaf50f7d7c969704e41338b16fc7676ed1ae7c", - "258f11ab4103be0fe6868db4ffa597da4a4a04e340c1939e8c4bfa6360edc3b6", - "6563ced2bfbe24cea41f48773df84e496cb1aca4eb74c08dd445e21ef62958b8", - "24cf6e083ce8df481c45ab52bb5524a1c75c65ede65c1498382483a6b3804e41", - "2345f6a00f3124de0de669ee21cae679de9bfa4c0ba35a7108731e98433c2c99", - "0adaab84758894484796ab4ee28f6bf78216145ea73bdb228636d4ad23c549fa", - "bc711c3e82d4c1628fe38f7d5fc5a7f1b16d8fbaeef4eece212f87bd3c80cb2b", - "93fc2a5d80a2dc24b59a144f4201c196509285caa05d23bf0dd064c353f6e379", - "efbe6cef7862e5a5e05ecdf878c9598dae96463b336c1f664d34bc3f2ba094df", + "7f6361f86b7eb9156deef7a7e7f1eb630bba012e2a4f4fdef36ccef79dca4bb2", + "6d46731cabb8b6eb6481239fa48d0657711160e6e04f5dd6a74e273609a1f4e7", + "2c025e6a212814b3293278e336d8c0a07579eb207b3afaa50fa4f868d057f81b", + "226071546fd8a74bb075f411ac717d806afeb783e78a7a89166db88896b88d2b", + "7eecd3367145065b787006ba3da9be37c0dd6c25ce6acbd363143dc3e7b30bb9", + "e1a8094022b30a6d30f5a5166ee06ca44f103251d3931f34695279c7bf5ce785", + "77dfe12e184346c1c06acefeeaa802a897b11cd9a3687629fb0bd0d75bda5937", + "674c5487cc625bb6b7aaeecb75b001ca472ead48b18d4d9938a6af7d562e9479", + "6b89fbce1f17dcedb368f7fd8ae181bc07339f5aa66484c24e745932709c0df5", + "2f8ada99fc673a6c74c5984be0b9b83354d0a13710f260a8971007d9c7fafcd9", "1fc0870ddc0615cf3a850ba75c93e44d6f949f993df65404eebb5a423894830f", - "e60bd20e33838f03c0fc24180ba67cf648cd0174ed872bc2cb751d4d07e362b0", + "33d471ed6546cbbc5a2336781a285382c2d4962fcf7cc03b6986b82e9f87c0d9", + "4bc53bf0d0c7c48275730ae382231fc11c751d1c884d1510638f2934fa0a60eb", "9952053b4a9ca4225d7f2224a4d8d40cd2b558092956760203d9a25e6b9ac083", - "d5458c02dd2c310f6c32b1691e9e86c583055fcad08586509796d2ea80c983fd", - "1283165b318a3062dd9cc5d8fc387bd2a6d7a34a675d97d1a3783376002d2f8d", - "a9350f92845058d80e21fb4d349a5fb0b199a31e49b7fd67d429c6ae77c28301", - "65fdb3a7251cf4e52b1853ba6782706f0dc40d3f2af2025a1e35f2e7311580da", - "193b9057753cb8184872d2093eaa75cf23ad4cf589b94f2d08782410a2bfec89", - "aee52a52b6e560370b92dc15bdc5fbb9ff47e86671e715c4411a35148beab021", + "367e64680d1ad9db1cc7f73edbfb5f0f882eae7b5664d711782dfb8281bb5e37", + "0adb84cfab5b1e640f0383beac85d76b36ff37c6a1b35c09afad11ee270ccf61", + "889b221f8f14823e3e518161f0e716c8d39bc8a67d245f367a4f17c50fa6da2a", + "28f978aac755e315b6c316fdee28b84db51adeb90d838da579fb986b12c1c4c2", + "9318a3e87a8a59ae900c564ea65f17b9643c4959e69f9acc33fd3afe68043c9d", + "145c24c294704b53a8280ad1bfe7a1ac2eb6b79c738aa9283899c6d2309e48a1", + "0690c7e7f7d58f5e2dab5cd421e80f76b7827027fd340b3fa5dfcc7f2f77f01e", + "601e3768fcfdf06688ad78ed4adb1cf63371880a0d779709fddb0aa45f0f484c", + "9b3a4e837c7860e00cd637a6b454fc89f1bf0c0daf4d48c9265c1866323c2708", + "3addfe58610cff503feebcd253b1683311668adede2885f208d548d7bab4f568", + "9fdfbb12916edc6a446bd5293b26c395c408fcc00c4db8f588ef79e82cfa6a0a", "1a31c6fdb22518d0d65648c8cba5faa415f3edb2b26871cf213e7fb0f842525a", - "aa5abdae6c76b82262940fce35f71ac955e1ac53b62f5053b2450a50eeaeeac2", - "9769d2296318719e2f4d5398bca23fe25ba5e9d5818a1848a50d14ac68affdc2", - "3f28d0375be777b072d17bd15fcb6d8f3d229aeeae24a4caf1dbc8a1e8e6ebcc", - "1efa50e97421c59c77e55e21d08cda0227b5d9c86dc0a03cd653cd36c1e93099", - "144c479641b987bcad9980a998ca75ae1c87454c3d9b4db11d9ad2276ee562d3", - "ab8a6fb4fc0a3687b3edc2805560f10449e4145b7230146554957002f839427a", - "9d79707f5d9d5be9cada118ecd446ef5d814fe8a635efdffb494421764eca529", - "c1225803b13d1290b386ff8bc0a52434103b72f192a5a1459a7f33041f13a0e4", - "f5aafe648888598e64dd3b27e9e03a9c7cf412e7d8432d539661eeec6ec2b6c7", - "5de9ead4a28779ba0565577f84f8fbc82f00c289d68526c1042b913ee36b6822", - "5baf70d5633815805bae7c87a948401e3dd0a978fa6e0c494eff2de89af2377b", - "71eedacf2802a06092f451ad06bf29e818e1c161f4b05343273ecbffe1d20414", - "c20b7b6c883af27fd85c181b8a642c0eb77c4696b97c3a63148e1b44b0da8e0b", - "1ca87da8eba683c9e8db54be27c4ccd8776646d3d200542e2c8af5c8800679e1", - "3c42e9f76aff34f9f68079f51476f580ad8f961d802768c15bb654387fd6ab02", + "f0baea8b03156d62824e26ef3ad3e3e443935c249a066c64fa75746855583893", + "24cf6e083ce8df481c45ab52bb5524a1c75c65ede65c1498382483a6b3804e41", + "61b1bb423c174944cbd04e5babb53b2958417a722920737995269ab6256d1d57", + "bc711c3e82d4c1628fe38f7d5fc5a7f1b16d8fbaeef4eece212f87bd3c80cb2b", + "585fd72ecc2a91a1bc0c78719a11980acfcb66b4e2863de50f036568e731304b", + "d98f0e05f6ea3d36a1496d9b7378bdf7ee24759a019673b684d85bc87ef897fd", + "ecf36cff14171a6bef15439075c75cd2ff7a6d4d53256d6fe0d6465f539f8590", + "6387c3c0941f857adcd2b0256655f79428cc07f2b7ba070a8aa2783abb7b6040", + "bba871178ff35fd9ab9f90f71f2806de133af627fc9da47ccf0ff6342697b221", "e82b296d245d9c35ae7ad8a0d44efbdc9842c550d2f348ba01442f33ceda8f0e", - "e89f1c2db473a493a776ce9df0b2d58f0b286a3dfc84f8bc04d4f7e1765589cc", - "c1986305bc63a407a1e9536ed5759382f9c1c68f38b542fcd8a4605d5731ccfb", + "f7c752272625c4559d6cd2634cdca46548d5b8d3751d7f9e2f70bacc077fe9d3", + "50382f73b340d1128dece5676acbb66f65c21bf4e3cb737a159cd06548ef436a", + "70a89297bd7df6207421ebbc2765eb2fcd86ea2d82943aaff509758d0bdc5a50", + "f7d469e2d87f1cd4f497f53e4d08ffdef4a79d618e9c4de2380206897850a990", + "c92dc55c3ff79d2611e8dea10815979504a7e38bd3407d138dcac01e20113f47", + "e9e9ff82ddb1af61f963cce33a7e4074be6511712016852ff96d44c6b75c8934", + "b2833517fbc1025ba2aa10fb6fe907bfc8c763f4985fcd731909de9589e32fb7", + "30fffe0409b4282d97613d9a8f06bb104271815171a395ec1377bc727709bec3", "84c63cc8e6839e96e1964e65808aebe7a84665a20280beb1ed66cab5d7962343", + "904591e65f474abdc3519e6ca15636433f5fa3d415528b1c17f3fc1bcaf4b08d", + "b901395305ff4a3afb712cbb521c193bad392a59135269e433ec3ded6438399d", + "d3eddd0ff8a24a8cede9b1a2003f1c3ba5470464c6d06e4eb88ef34fc4044af8", + "058ce94a0971b8f9b22c09dd3cb112fd72602e2f8d926eebcabda819a89cfc1e", + "fb430cb8fd1b94f0488f4203aa9bdcfa196f0c1e2f90d6f9eae7c1716a69e1d3", + "c073aa4abcc1ed47d42bd934a97c99721e010b2d9ab0395e5988a6696112f498", + "fa121e7e9987072e3bc5f3d00be88d2eca2c6c8054aabb77db2f15a0b80af534", + "9cdf4d2e703aa4e01b79510c25dedf13993b22568c7f3c109533ebb5bfaa79f7", + "81340d0e94c61ec1d9e6cbc31879bc02ec11c09a4ff2fee7fd18f73ab4d9dfd5", + "46d8d8962bbb4439de7b56229f4d6c4ff127c2c9b3ae65880a711a0b1d584fb3", + "127e3c8116b95a196f90f74a89fd5493065cc76356f7e0e67a4c69231fa0587d", + "bcde38c840d05e8d201b1fa0318c9306ecd8726b8a47803eb62be0c0af923952", + "aee52a52b6e560370b92dc15bdc5fbb9ff47e86671e715c4411a35148beab021", + "c4c954d832d60911284b27aced0054179fb98d93b78716916ee218be20280c97", "878cd7b3e39d2640ac24619336fb0e37118133fff3cdcef751701d7605fb1db2", - "6581ff062d4519d74e342cdf4678d65833cd183d39a1e312eb86215fbe1f051d", - "a8fac621ace5e557fb45038541d925baa0f1e943ddea9e6240b40b68be88840b", - "6c22deb9ece85af9901f7e05b1ba6b403d132617983e6881af7e2762454ffc02", - "0cfc49a35824b6ccd92c62d90fa34903d6160f7a2db47baeb1cbdbbc73d85205", - "0efe969055f2e039286e92c09f282b2bf2e2ed3afe692d5edd45de8962612c87", - "9057961d9dfbc17121fe33f3549a2ec4c0adc4940530ec1e9c10327e5b40725f", - "1711107b207040b41d4ec0bbb6dae411462497f510130b266f1c0f0d9af883f3", - "4bc53bf0d0c7c48275730ae382231fc11c751d1c884d1510638f2934fa0a60eb", - "33d471ed6546cbbc5a2336781a285382c2d4962fcf7cc03b6986b82e9f87c0d9", - "ae3152238e659376d17e4b53675c73908428581d8c9701cac8d96c46b636693b", - "ee53c0396b21917b38ef91a015a3aef54efca1d7d7d891d1083e29d17ca2c23f", - "0f108bb8ea9df50f87c2b5f2a1972cd5625ec19af812cc81f94d7c40946e13af", - "3cb27275c75ca6758b5f8cb661e11bf75565e7c405b557ea4a480925d8543d91", + "3574be2a3098e96203e6e65e55c9405c0d634870b52a29486004a1196db61135", + "04b7d9e6f50104f0adecb02f8a822f136e9f08e72764d0f2a52883ed167cee1b", + "92183a3e84a39d2c684a88907d2243986878045bd4c8f4fd3e9b1b86a7845e21", + "9e4d915792f6046a949faf57d745f5fb071cb6cffcdc4b682ff07d9d375c9282", + "6cf095fd63cc32bfba07f6716b0cd98a641a8ba045a72ae526ab1fc9201362d8", + "b265e5b4e9151d99b57c29c29fbfc1410567aaa4a10cf660dc3c83df911acbdb", + "7b852d193bb997a2aa5d1513ad6dca4298d4c16c7fe51e755bce3b909f3e5796", + "c78f43e7b08a1fc0d109fb41b365e2eb8e7a59d8c2fc57edd7bbba9320e08325", + "231cb12720ca430b6d9eb2c412bdabc30aa034ad247bcee7118375504055bf38", + "47fa3e4dfeaa332f31a1032434a5f88a326eca843b3fbd0a869cf832f6c41445", + "c49a549986fd8609ae2dd232998cf111a040f1710662f2634abddc309fcc8f33", "9729ad27bb080ec58e488c007db956580ec5a1da104c8bf781b7435abfdcef19", - "127e3c8116b95a196f90f74a89fd5493065cc76356f7e0e67a4c69231fa0587d", - "0f8f58db710e3c3f7a6140d2918bba4677e6f39f0c25465b30d2890aa6a1fe2c", - "d11aeed95f13d32835343065f87566140a9c52c0e3bbd64e9eccf1f4f3c1a0a8", - "361654e72b874d9ee8595c0687c6df4c0cfc4e74521773a3951822454035caa0", - "51859aadba957f44ec63e251d3b9ae1bbe1aabc772ace95af35c17d36c17e74b", - "9522c0107af23703cc049994e627f4eca0af1429fc139f8e1197280e32b7cea8", - "6520bd984294b68dab2ea2e75528b786d308be9c24f34a2a30c9526d6d13ef8f", - "e9e9ff82ddb1af61f963cce33a7e4074be6511712016852ff96d44c6b75c8934", - "f8ca5232573eb9de675dbfa4e8900b43ffdd49da6078eae1ff9ec14c9423012b", - "7bde57d27bb5f979d51de6f08534856b635e9ef43e22f140c5371265e47ac3ec", - "420fa2388adbfe2819ad3ac8931c9c7260deb8388a2baa7e35f461edfb05b9ee", - "2750832bb6680e2ceabe653df70bde6bb78d620cdbe6095a622a43481cb153d5", - "6d8509ea2b05a396a74f733f35aee83e8dadfec5e9984c4a79db026358191d63", - "b1d7c47d2d9a5be52d671a6deee803f4afcb437d5695807f6ac909e39cf706d0", - "c5f8f2ac3ac1f8dc1b32befbce5300382c075079a97a50bfe34bf81413837ed6", - "70bea43ea7f00a49e7f5aa8205427a7b4d494dee674f2b28736a9a6724e78031", - "055c81fd0b51ef3d502579eceaaab352f46d315af4291075efa5c6618b0ff587", - "ad352fdb134281f34dce64473154e7fdbfffe91294ff73c71a7d72d62b8790a0", - "229643ea98c02a79267ebf14b7fa0549cfb009119558c73d22711171f2a41925", - "33bae1b7a1295727f6272b59dc9fc7f7ac44224339c47a6260735423ff27b9fa", - "06c91ea1cb3549252523ca57367635a17d333c4c3acd08900d606a6271e29b1b", + "d4edfea5ce889a964ca467e4b8fcc99151040e5966301e6954b7aabe881e947d", + "71eedacf2802a06092f451ad06bf29e818e1c161f4b05343273ecbffe1d20414", + "137887df0f5e3b846364cabd18382971f760efe492a50de26014cf73c666a288", + "38b757aa4f1755caab8a1d0ba243870c3f2213439f8a0126f988829dfc8fb6a0", "30e36b1da847ed5a9e2354eaa771490cd6bfd0eeca28ec7325d26c9b03c7da7c", - "4a7bc8415234d1a6d707aee8795c2907a713768e23879dac5d1e4a8fdad2d26c", - "187fd34b9524bc0a01feb5e29e63e9d88a67f39f0ee0f44bd5d6940f37034a69", - "8b13a6fac69e4b2fdcb9e2455c4a1cac10c5718ecad39ae8c7e5e68e91ae039a", - "058ce94a0971b8f9b22c09dd3cb112fd72602e2f8d926eebcabda819a89cfc1e", - "3933d516b972d38318c3e390c32fae03d09985689246fbc21dbd8e2c4e389b15", + "50d2a28b64ffc1bc2afa5c946d0ab8224451e32b0a5601159de815626a67a994", + "9522c0107af23703cc049994e627f4eca0af1429fc139f8e1197280e32b7cea8", + "ea089ee458d15e3675a787b85ee4d97cf66e1b7b6f4478047cbb5c93e0051e46", + "ae9081be6c1330cadf4dda2441b3b8a4fae098377998e52159565a78a51a68a1", + "04cf837ce08b308f3f81f18e0c42b260eadd97ff3c5994a1f53ead0aaa189155", + "0159f7b0724766b6be7f8c6ad25c4203e71442fb874d97612e7c42c3607c0be0", + "382edbf9f7b89c23e0f5c027ea10f0c9246eae256ee40e7d96295b575eb5b4ec", + "eac617dd0312a6190d7187a99f45f5a11aa992ec178d6122820e35ec7039b0cf", + "0adaab84758894484796ab4ee28f6bf78216145ea73bdb228636d4ad23c549fa", + "a663f1d1683424a0b6f04b09382005efb99f656eed5e2e5eb031d7a4439337db", + "2e3dcbbaf130913daf9fc2285a5ed64907b7c40a6f9b58375e42446061c1da58", + "65fdb3a7251cf4e52b1853ba6782706f0dc40d3f2af2025a1e35f2e7311580da", + "ccfc5369d5177ba2e404b24360322cfbc21288cd0586d5795113bf3bf5192123", + "c6b99141ff324c2d578b8eb2ce10416f2c921bad999efaa04ef03815a92e6ed9", "16de5d1c3de6bd7b0333af4eb59cbb09e24ebb4f5e624bf51fe52ce99fd2c12f", - "2a867b5ba607fbc0034143ad9a73a9ff1d4d5d61d40c863189be2e362b8774cd", - "601e3768fcfdf06688ad78ed4adb1cf63371880a0d779709fddb0aa45f0f484c", - "26395cb7b5d71dc822e627ad62532c50ac8b7859e751cd366492dd9d8302d77c", - "eefb547412d6d823f0fdaa03fdd39134647e03416d832a3d831c68759658694c", - "c4c954d832d60911284b27aced0054179fb98d93b78716916ee218be20280c97", - "c6d1b9be236bf713eb607a56077a8d59bf95c798bc2802ef370ca90ba67b5d1d", - "d9ad70aa35c0035e940cd77b0b65e5937751bf95bfe5c3aaf8611fc6f60038f7", + "9d79707f5d9d5be9cada118ecd446ef5d814fe8a635efdffb494421764eca529", + "c001a2055fa7ebfaa9d6cd1e09fb719aec9f35a92d2512f5052608620e8286b8", + "ab70834eace0b21d65663427c2080f6f6f1068e02c606d8d5fda5d4978e585f1", + "3af02b0a3531d42c67669404f07cab092b37ac163ad17bd17de4477735158057", + "156654d6539f98b230a3f0313c45e1f4e8f34a728285ce2210da7ff90ffea17c", + "05e2a4d83f06ff5a21fd420a4c8473692f8297eb54c8a0b6bc9d441cbbec06c7", + "d9ef06f2386a1056f3c5b5ce9d840d3f0748c64b72479312ccd4017105ad37b3", + "c1225803b13d1290b386ff8bc0a52434103b72f192a5a1459a7f33041f13a0e4", + "694ab7bcc92fc9e1e2b280def27f6d48cf0481218ec54dd42f7e74f5d26c5cdc", + "9057961d9dfbc17121fe33f3549a2ec4c0adc4940530ec1e9c10327e5b40725f", + "553706a334b44c34222dbb5738ba5f1cec9623ac44527317db8c6eb667e10ff7", + "7e1da283afe6cfb54ec0b4383085c5ee7f1cc0a22bbe5085991bfe6fcc37cfd9", + "f9b390272014d0adf3eb81d7c18b13c18374d61c70f54f93e48d995a770e85d6", + "33c83495db6bbb9dbc0eb17cfc5a90276a6e04e9efd5fad4e80d9b6384c7fb97", + "e3319f9b6323d1d3a70ff4b5b1feb24de451cdb9ad87b54f6d3f05292532c9ca", + "6443a54392c6c2ffd8b71cd59e8eb07ce3d35c6cf05b6c2812eb2c8b4356c51c", + "fbdbf05c2585cb8a7d02e7c65822faf235a34a39883434f1184271979e9de680", + "aa5abdae6c76b82262940fce35f71ac955e1ac53b62f5053b2450a50eeaeeac2", + "a9350f92845058d80e21fb4d349a5fb0b199a31e49b7fd67d429c6ae77c28301", + "84bea139e6d2c92f41d369ddab69e59920f5368c9d7f953ff12795693220a70e", + "06193195d847c6439c80511e8369f3df6d9622cd278e70023b6b53685352d7dc", + "2fd111f09384a1fecbec43290961e31fb10de84669b36ce0408e356a13486468", + "6581ff062d4519d74e342cdf4678d65833cd183d39a1e312eb86215fbe1f051d", + "435b5e43605476a55dfcf69ba748122c3d0c64fcd0ad9f4f5abe5b73041754b7", + "38b00233b6593529cbbb320b52b0db7b0e3fe33d0836c1552d4a2abd8e8de08a", + "cc1639641bf3bc5c8a212637ad1e657c7bef493e5f1e4cb2e675281c9f012014", + "9cc69c7b9412cb04e26b4460c064913224e91884abdc631ea72adf6165aae3eb", + "6563ced2bfbe24cea41f48773df84e496cb1aca4eb74c08dd445e21ef62958b8", + "9769d2296318719e2f4d5398bca23fe25ba5e9d5818a1848a50d14ac68affdc2", + "fd35194e60046ea556f48fcd6e2134fc4c575b6299c7fe3a83ee2dd81b177928", + "aafe2de1b76827c82251afb66a8ab9132228d42cbdd4d05b808d1c26fa132786", + "cc79abb768ba7908d0d826756c193ee9038988fd23bd65d937c9e215b2099451", + "f11943b0df743713b484498b2d7c54c6fa7573335f229fd9c6b849de71a321bb", + "8bb683f36165c1e9d54a20b4260acbd0d490363f25a309c77df8d6fda7f8caa4", + "c88ab28b22f68d32b8c7844bc633957b1b6ef37c17b7ed2d070c7bb06959ce33", + "4b8d6684d906de76ece8424abb2ec10ee4567b0f37f1d8106e021e5a5c1aaaf8", + "d4786bdb82b223cd108ab909aaec54800f453e874820503ea262c797c5aeb440", "6889a376a90aadcb930e033f17655441e7d2746beb5b87016744160357689abb", - "9b3a4e837c7860e00cd637a6b454fc89f1bf0c0daf4d48c9265c1866323c2708", - "588a6125607ad9825b007a997b338fe9479b264df5f262fa2069fa27cfe69f72", - "f0baea8b03156d62824e26ef3ad3e3e443935c249a066c64fa75746855583893", - "eb4b5f06bf718a3009d81dd4b48e340c78b25d8c408818a4e24ccc7e4f01b588", + "2bdddf21ce77e91c61618b5c79e84f9225adef520c90ca1ef5beb67ad564ea0a", + "4d77504449002685bdcb8a475b7ebad232ed99557eed73654b41b2ef63f8c8ea", "a3be20a6b3ae3f4eba504fe66b55ecf273346cb94a42fae782c6e3c81ad3e39e", - "696c437d4b707021cc10d282c14b28de3d1a4b3d268a78cdeb6acbdd897f9ecf", - "81f97d6db3262a0c976a0c0b906aff8e2d6f7b995ac3be30dbf0bd1d61ee7b7e", - "0e65bf0411adaf6e983d34743406483d1887887463787896aac310ea992477f7", - "29f9a41cbf65b867215d71b125ec81afd9d6bc0d85742744ed12e686ee6c65a8", - "6387c3c0941f857adcd2b0256655f79428cc07f2b7ba070a8aa2783abb7b6040", - "698a2d9fe3ccad20b21442c8d4e65d32349c3f686080c5c88a800329875c1475", - "f1c862dd8ae20623c53ed1dc5499b3a70cba7c2227989239fd979e26d981cb21", - "ef940bbc4f399edfb66ca83ff25e8231085c7c005bdc2f1840c84755253470e4", - "eac617dd0312a6190d7187a99f45f5a11aa992ec178d6122820e35ec7039b0cf", + "1f6d5b4d83b115a1665fb5fe67e1e132c71cfec132fc459815b5c82d55e3f747", + "8b13a6fac69e4b2fdcb9e2455c4a1cac10c5718ecad39ae8c7e5e68e91ae039a", + "beea3ebdac12cac697e33da86c6d19b0e6a42e704f0c1c587da96c88ce2d2f37", + "187fd34b9524bc0a01feb5e29e63e9d88a67f39f0ee0f44bd5d6940f37034a69", + "9d94d3af56cb39bba4b78f877369526da57dfaa57165e408a0cfa5c78efc828b", + "b9799544b1d4cf1285ef5b452cb75a3750d282f922bbfdb3feb3922530d9464d", + "22ae5c0f070ee2ed489e9a62060586a525fc395745bff590de8663042db8ba98", + "6520bd984294b68dab2ea2e75528b786d308be9c24f34a2a30c9526d6d13ef8f", + "005bc0def684ab05b2000d4e145b0a2b19f41d784451e8110529bc39989e631a", + "f86927f01bb3c65e813286d9f9900a50450b21f3cfa5f755a5f3c6b9b91868ba", + "33bae1b7a1295727f6272b59dc9fc7f7ac44224339c47a6260735423ff27b9fa", + "fd5fe9bbece9e6104738ca999f4aeab83452e4d9c549ef6da2680f08c43ef9ff", + "a8fac621ace5e557fb45038541d925baa0f1e943ddea9e6240b40b68be88840b", + "81ffa5deb6c7c6350ca83b2b69f06e184541a1ff94de7a306c5d8c68d167bde2", + "fd2b0c536b28bd37f7a8d669bfc885210a220c8b32b20760665b0bdd5902b695", + "1b13bb59e58af9c5e66d99b6bbd81a81009c20d338098f5479cf0f7109a6c4b1", "02819667b9ae1915b7b66d8554bdae93d41978b3df169a9382c1401ee98a13e8", - "669ef97910bec0b49906da900ec79eb63a31dac9ca1b2e9ecbfb687eaac229a5", - "ed6d1e4515cb8779ac1c1452f4a36b6cc386158a3b8f7a2b1f6f9361e6dd4b79", - "694ab7bcc92fc9e1e2b280def27f6d48cf0481218ec54dd42f7e74f5d26c5cdc", - "e3319f9b6323d1d3a70ff4b5b1feb24de451cdb9ad87b54f6d3f05292532c9ca", - "c92dc55c3ff79d2611e8dea10815979504a7e38bd3407d138dcac01e20113f47", - "b265e5b4e9151d99b57c29c29fbfc1410567aaa4a10cf660dc3c83df911acbdb", - "0435239c2645719e56d36c2bbf5a664d0258b46b880cedc74224ae9f22046f10", - "8fe9de6060261e6cab551ad4b9e51c1443580e9c5d7fc491bff8965bcdcd9f9e", - "a251008af9fb9b375450c87ee7e8b8087c232209dd2e057e8602da4ecff34f3f", - "2f8ada99fc673a6c74c5984be0b9b83354d0a13710f260a8971007d9c7fafcd9", - "5f967f10274707d72f95495576f92a0ae743de88ea11097d97fb9c700c4d64ef", - "ba8c710825476d841d91ce744fa41dab2813782b49144ef4a8417a4b89957e60", - "474cce92157003183710b0263e7ae711b51df7df2e44cd1379340cd455208e34", + "2345f6a00f3124de0de669ee21cae679de9bfa4c0ba35a7108731e98433c2c99", + "c1f11d625d4747429176e94beb61f2a1021b0db4157306c4c38eb76941f7f111", + "ed619ffb79cfde97b52dbd9413a2f433cf9f682d4f344fbc441ca296a10ea154", + "675f1450acc71a3646f97720396d7b5b926af373c3b36925f67b0cfef3aa3fe0", + "ae3152238e659376d17e4b53675c73908428581d8c9701cac8d96c46b636693b", + "0f108bb8ea9df50f87c2b5f2a1972cd5625ec19af812cc81f94d7c40946e13af", + "51859aadba957f44ec63e251d3b9ae1bbe1aabc772ace95af35c17d36c17e74b", + "34bf8ef6125a16765828c96a5e70c526081290ac6b5fbfefca5e090605a41fbf", + "7bde57d27bb5f979d51de6f08534856b635e9ef43e22f140c5371265e47ac3ec", + "d73adaf05cfad3db9f1c326b6ccae765686018b08b4233ff0467f50b2834c9ec", "db216ff4a8bc4d6247358349dc5b7d9145b96518c6d21f39c7837d9f6af56c56", - "2c3c412a6a3d14a2e5e83305a2833f1f44ffa53f8f2e7485cfcc4e41a20cb38d", - "9a9631a9216991912c3c4eae5b1e166e0773ca14939d8d028dd6f83dd118cd0e", - "3af02b0a3531d42c67669404f07cab092b37ac163ad17bd17de4477735158057", - "ea089ee458d15e3675a787b85ee4d97cf66e1b7b6f4478047cbb5c93e0051e46", - "3574be2a3098e96203e6e65e55c9405c0d634870b52a29486004a1196db61135", - "77dfe12e184346c1c06acefeeaa802a897b11cd9a3687629fb0bd0d75bda5937", + "ee53c0396b21917b38ef91a015a3aef54efca1d7d7d891d1083e29d17ca2c23f", + "b1d7c47d2d9a5be52d671a6deee803f4afcb437d5695807f6ac909e39cf706d0", + "6d07c0df139d7808ae024921ca98851e94137963c1d2677088a6f7fd7d19be3c", "a77c6ceb112f1030dd42fa75f9cb6a7b588c3e206fbbc7a4ac54a31ed234f743", - "cc1639641bf3bc5c8a212637ad1e657c7bef493e5f1e4cb2e675281c9f012014", - "7b852d193bb997a2aa5d1513ad6dca4298d4c16c7fe51e755bce3b909f3e5796", - "0690c7e7f7d58f5e2dab5cd421e80f76b7827027fd340b3fa5dfcc7f2f77f01e", - "d793f12f3196b6f62a9755ba51e917154bf605a614cf11ae3b21a4592faf14ab", - "ed619ffb79cfde97b52dbd9413a2f433cf9f682d4f344fbc441ca296a10ea154", - "889b221f8f14823e3e518161f0e716c8d39bc8a67d245f367a4f17c50fa6da2a", - "50d2a28b64ffc1bc2afa5c946d0ab8224451e32b0a5601159de815626a67a994", - "4daf823a4aea9cc5c63aab5baf4e63e126a7b87c2f0e9f63fe72fb6748d8ab9f", - "7eecd3367145065b787006ba3da9be37c0dd6c25ce6acbd363143dc3e7b30bb9", + "8df47e69174ce212386ac45f45ca032c549f5d259bd006c822b6858ef968a004", + "f995c0d9aff5f87a84fe5566bbfd9b43605daa4f527c561c2dbab64a984cf767", + "fda628eb7a375ad99993fab807a9c10256be74638beac03bde0dd92e381cb171", + "2a2c6979e4e4e17158c5e8c37fd227b3f19eb44023e15852a18ccb84f3c4dffa", + "2db5d0ccdfdae2415c985be14d4453fee4b43f4a383e1e41222ffc558e5338b7", + "06c91ea1cb3549252523ca57367635a17d333c4c3acd08900d606a6271e29b1b", + "427d2cdd2dd0925b5700e992af834a187040405484d34675e74ffe1a06dfba5b", + "ae85030a6b7290cd9f12420cfc5dfb15445fffae1041d0bc897d6b31b2ed39d1", + "be9d90c84774c7436a447c2b18075e201e318faa0b09e8a153bf0477a9b1d133", + "4bc6acd79c4607c8f99d519ef78620465828b23ed4d3adaab913fcc7f80d39f4", + "d0293a95aa21ad014ffcb803661a7486a9328b62f1ca1c42a22e333959c54798", + "13d3c3a38b355357d316a540eaa34c9cc5729a5903c0bc2b4961f30867b4cd53", + "eb1f11f5d6e3e0e044527d294b58ed44acc8e4e150c1915c1c067d9334d7fee3", "edb14b6f79ea6f7b38459572c67c6ad6c74e590f2642d62f10c5110c2c2c3efc", - "3b70c21acbcedca1fbedfaa28cb0197ad39a10fafa87697d9727dfd320facf95", - "5c44a3ec370d76330c3ed037cab1591d29a6e6f08623a28ee51088676d72ad49", - "a920b7f2c4bbc5906a9931e4c65b1db4447d33e3da5fa64d3268ab23e2e4676c", - "526f34c2f3ee5be7c56b587bea87d2d359caf2be0b9a35ad10d5509b033056f3", - "05e2a4d83f06ff5a21fd420a4c8473692f8297eb54c8a0b6bc9d441cbbec06c7", - "ccfc5369d5177ba2e404b24360322cfbc21288cd0586d5795113bf3bf5192123", + "2c25c8cf3b29f69d4ac2ae8cf1b363fe1b1a737f950861b7ec05abd44dd0bf9b", + "df3483cec3ebf68c7c1db1012f3283f318462d00217b981a0053bbed443e4ca8", + "56f48b50495815a4962575e0d99fadae8ac49c6b4e1820b2f3e8f37d4a098b6f", + "51e941d4706e69d8fd01938bd959e65fb707f558e70d69077d2a06f234e0bc92", + "bfc19cbe9e812bcde1c2787845c65e1f18b30992994c6b4b594054722cbabbfb", "ee5d1b5598cfc3c3cff10f2ab32e5ce245d6291aded4addf17d1af640945ba9e", - "38b757aa4f1755caab8a1d0ba243870c3f2213439f8a0126f988829dfc8fb6a0", - "46d8d8962bbb4439de7b56229f4d6c4ff127c2c9b3ae65880a711a0b1d584fb3", - "45e8b425366f52b6b3445d5892f2eb0cc656e9c11c1215926e81d3197f9ac9da", - "145c24c294704b53a8280ad1bfe7a1ac2eb6b79c738aa9283899c6d2309e48a1", - "e4a08ad2fdb152fe225ef710cb0253bd4a885e92eb38168dad9836b8d24399b2", - "bfcdc1ae5d4b4717818d32a93bffa446176d76cce25542da0057331381ab60de", - "f7c752272625c4559d6cd2634cdca46548d5b8d3751d7f9e2f70bacc077fe9d3", - "301598e42c14fbfad17534618890846f62593ead552d2ca08bb23358df817340", - "22ae5c0f070ee2ed489e9a62060586a525fc395745bff590de8663042db8ba98", - "9d94d3af56cb39bba4b78f877369526da57dfaa57165e408a0cfa5c78efc828b", - "5c7f63b879628b5837c65e66e32eb3b6ea4ed72d502714d1ffafa788de01652b", - "2410ec7188a4dfc480796e45c7d87486d8037104d8da25fc193cdeba9b3353fc", - "d98f0e05f6ea3d36a1496d9b7378bdf7ee24759a019673b684d85bc87ef897fd", - "beadfad0756a95f260ee9761449f948b423621d3aa0292cadd6f0ac52ec02451", - "c2ff3b939c5a078b9b8d7a62696d8093219e90aa953f58731ef886465cdb7c1c", - "06d66c66a8c313fca3a69f167c7f106d110cf63de52488c20b3639a5cf6dcdcd", - "33c83495db6bbb9dbc0eb17cfc5a90276a6e04e9efd5fad4e80d9b6384c7fb97", - "c097cfccfd099f31d72cdde8b95c769bc2223756b761efa3771f84cb43fb0c6c", - "cace8f4e06b62075dc7d9571d82a9e9411f5fb06353f13c8bdef619c9dbf84f1", + "aad0c5324c277fc37e3823b1d36488ebbc8fadd621c38a4d135f54ed8e535e33", + "f71c8c65ef4784fb9a56f2d6ac1cc33c6779b1ba037f46882e3413abe782a3be", + "193b9057753cb8184872d2093eaa75cf23ad4cf589b94f2d08782410a2bfec89", + "b3c44b04cf3d06726c0009807b5327747d2b619414a00075219451f715ff5108", "3a6ba09955510536af9695feae5a24cdd6cea09f9cd73b09fb15ab66c550fcaf", - "adabf2b5acce4b8d6f56c3b2bc1a3edcd6cc235568c41e76cc92e9f3545a3fbb", - "6d07c0df139d7808ae024921ca98851e94137963c1d2677088a6f7fd7d19be3c", - "962787e7cfc0bfe373a2d1a84ed9d9f852378211910cd17f9872b7f0aa6c9546", - "bd6a4024820c9e62778bc05dfaa579bfb7758eac3f247c7ace052a6819b2d369", - "308e9ce92c44cc4a703f53fede587fbd0dbda03d9c8e28e378ceee761788880c", - "6ca496b5208e0f65e3cdb74e4281c96bd19d5d6a2380fb228b449070fa74836e", - "51e941d4706e69d8fd01938bd959e65fb707f558e70d69077d2a06f234e0bc92", - "bcde38c840d05e8d201b1fa0318c9306ecd8726b8a47803eb62be0c0af923952", - "a79c5f67d9df0482560654587f9e626b13c70286d40298eb94983dfeb54a4fe3", - "04b7d9e6f50104f0adecb02f8a822f136e9f08e72764d0f2a52883ed167cee1b", - "9cc69c7b9412cb04e26b4460c064913224e91884abdc631ea72adf6165aae3eb", - "6b89fbce1f17dcedb368f7fd8ae181bc07339f5aa66484c24e745932709c0df5", - "fa121e7e9987072e3bc5f3d00be88d2eca2c6c8054aabb77db2f15a0b80af534", - "5abb0cc63c343717c4b888aa07c3e27a494cc66603c324c84e281bdf699e2493", - "2aaa3d7afda10e326fbe98b639009175072ad33b600469e538fc036dda97e679", - "45a91a8271bcb7b2bc26267a5399f858426c8279be62964a814c7f582029703f", - "b8f733e25d9f1324aebfb5fb4eef8183b9613cef205db92ef0bcf528b4511230", - "02f79b2dd1e381d0e5bc85b8c28f438a60ca37d6ab65b8bf5577cc0dfd71bfa1", - "b91ff2bd37fecac7466d756d78d259905fbf263a7d1b7e585ef129c7c25beabc", - "e0120516f77603b503524dd22849056718a05c4278dc471bcd179b4e3dcae873", - "6085ca91e1ae99b3c11c214ee590773da455ec915fa8f1b2913564b5e1e0cb02", - "ecf36cff14171a6bef15439075c75cd2ff7a6d4d53256d6fe0d6465f539f8590", - "acde001479007c4a3d677002b308963821c668fda51fe5dc8316ac5920576078", - "61b1bb423c174944cbd04e5babb53b2958417a722920737995269ab6256d1d57", - "13d3c3a38b355357d316a540eaa34c9cc5729a5903c0bc2b4961f30867b4cd53", - "063fa5351967e1052aa2c17cf70fb15a68deb420b849924a3ee99072e5e962bd", - "5b44da5a1418e7f8b6104d6b9dd4bd8a77644cb853e19038d2e07aad1663cd9e", - "675f1450acc71a3646f97720396d7b5b926af373c3b36925f67b0cfef3aa3fe0", - "7ec488e972950b9bf1e0f6ea276f7b4fba507452575be8309418f51e957c74dd", - "104fc42198ace9e40ac5fa6a39f5ecd8d3e45fe1d28ba59605db361cf0007634", - "34bf8ef6125a16765828c96a5e70c526081290ac6b5fbfefca5e090605a41fbf", - "5f1eb2233ef1bfd5ec0ad591553d7887f77a7ecde545fe473c20374e32cc2065", - "19182796c8dffe9a637d2fba86f5658e2ca1a42473eae4d680ee7b06d5a8fa9d", - "50382f73b340d1128dece5676acbb66f65c21bf4e3cb737a159cd06548ef436a", - "1d0a9fed02ed242a8c4fa86a08c5524ae8970ba91b3467602e6b3a9482a6e92a", - "30fffe0409b4282d97613d9a8f06bb104271815171a395ec1377bc727709bec3", - "fd2b0c536b28bd37f7a8d669bfc885210a220c8b32b20760665b0bdd5902b695", + "361654e72b874d9ee8595c0687c6df4c0cfc4e74521773a3951822454035caa0", + "e89f1c2db473a493a776ce9df0b2d58f0b286a3dfc84f8bc04d4f7e1765589cc", + "1963aefdc9f37f968baf0352cddc252c2cddffc65f43396f75628d5f409fb53d", + "5baf70d5633815805bae7c87a948401e3dd0a978fa6e0c494eff2de89af2377b", + "588a6125607ad9825b007a997b338fe9479b264df5f262fa2069fa27cfe69f72", + "5de9ead4a28779ba0565577f84f8fbc82f00c289d68526c1042b913ee36b6822", + "a9f731dd11dc76af9dd18ae62ec4ebfbe03888a01c37644fe4b5c2fda1deea1d", "57e3b3fe84b2a999c43764b1e326291aa537c8b19c3d0cb0dbc777270c83a4d5", - "27f2fcb68e28ad912481ad08e32a13def5455513fd76f58be2f137abe548443f", - "ff1ac56824f6ab3cba428824f83eea2bd0cf6cf582a2a2acead8f58e7e9a332b", - "904591e65f474abdc3519e6ca15636433f5fa3d415528b1c17f3fc1bcaf4b08d", - "d9ef06f2386a1056f3c5b5ce9d840d3f0748c64b72479312ccd4017105ad37b3", - "7e1da283afe6cfb54ec0b4383085c5ee7f1cc0a22bbe5085991bfe6fcc37cfd9", - "69defb814b515fa41240c26f039750da87ee98ebf02eda27f6e1a99f701c6e46", - "8314a32e2eae7989392b1b7c9bca3e5ab24f4832dd90ecfe861f54fcd59b85df", - "5d144b80daf421a7b516450994b82dd95923e679af47dd264b2836a0aa194083", - "fd5fe9bbece9e6104738ca999f4aeab83452e4d9c549ef6da2680f08c43ef9ff", - "650c63120cb0e1f90bed748a4e605c56fbeb4c613bbb739810e7b93f46a22563", - "1b13bb59e58af9c5e66d99b6bbd81a81009c20d338098f5479cf0f7109a6c4b1", - "3bc385ca4c6cf18daaac4ecd93193c5fde8cc25d47185f0cbba263c65223d7fd", - "d04d0eef9b774170e6640f10b47352342bc1775817bd54a0eebaf00eeffe303d", - "6443a54392c6c2ffd8b71cd59e8eb07ce3d35c6cf05b6c2812eb2c8b4356c51c", - "2c025e6a212814b3293278e336d8c0a07579eb207b3afaa50fa4f868d057f81b", - "90adb304a2c762073f5f3b48670b32d78bd6b812fb7bc5da5ba534cfb34c97cf", - "a663f1d1683424a0b6f04b09382005efb99f656eed5e2e5eb031d7a4439337db", - "38b00233b6593529cbbb320b52b0db7b0e3fe33d0836c1552d4a2abd8e8de08a", - "fbdbf05c2585cb8a7d02e7c65822faf235a34a39883434f1184271979e9de680", - "b901395305ff4a3afb712cbb521c193bad392a59135269e433ec3ded6438399d", - "0174788d03a36cc41b824b01edf3233f5c65c3d0cb5cdabdbb2a5d73085aa900", - "f7d469e2d87f1cd4f497f53e4d08ffdef4a79d618e9c4de2380206897850a990", - "c001a2055fa7ebfaa9d6cd1e09fb719aec9f35a92d2512f5052608620e8286b8", - "f8d31713cb67603d5c34edf98fd214c3c80e5a05dd9156bf1ac9867a5ede79f6", - "f11943b0df743713b484498b2d7c54c6fa7573335f229fd9c6b849de71a321bb", - "67a5b07cbbe4b5db13727a0cd160014c6903d51d95119d3fbb8f456318a17f04", - "181cdd6b798e23b2c434ca6b0bf654521f25b0a5579711aff262d06a5b1e1f55", - "da5a3654e3830fcb2e1e25027bc74fba028366e8ba04b35ea30d733edda3b996", - "ca7e2c4dc03a37d5b3c89ac97309cc633b9df14ebcef542b7657b157972c17b3", - "9b85118d25266e6c2cad5e416f1b6e1e82c988f8605bc3ba4967faff8e1e447a", - "5e46d7b9dc04a377f5fb98e17e653c1291213e3bfd84bc8c97fdddcbb98e84df", - "2db5d0ccdfdae2415c985be14d4453fee4b43f4a383e1e41222ffc558e5338b7", - "6cf095fd63cc32bfba07f6716b0cd98a641a8ba045a72ae526ab1fc9201362d8", - "c8c96ec2a19a46360779516408e760513800153aa9f82daeb7904c0bbd6c536c", - "2fd111f09384a1fecbec43290961e31fb10de84669b36ce0408e356a13486468", - "553706a334b44c34222dbb5738ba5f1cec9623ac44527317db8c6eb667e10ff7", - "4f7a63dcba6af8bd95b5f3a56119515c47c6ab861ce4a00fa66fae9204704973", - "81ffa5deb6c7c6350ca83b2b69f06e184541a1ff94de7a306c5d8c68d167bde2", + "2587422519b70b3e8fabc1d57ac231c70314d3d4379173c7a4dba4b62d628fdb", + "1283165b318a3062dd9cc5d8fc387bd2a6d7a34a675d97d1a3783376002d2f8d", + "1a5f9a9df36204197fd9cd094ab9e8fa4c35f8d8485fe43ace3df947282aedd0", "991297cf398feb60dc7ee2d411ce6f281c46d39334a2b955352b43b4c0f791aa", "2277432fb4f889a564d156b1f2f4198b58230391d2a40470cf47b5dfd0303f4d", - "4d77504449002685bdcb8a475b7ebad232ed99557eed73654b41b2ef63f8c8ea", - "674c5487cc625bb6b7aaeecb75b001ca472ead48b18d4d9938a6af7d562e9479", - "17f8ff24568befa47d96b201efebe4ae228ab47d4d23df5f379c4b2329b31a49", - "6857fa4cc0e66f64edeae5308a7e12261d3f7a37d70e353931a3ed6ffd146639", - "342e9391806a6de367c1196d2b3193a574d88c945c41d700865995ad5de7ea87", - "bba871178ff35fd9ab9f90f71f2806de133af627fc9da47ccf0ff6342697b221", - "77f3aaf03a6182ec4808611712861b5831f7001b3acaeb821cca2277000a3226", - "1963aefdc9f37f968baf0352cddc252c2cddffc65f43396f75628d5f409fb53d", - "2a2c6979e4e4e17158c5e8c37fd227b3f19eb44023e15852a18ccb84f3c4dffa", + "b91ff2bd37fecac7466d756d78d259905fbf263a7d1b7e585ef129c7c25beabc", + "f5aafe648888598e64dd3b27e9e03a9c7cf412e7d8432d539661eeec6ec2b6c7", + "3c1ca37d9fa887c398cca2acb18732a01fa58b2c7455c85907912e21a0988f60", + "a79c5f67d9df0482560654587f9e626b13c70286d40298eb94983dfeb54a4fe3", + "07b7bb2b97f20bb6041b0d5b2364db32a3940ec6352c0eb7a70a47c207da5208", + "8f7abc394bbdfd29902dea9d5207c0dc667a3e2c7386697016a050bcde842d9d", + "e60bd20e33838f03c0fc24180ba67cf648cd0174ed872bc2cb751d4d07e362b0", + "1ca87da8eba683c9e8db54be27c4ccd8776646d3d200542e2c8af5c8800679e1", + "3bc385ca4c6cf18daaac4ecd93193c5fde8cc25d47185f0cbba263c65223d7fd", "06796bbd81838b5d244c84d0fab86929d104dca7bb231d852682a2f68e8dec37", - "776f95f8ea0c224a88180238d1a3e6c391bcf0662ad7589cc3a24e6efc030878", - "b143b5ec2edeec197b741dcb94f7e8f4aab0c141f9142633b90fdf192b9209f9", - "386f862bdf3c70fd3fa77ac830a1cd6d95b6989ae2f83eb1de0ce6653afad3d7", - "47fa3e4dfeaa332f31a1032434a5f88a326eca843b3fbd0a869cf832f6c41445", - "be78e7c7e2f933b94ac6448539ea741915ec146f12a1b0b5c85940d752806cc4", - "4ae13f4e8efab1ff6a572c487c2256cb921ce1a6bf8dac34280c31c77b26791a", - "84bea139e6d2c92f41d369ddab69e59920f5368c9d7f953ff12795693220a70e", - "28f978aac755e315b6c316fdee28b84db51adeb90d838da579fb986b12c1c4c2", + "c4d2c157569ca644a1cad727a2a81b8d89f60b97cc75f0fb99bf18891ccfe259", + "420fa2388adbfe2819ad3ac8931c9c7260deb8388a2baa7e35f461edfb05b9ee", + "21ea53794b0650ce1a40eb55a899b63845944a80ae436a98a24878a7d1a12925", + "bc55e2a705adf02b47d969f0477c1c742049caf6b8f2a2df5bf1f7ee05de3099", "af013d7cfbd8b67309a91737a95bf7a0f5530c6273371ce42be256272d4074e8", - "b098b3589ba14fa301d6c98f4845845f639a8cd1f58a87b68f9bafae91ce556c", - "fd35194e60046ea556f48fcd6e2134fc4c575b6299c7fe3a83ee2dd81b177928", - "ec710f748d5f7498d7b9da0582004c98aa7977f301c4e3853607029b404a43d2", - "a9f731dd11dc76af9dd18ae62ec4ebfbe03888a01c37644fe4b5c2fda1deea1d", - "bfc19cbe9e812bcde1c2787845c65e1f18b30992994c6b4b594054722cbabbfb", - "c4f12db3c7ea9bf4cb1a02e5839f384c77cda688a8d5b9f533a9eaeb6cc2b3e4", + "a5d1e6ae6769d7b2950f26628e80039893504bac8f0bad9beba7a028f641de6e", + "cbb884549963523a63790861272525223bf4fecff2555f82f8a7773cdf3630c0", + "342e9391806a6de367c1196d2b3193a574d88c945c41d700865995ad5de7ea87", + "6ca496b5208e0f65e3cdb74e4281c96bd19d5d6a2380fb228b449070fa74836e", + "c20b7b6c883af27fd85c181b8a642c0eb77c4696b97c3a63148e1b44b0da8e0b", + "26395cb7b5d71dc822e627ad62532c50ac8b7859e751cd366492dd9d8302d77c", + "ba8c710825476d841d91ce744fa41dab2813782b49144ef4a8417a4b89957e60", + "7827ea6d7e72061f13bc13018a1a8d3e3fc9b202448dab4cf90a039ff76dd4a3", + "258f11ab4103be0fe6868db4ffa597da4a4a04e340c1939e8c4bfa6360edc3b6", + "e0120516f77603b503524dd22849056718a05c4278dc471bcd179b4e3dcae873", + "64caec4a2db915012d162a3e71403409d85f13e85a55031f81a3ee88f60e9caa", + "4daf823a4aea9cc5c63aab5baf4e63e126a7b87c2f0e9f63fe72fb6748d8ab9f", + "45e8b425366f52b6b3445d5892f2eb0cc656e9c11c1215926e81d3197f9ac9da", + "c097cfccfd099f31d72cdde8b95c769bc2223756b761efa3771f84cb43fb0c6c", + "2da349e4991d8577f5d2a29bafc29cc2bb7294e41dad4c6694c09d38b28d5e23", "35d002793f64fc58fade43282554c7724dfcbfe74e006ca52a4cae36f63cc445", - "0502ac5de51c25c74aba902e24a76c7e5c4ace28de13ad3cc9f8c14323c9c2c1", - "ebc39eb660dc319a9e4376040c1c2da561c674e9cfde204d37f7322432dd832d", - "f71c8c65ef4784fb9a56f2d6ac1cc33c6779b1ba037f46882e3413abe782a3be", - "4bc6acd79c4607c8f99d519ef78620465828b23ed4d3adaab913fcc7f80d39f4", + "17f8ff24568befa47d96b201efebe4ae228ab47d4d23df5f379c4b2329b31a49", + "d5458c02dd2c310f6c32b1691e9e86c583055fcad08586509796d2ea80c983fd", "f1667fa23b9a4f52e0b8f4b43df8c29cc580038882bc339c5c76904ec090bd32", - "f9b390272014d0adf3eb81d7c18b13c18374d61c70f54f93e48d995a770e85d6", - "be45b40ae288bac756d8877e2ddbff7337391f2f2b45ca7532bd999968d88c07", - "1084238b168d9b72b6cbc7ff97d47deec60a542812d204d24414730fb453ac3e", - "25a7d86be9ebd83d933157b8a785aa555a771922fe871306921041c64c5fcab0", - "1f6d5b4d83b115a1665fb5fe67e1e132c71cfec132fc459815b5c82d55e3f747", + "f8d31713cb67603d5c34edf98fd214c3c80e5a05dd9156bf1ac9867a5ede79f6", + "d04d0eef9b774170e6640f10b47352342bc1775817bd54a0eebaf00eeffe303d", + "2a867b5ba607fbc0034143ad9a73a9ff1d4d5d61d40c863189be2e362b8774cd", + "7b76772a947ac74fe536460d6d780f10e75cf9b8c82714ad8a8f25bf3be9fe3a", + "f8ca5232573eb9de675dbfa4e8900b43ffdd49da6078eae1ff9ec14c9423012b", + "eefb547412d6d823f0fdaa03fdd39134647e03416d832a3d831c68759658694c", + "ebc39eb660dc319a9e4376040c1c2da561c674e9cfde204d37f7322432dd832d", "1ec69e76dcdf206a4f1a0f0c7628049e2669064aee42722b005aef3a95a06a88", - "d73adaf05cfad3db9f1c326b6ccae765686018b08b4233ff0467f50b2834c9ec", - "e1a8094022b30a6d30f5a5166ee06ca44f103251d3931f34695279c7bf5ce785", - "a27cc84c7a2fa588d2356e307964ddb50003db0dfa4bcdb32812b933af0e6faa", + "29f9a41cbf65b867215d71b125ec81afd9d6bc0d85742744ed12e686ee6c65a8", + "2dc0eacf9a6f4dc206d093d0c1c9570fb0760de5fc0dd6b6253340b6e4349273", + "181cdd6b798e23b2c434ca6b0bf654521f25b0a5579711aff262d06a5b1e1f55", + "a251008af9fb9b375450c87ee7e8b8087c232209dd2e057e8602da4ecff34f3f", + "9a9631a9216991912c3c4eae5b1e166e0773ca14939d8d028dd6f83dd118cd0e", + "2b559a2ba85d21f88a81d836fb1a90590ffc5d0cc9b03aff5eedd2a994051ff1", + "332a5452d42a85fdeb66e3fe55d3aeaa7e54a989eb35e98ad14569834086dda1", + "e05d64e23e9c6f345ba316b086dec26f89baaea52d4cc4f8f5e5bea4d831965d", + "27f2fcb68e28ad912481ad08e32a13def5455513fd76f58be2f137abe548443f", + "c99c4ca2182e8c2af3c00e66af5f6712d16db9d1153436ba321699b9c8191138", + "8fe9de6060261e6cab551ad4b9e51c1443580e9c5d7fc491bff8965bcdcd9f9e", + "37ec140467034946ae4494d97e562d888b3551a48c3613026ac747ba176781d8", + "d14d6737556ee8afbc2bc96e844f170c5ad657bff70cc1a37fdb4f1df0f3e9c8", "b111f71e09ab60096d0e5b091cdb40bb66fd1a429a15b6e60b1f990585e8bb7e", - "005bc0def684ab05b2000d4e145b0a2b19f41d784451e8110529bc39989e631a", - "3ab73ba9358ee8ba6d890298b377c7ff57d1dedfaa2a289393b44f92442b5775", - "beea3ebdac12cac697e33da86c6d19b0e6a42e704f0c1c587da96c88ce2d2f37", - "427d2cdd2dd0925b5700e992af834a187040405484d34675e74ffe1a06dfba5b", - "fda628eb7a375ad99993fab807a9c10256be74638beac03bde0dd92e381cb171", - "aa202dcab28dac61f295c429016209061b899241dd4d8ab7ffeee1b4163318dd", - "8bb683f36165c1e9d54a20b4260acbd0d490363f25a309c77df8d6fda7f8caa4", - "85fc4ebc2aebcb62ad86c36c6bcebc06dbb4b5c6e23b7dc7ddb65cd8b80f04e7" + "3c42e9f76aff34f9f68079f51476f580ad8f961d802768c15bb654387fd6ab02", + "1efa50e97421c59c77e55e21d08cda0227b5d9c86dc0a03cd653cd36c1e93099", + "cb6d1da75dd5f1cfeb58c3929f2c460747fbe11ca27b7ce5d41eac1771f8b968", + "650c63120cb0e1f90bed748a4e605c56fbeb4c613bbb739810e7b93f46a22563", + "063fa5351967e1052aa2c17cf70fb15a68deb420b849924a3ee99072e5e962bd", + "7ec488e972950b9bf1e0f6ea276f7b4fba507452575be8309418f51e957c74dd", + "1c29c23fe08ae7ee3aa5d0800984735f561487346735f559389a274a2884133a", + "69c3320bbe0d9afd6295e7408c03018ee0ab11ee99f687b9f768744415eb5f0a", + "9b85118d25266e6c2cad5e416f1b6e1e82c988f8605bc3ba4967faff8e1e447a", + "02d45d4839132641c6bc0f6a11b85e68fe1a943047840d577450e04f841c5719", + "c6d1b9be236bf713eb607a56077a8d59bf95c798bc2802ef370ca90ba67b5d1d", + "b098b3589ba14fa301d6c98f4845845f639a8cd1f58a87b68f9bafae91ce556c", + "a27cc84c7a2fa588d2356e307964ddb50003db0dfa4bcdb32812b933af0e6faa", + "ec710f748d5f7498d7b9da0582004c98aa7977f301c4e3853607029b404a43d2", + "75c06831b1b8504a58f541336a2c50688c6be0b0bc363f81d4ce4f98e7dd844e", + "de282018eefb28374d59be377021308d2ea51621372507b73f0f1a2b24be8197", + "2e092af8db369daa5c108b4ae3a8ab26e2fea0682d287887dba97aad7ae960e7", + "0435239c2645719e56d36c2bbf5a664d0258b46b880cedc74224ae9f22046f10", + "145e5e78d43b7332bb8ddbf991cd51765f64de5115fb7dbf3a2701813afbe53b", + "90adb304a2c762073f5f3b48670b32d78bd6b812fb7bc5da5ba534cfb34c97cf", + "60fe1618c69590f51f0bf7c233b162ef4aca4cd686719438462681d867117d9a", + "1084238b168d9b72b6cbc7ff97d47deec60a542812d204d24414730fb453ac3e", + "37948f96f6e10ab685af7cb146fd95fb85c5d713f150f7cf5241434dfe857e68", + "5c7f63b879628b5837c65e66e32eb3b6ea4ed72d502714d1ffafa788de01652b", + "6857fa4cc0e66f64edeae5308a7e12261d3f7a37d70e353931a3ed6ffd146639", + "ed7b31b0c9e40f0fa1928235e4685dd8285a6fbf9a4ad8635bff7abe7269cef1", + "5d4917445a8fa4db7d29a394c5c559d55d9a9a2cc4e46c3b49d7687b11d0a8db", + "6c22deb9ece85af9901f7e05b1ba6b403d132617983e6881af7e2762454ffc02" ], - "last_check_ts": 1783613636.5872803, - "last_data_change_ts": 1783613636.587283, + "last_check_ts": 1783745565.4327576, + "last_data_change_ts": 1783745565.4327612, "last_failed": false, - "last_success_ts": 1783613636.5872798, - "size_mb": 8.067543029785156, - "time_s": 11906.136130332947 + "last_success_ts": 1783745565.432757, + "size_mb": 3.441317558288574, + "time_s": 11302.616640806198 }, "mercor_eval": { "last_failed": true @@ -5157,7 +5493,7 @@ "url_etag": "\"1ab116918dc60569cf10f188b7b9e05353750460\"" }, "mt_bench": { - "data_fingerprint": "eceb7bb89c45eb149d9c98757cd25bc72dd3f56e22c44fe3010597ebf3607a6a", + "data_fingerprint": "", "entry_hashes": [ "5627d7062cd8e3978c2379b082783dad9d09c872091a8c3a9d817de1102e90ee", "55872308885b9d9d4211fa7c2ebdefeae8a06d5026aac35315edf6aba0439ec4", @@ -5194,12 +5530,12 @@ "582ccf46f74ac785b9e2d0cebefb1b34621781108d5d81007e476e28db15d043", "339dd74e428813a1149f760df2729610a5035234ca4d56b279e82e7248f33fdf" ], - "last_check_ts": 1783601733.2563753, + "last_check_ts": 1783734264.8814628, "last_data_change_ts": 1783518900.6594465, "last_failed": false, - "last_success_ts": 1783601733.2563748, + "last_success_ts": 1783734264.8814623, "size_mb": 0, - "time_s": 2.8338077068328857 + "time_s": 2.330172300338745 }, "multi_swe_bench": { "data_fingerprint": "89a0196eeabde307083ea8312416acf3aa63dd9ed391ef39dbbb2f0d1371f3a6", @@ -6660,7 +6996,7 @@ "last_failed": true }, "terminal_bench_2": { - "data_fingerprint": "78cfae2ca8117a31450fdbcc5ade2f81f33f7aad58e7db11989af1a2589b094d", + "data_fingerprint": "", "entry_hashes": [ "b5199e33685a721667f574501c8a3cdf35b7c0c240eb8dc2aaa4724c2d2d0d1c", "c884b75187ce9f0bdef862b51b86fede11440973c782aa1eaed81c2672015ebc", @@ -6778,2047 +7114,3679 @@ "c27fea598e7b0582128c3e4b93dd2df4611a1033a32a01220d75a393ed256468", "ee49d111c4441b049d3f76e78dbbc6831541b2227532965b0cdb9e3372a58efd" ], - "last_check_ts": 1783601733.2561557, + "last_check_ts": 1783734263.7708974, "last_data_change_ts": 1783518905.778856, "last_failed": false, - "last_success_ts": 1783601733.2561555, + "last_success_ts": 1783734263.770897, "size_mb": 0, - "time_s": 0.6099386215209961 + "time_s": 1.0067903995513916 }, "vals_ai": { - "data_fingerprint": "b8572ea2844882cd6a310e3c4bde14a2734806cc20694bfd64634a4938f6a4ca", + "data_fingerprint": "790e8aa0e939cfba53eb581624f5ac08e44cce17959adab5e0cc278cc8ab4035", "entry_hashes": [ - "f3abac4dda0c4295e245210c5f58788d3882d0585eda248cbba53fcc5b6ad6cd", + "420fccd9279714ba8b4a0636e0944467d9987dd2e370ce430acdae23cede85fc", + "8548b6fa87f0165d51eff254db2dbcf7ce780af7095b5eb4f0fa0092619f30e2", + "da8734dadd1dec2229cdc5c08a4b0834b78aac4d42f3b9071d1f543bfe15f329", + "90091185817e3a7dc46523e28be2175d7fa320fd617bf3bf65ab77f62897ecc7", + "2f5f9101faac225e2eba2e81f4c8fe838853de4062d268d57e33c95ad3ecc4b9", + "0c9df54d3eaf207cfa508c496b02851d47a12f859b9500ef3833266cc9324f4a", + "ca79c6c62c6df3f3fb9a3f8dc35d6b3f0c27842696bee06674e277098d2bfee9", + "9ab0b13a752b5426138a67296dcf4b5839f02679f4d73863268e7ef3a9d7a8be", + "e607b3ba9258e6da0c69b4c244e70dc2f7ddbc8cb3159bbb7e0993b15eda5538", + "4015a995eabf6fec8db30677c1ab2b1ca47a5e68ac787cf451e31cab599f7945", + "f59b1a939eb835d0a79a5a19112fde369c2fd6f490b798a8abf15c0ad57473f7", + "392780b5e855a3893c688aaa9f73b92d4c9fb48300709a3610bdd08c3cc62fed", + "7e4cec299b13140a36c584c68e6248cba694a7d11fed802cc84e86fb243c55c4", + "e6e8d83e4b4b6c4ee994b0fcc27769e763a41ff503b9bb23d7d5e10bb43aa6f4", + "55e3e88516d220a0bfaf15a96043942e0654b819854ccfbba0e32c063e84491b", + "665d2d8fbc57173098a368787972734adfbe322ddc0c863306c92135d7dfe8f2", + "86252d62a34cca6e1f62af7008c849d23f8f42781402afbfe95b0835d21babdf", + "b4c63ef432638ece3a97330ee57347fffbf13e340181398e0f18805bd4b1034a", + "f23e6d1515622b6acc45d60b922429578850f2135b35f06e6650995bb52a1047", + "2d6a249854c7aada198109caefa140ad45cc8cdfa91db7d60c89a9900869ef24", + "b069f9a5cbd7666456422aeec78cf860ca1919438c7285edc9d9a6e8b9959bef", + "c83002a08dde41e995f993c25e17da5ad91a595653144fee8b4a97552a5ec748", + "3ab21a1840919416479b4c97b0288b5483a87be99800a6f1e85a6928d51faf9d", + "0586e29c7a4fe60ad3a0cddf5bb459408227d8a5b3b3995a8fcd2a5295673c2c", + "8b4a5b606c84d1f1964ef322d37398631ba960052382500f08a6aea8ae9f3bc8", + "e76187d83161260404a646608eb8144e15a23961973ee55bb2d455d09dc5a4e9", + "2d8975a767cc0d5ab650f12a0e702240415915c564935f2d9331ad57d74ed714", + "13b8bfe9a470347bebc83f0eda4e934ac8e5f6914b3453ad54d13db294dc4b9d", + "e330ceb9f47689a868f0cfe979bd6a3f484208b1355fbbb23947bb666cf5ed9d", + "31592fda95b494176dc595fc8451f00bbfa874e3169755718a442b11db80951b", + "921e0ee8486e191e708c8854157ccac720f297ce7d5bb0d4e8cfee288fb51a6d", + "369a495ba82a9500d406e2140c4165c7808a06cb6a538b66ddc106e1f15c7a81", + "285cd964d0145692d1ec4f5bec22077fb7a532b2d3bab4875b3c58190bf69b1e", + "753cb35ee4847089ba1850bdd350dc892fbe6a5827bbe4bd63d13560f4353e5c", + "7de21774a9fef587a2d1c0f8c90a029dd8019052dc22521d8d814dd69ae989cd", + "cfeed079164625f28d6d079a0a8c89cfc8c2b6ca26b70678b8f4d167528b719c", + "2bc6f87770ffd5cf4751b4511f2385ff9f6fe69f20f395ea6eed9c527a27718c", + "448c626451a34f507429651aec5edb1ce2a0543d21b247a932142314f0646146", + "9d531b8d19ad181fbc615f67e632507b69832dd09e25b01968be56d6cd6deeb6", + "dc5bb223fa18158d1b823496c603d92d0cb1c3f68a4ec7a9fd23e538f73921c2", + "2c30f93e9821bdbd4991845edd4e51cdd1d9b62088528668f3365fa6cb4a2850", + "376a61f7c49c2df5361e97cbc47bcee784b934e80c1f7a7f6096cf80915549a9", + "4ce5774949472162766df39f9d4e266e912dce551ec1c2c08ba948a663a7e707", + "f95c2dd57c3bc3106097a9992c39bc384cfdbaf123b8abfa3d124be6a21a26a9", + "f380169996f8818fa9a28ffcdd41fb3551bf9e3b24806e5288a355468459a6ec", + "44d299f54b4fedfd083e793934879a9c28cc58aa6c2e8d61437c26974a511ce2", + "074db63c374fcf31ad336df19116bdf32b62d467d6e2673783c12f23adfc3aca", + "ad3468177bd0359dea5db1f7d7eb2b9b4b12773dd6f9d6681d6221efc7fc41a3", + "288a86ad047a0585108cebcf299076e915c56221d13be8a4f5858e256a8c3b6d", + "308a50e336c70a594233aeb9b2a33e4af00e6b0f0237e28518f9cef095707d68", + "55f9987ea0b7c7b9ba84658b25c4bcca4487199955231c2c05e58b9e95245f48", + "2c12f20911fac645e601011e265854d4e74453f56acc19839ff5815c0455cbf5", + "6799104a090d2a71965105696c4b455ba75e3f138d796157b8236049fce252f8", + "762b2fc5c4e101a264ecbe3a1d1ccbe9130c6778047638f5933dc6957d737d86", + "1efc86ec61bea58711274789cd2e54fceec96d64c6e206c3830df953b7ef5ae4", "7ac874aa107bd831ccf9837842937edf229b2cf07cbfea3d962080d7dbf934ab", - "a084ee3071b89393a059f18ce4823dc37cd35ac44c075adb061a99887f6b46da", - "33db90b2947590c9bfe29b276715249ec08b2e8f8280c5a1c980040c0f6b7abf", - "78abab345e65ad3d36c91f493a6a2ab19c535546a502ae6fd9b5af463f36c296", + "9f88b2ba9ba48c41b92c214ba4bbfadc565187811cdeeb27991e0b06bc83611c", + "8e75adaa8b7daf136ed75212dc8ec5dd1d63110fdc3b7c852c67af847ffe5e41", + "3b8093c2f8f2e5ee1e567486dd0a98655fb606f814b6f26a3091c3ff4984d069", + "6df7219f6da1ead40dd2ef106364d215459f051df739a81ca5a537befbfe4a80", + "a48ba209db73cbc3c7df88b1124135a33bf77edb0e30a8cd0feef580989baa54", "e7c5f858ec28ef689bc54b835062df058f0e1caaf6e6ba10fb72609934ea61c9", - "b6bf8d5a8e0ac86d267242fc2a32d27bd687dff1db831d91492e10623ecacbc9", - "3ff1e4cf77f208defbd80d37c35e541502e922ac0861db2e15ca9c4b99e16d91", - "f4d1f72945ad38c496a873e95ddaf51b0e7a6523c76944cef6ccdddb6a3b629b", - "b5421fe28d8a31ee595864ade1518e5cf4438032588e99b83924c4abc4e68d1d", - "7d849cb1fee19e70c20dc42f9c222803f63d02d272e26ace65495edd487f18c0", - "81e003e11e2af935e793e89c29fea9ec4b417d4a654548a1b4b699cdb0b75b58", - "721949870a6496848b29ce4573c958f89d0087f65deaef9aada8ee601bf0edde", - "71f99a8c50b64383080acc8231599415d4b9bf3052cff0076cda62d08fa2d4aa", - "9003bd9e1bbdc411567adf4c1968f771eecd38b5f7e4ccb29f0c48027642698c", - "83a18a933e5ad42e7d1cddf01afc66538b48042db7f9cc1b40a957f8d0d55a86", - "cca409ba122cf5447dfdfc035d2ea09b9433bd99df5f0b65726b1f13181da4e5", - "45b4e658f32ed2d317853888bacb85450b6489205597ca61b2cb4bc7d00ede10", - "68d0f30a48b953088c1ea2eabbc10014de9b03cda792457c3625cd3cf989bc94", - "f999ee53582df9bc97c520a3cc4aca59eac077c8ef8f45ecb15ded9ebb4a856d", - "8cb1c91305b3811ae811b1b08eb66a8ba744d6b46f0d3fc96621752868676fd2", - "8b1b96adbc362f36135b93b1b4ec5517fa525974256d702dddf2fc5c5d46c009", - "4015a995eabf6fec8db30677c1ab2b1ca47a5e68ac787cf451e31cab599f7945", - "0c0bbb0bd7d4c6f49b6efe2babd37f8fd41d24e5c2d1e83a04333ecb4fb3309a", - "690da01f2918633c85fade0e95f2b7e1ab0d61d9184bad3c8d0301e2349f7526", - "de4844f82b6828b987e13594e823454de22fb71f401328326f0415968def7906", - "138df56e342ded7e936a34975bfe856c40e27612d8d3ecce9b435b8ebaf9308d", - "c99eb02b7ec4c992bc620936075bb01ae9e5cda91902333ffc2f5819b1d6c55b", - "e9097524be7fcd7f250382fdaf57bfa38c12fc7ccc87bbf5e88f31dc253cd971", - "0b9e4842c8f864ac865772a4f51f7be4687eb147dfe75616ed6a83bcbe708cfa", - "bd505afb15257617bb2661a5f5e7e2f8dc3a46024290c801f386058e840a9a1f", - "b2f2cbf8e19602c009c5d1323d6bd02a638a99eeb87170e4c880344f3168c56e", - "d5b3b450c5dd1fdcf2f71f81d472fc97bcf294ed056230260a29ba4810f54267", - "f991d3b9ad54d0fc4caea25f41c9cf9413fc3992716ed3beb5fd823dad131a93", - "211a4177c8a4a047acc6146ba4940f224fc202f8e1333d9d8864ac277d4c3956", - "ef2d039f4a10fea1fae019842efe40f559c3afd486898949357bc87b168e9890", - "86087175c1cac386cb0dec67a01612e186cbcbbc710d76698660d042239bfa19", - "8df2346679f6d59271180e74c67e797f11fd9bbed8a838977cabb0e1253ffbef", - "063234a96d8fb3bdd0472e4e5dc69bc26df3e2041a403139f003d1a78398887a", - "f12e397189f140b2bf2aa87700291939c7d3ace6d23241970793f2979b05a8e0", - "0984181b5c87f473e7e0e8ba9f2a3411d9cb320d1d026f732e2b066b4e845286", - "5af7e35f00e987831f21b9b4c5065a29633ba6d5bd90c3ec22e2f596c2f133cb", - "79970e872e3a053407f48fcbf104e98eb0851a7012c29106258f5d5d8ebfa8dc", - "66e39bc3c9fccfee5735a0790386afdbfea1656f10ba45f184ce834494386312", - "e0ba84c2e369f0eef7f171d6484ebf12c14fa266c671d7d8b97ce948841afef6", - "35eed8612af542b5dd61e60916fe41ad44e714ae8166fe946c6f780395ec4e1f", - "eba5bc627b7cff53863ec9f591e97afdedb26e88bb6c5c18b07990e90f299ad3", - "4d3273e3d504d5da278368e7ba52842c9eeede959d53fa0fcee712ee74d297c0", - "a4dab94534075fc96bdf2c1847a05243f0026014c26641ae19d4e97b8b41f6e9", - "a95c959d1b5eb0213fe6ace31fc401c16f637b22443a5b1e27812b773d723772", - "181ade391d0cc3decc29f125e6ebbe9e1dc25856f137745e1a0ab480915fec07", - "09909c5dc7e424c0835a82121c5b562001ef2eef330f3a2f46f4ff40a55c7226", - "ae6b57296037e98943107617cbb572d31c8d8d5c06b0117aad7ab81de5af781c", - "73570ecc3b135cb879d56d811eaa9f2bc4727de8340ff8e71eab771445ce34ce", - "a1c26c672d1387f0b6d6dd0aae23c5e6ec491a697b30656c88763478001b9db2", - "cf3e0af1b115ff1f9efe2c7677551ca1bf378b7317f29ed9a3eeea01ed0cfd63", - "03c80b354d30fad2c1f0b0c866ddc22e3e3854e659876b31b3c5b4243437a786", - "16e5f6c093a1d94fc955efbb8a711fa44c1368be75b6257d47f3fed560cfc8a6", - "5bbe93f64b4dae8f7e73fe913c41ba10c6c17363495213a767fe9e13a1fbfa2e", - "32ffe24132ec802ac7da12ace5383c12c4bcec2845f9375a07c55cc778910ac2", - "0f115d4e641b1ea7f90ebfcb220a178d49c79d0e55943bfab475bf24a5118411", + "e932c0e5f11e82653ca68d63f4f211dc4d5c9a91dfe0544e11f1b265ec908484", + "5500dc3599a5c26a7f0aaf4c84efd0b860b4da8b4ab14d339353897e00a8591f", + "471a02416242e877b8d54043722013ac3076e36bdb374b9067036b4f9b89cff0", + "3d3575c25416a3450044015d317a40bec3b381d10e01fcc47f86f62f317b71b7", + "266d17021cb92bed094016ea30af38540784ecd5653297579ea386867eceffef", + "f4a97219b1c98c21bc472342929a32a41905dc7f93e5a59021f30a6f399f56ab", + "9bc99fbe9f43c2e25ce7946b14dc1f7f463f3ace56726e03385b34e6ba1739c7", + "ed21aaeba75ccfa30b8daff0f78c9656b3fa7b28cea90bfdb00933952332ac06", + "731ecf6cdfc1e454289623816f8e903bd27bb94da8b5a92398767e17be52e51c", + "c0ac9fd1b245c102025e7ee95d22132de8d48b8eb0dbf0c240987b693fe29825", + "c24e68df557394381f2e33f766f676a0c737a3c6a9d6b44548e8416bf7ae923c", + "8587ae58cd80c9a272afed536784e6d23110824c910153b356e93e04262a558e", + "bafcea35eaf66e410351b318ea14b491b9b19413cff66a9f23885934e2ba2fa1", + "4a6bf7c08e3e87e265bfb6ab5cd1500adcf2fe2486268efdbbf6d729364cb79d", + "1ec3a93c5937a41c4220fab70f7998015326197de6fc15cf87dd206d83fe8970", + "1f393f07c5c15eae56abfd7e04267d87fbbf8bcd752470e3abd53b26731b52f4", + "2eb4e5ee892e15e659e6eeca9f91ede795ff14a41535bbb66c334f29425e0a34", + "fa820fa57ecf285f563aadee1b07a46b0fadf41369c4e85ee718a79061a4109f", + "df16175d100d3ca8dae7be938b81730f0d504cb73763ccf3f9862ca74bea0d9d", + "076934d984a4983a807971f88d2932d9a5aa93707ccc8b7960fb186cd2c3f4cc", + "1a0958bfcf2c4d3e47fb2420abbfd125e5486a3c284dd84cdfaa14cffd499194", + "864fc7ed6c240adb32dfc5026be261c62fc4d85039b07dbd30ffcb65da96c604", + "6df1831328b362a1d27172029d49ce788c81d05a0ed082008cf42f15b37d7ae2", + "2258971469974081c934c7da561b9b292b61cf4afefdd0a5249aefd994126378", + "2d64ed83fd3d3e87692ed8bcbfd4d250ebe699c848b157e564b7f32a56a6f9b7", + "2162a04d01ccf7d5840133a6c91cc5f76aea91b9cca0026140ac7f7cfe9de91c", + "fa53682b7d8bc9dbe3ef70338a73e98ec2e830ed7accbb980835678f6b761d5a", + "23352a00ec84c4f322d6de72ef6acc21b2d4e0e8112967ce12cf6667e8810381", + "7148679b5716ab36d9ef7aa14691047e6216fe09cef333349e3986b1ba205611", + "8fc380bb9cde0dade1fb4a5a3915ddadd3ef2685e9dc57b876ab01a3ed3f65c1", + "add43ada79c99aeab1ca3425b374cccd20bf9f76ab8885c3a0286102f1b40298", + "63f6ea7f3acff771ae0aa7bcaa4657afb18594bee6352a0cb13bae93eeb15307", + "cce254f41e6c9fbecd7ac1cab424b4e87a93865c97b8eeecebace7ee40168a60", + "fbb856efa51648f6e2fe4b809aae03baa80550d8174a0f127e9988d3ad742d33", + "5c53b45b800d27f37d6e34a223e35dc10b70ad895d5bc7372d27c70e25f8184a", + "efecd6fb2edf2e5866035d78eb27d814d91e580b0f2d1bda8125b7cd5747ffd7", + "cb65634e20748af548c8451fdcefaec9cba614ab93143c95c1f00b4e02e85916", + "b484f89b8e6e2d830e17d612707666e1b8cdc7e9aa6af796c66ea0d83385250a", + "29620d06e4798218421ab1fb237759bed62575cc3c0547efcdc8d51a9ccff2d9", + "f002c8e390756809403d3827a0036d55c66f8e93cfe40a15ba7e8fb1f1f15d53", + "23064ec6c4b0989e493135790b1db899fad302689b44738e78d8378aee05dc0a", + "db4bab04de0ad95abd4a71250b9922e34f2478b151c0f531899674add9ba274d", "1b610e697cee0ca97ceb9477c6456fe299958441f752eb4731876d340ded8e8e", - "4f83a47f7ad2954172e15267a02fd6d45a1c4a25ac41bac1870145e82b36a896", - "3c4e83c5897e4f461d5fbf862047c960a9c52243548459bd507332751038cb5b", - "b0d59fac1af486e6b3d0f79111cb06dc8cbca521b969df6045f287fcdbff0e4a", - "75c5474c019716c90b29a1974908fde3171f9b6cdba0348ed15ec892ac2baaa9", - "068cc498a583b6784cc3c1d66c6edb6aee1c1d2d65c37ca5d0a2132af4cbdff4", - "df831458ab8492b0c58746d8a50c5c687d562e86caa732f8dfd53f3fbe9f69a2", - "ab6ec4759ad0ee40ca177b5e631e51ba83c28c2ceaefc88d668b1fbb6fe686bb", - "721f21b02dd679c24137a5ff3d947e46cddb045944476388ea10165f99def001", - "ff5a102f32f3622e1449ef53a4eb3b380e6beb6262d55252c86285453998a92e", - "fb0196d94b6253b46053c9e2e90f0cb5fb210ea812a61c09aa24cdb4c3e49742", - "4270f1438c0bc2da121b25455f606ac22167d540e268bafe1a229804d146a617", - "957d1c54ce3910861ce1a5992a9b019e38b5c464e8fdeb6c148337f1f1c1738b", - "64fed80b2da32ccd08d0eda28a4a06127c5c642a27470c8dba7be63ee573ed33", - "1dac5cc749a1a967e6dc1a7f7a1489b30e18aac943b7c494b32d9fa4b64b4fde", - "daba4df0331421616bbd335859554290941b437e5434d4afe3f21ad9757fb093", - "f59ecdf3df94bad6680b4e5fe1ab796ae662757da4505dc3f81d06bb31eb4881", + "82e747b875a044815c9b216dbce3b28359bf2af340527360be5ed7af7ad40988", + "86d91473025938c807205ec27ff85d3a61f562c851933a354967ed8f167bf4cc", + "64e582b0e76b17b63eb31d0b452f7ab4da003d7a33b5741308518424e4e8bbd0", + "8f0be68694d4b7f7446ff4d38330a4da01fe6c03d641dafa8d0ec1dc7f9cec8f", + "ae197ac0a8b9dc2e1ff61767833b737ef8ede20e5549e0dc326d22e928650617", + "e9517e120602f1b6ff636ff71bf14885482becc111955877e8685b8b88c2dc3d", + "0d5eee783396740ba21f259bab0f9f0b45412cf5342ea1991bfa7c9d6fa0bca2", + "a38b87285b2bcc53a557830ab11eac4e26cf540660f757e2b2abcdef1e6887d8", + "26a206b0cf397170e33d3f0563a33bda7952d013869ce393e540173e1d4981ce", + "daf6af6eef54b9064a2b7f33f05e3d2c0568d17858834f1c88c2b2018c1ff70c", + "7724523ae235fe3d3f12c6e4cfa5d1cd5d60d0a9d08772d7a5770433915ed915", + "30b4fd05b9d54bd5355c82fdc649676aeba7f928e7a03d3c65edc8ec603fe0c9", + "d4c82a1ac7568ae9f1109a2f92ec64eb6457d59868f44a59e21f18c95df8589f", + "cb945d59cb46e0384e113ab1b35694dd7bc7ae1155f10c188aae67aacb89d274", + "ef684b13df1c74e09c7ed0f46c7e2bd0aa796f495b0cd2f13a421a6d4c67cc0d", + "b23aa79b96e574b83fe4db6edd58eda7555555721ffa5043921fbadcefc21b20", + "f4e5bf99c4d399644e4fef271fe15572798a898541cdfc3fd83604aff90f766b", + "735c53ccc4f378c61f9cc510467c5c8ebdd835cc1b38f2e58debf6cd3dfa36a1", + "532cf49b0c7cf490c9fc9fa051de7d071ebcf0dfd4f05242e4d3d806233a71f5", + "e4169c9a3def0fd60792576bd8f0a0bbc26d1f4885aab73f23ca00f20641343d", + "4dc649dc9f9e23f53ddfb0040952ea22ed081cfd790d022a3a8a227dc18c0dcf", + "2a9210fc0db739cd30ca6708260ef76360d110b8abda4e4a7f65a08d9788b525", + "9eb71febb5d1b582e5b3fe62b7b8f8db8268d3cae421cb508c1e09e773f75b8c", + "8625361ec0d404519dd589fefcf011742e283eef93728753fcb44e80ac6583fc", + "316bcab85fff739864cc0e89006b29a47c2894982941eb445c05a091063ba853", + "97411051eae641adbd84b7886c5f8f9db55b0c3e16dd8c12d1d007b8bb5861f6", + "4ef0c4471d9bd22bd866dba7a3e9c71decf634fc48fe9efe4d5a43650ad24c36", + "1817eaf7924bd706c53bf69f56ee5ca03c299deaf50ec3c5e504e7dbc8fa2bd9", + "d5a0b740696b36c80ccadb74037e650260d813ff9eae314ce5ba3f68c576caed", + "d607641b0a40dad0eef33a3bad27386f03d62bd7e7c9c9104863b68e34aa9798", + "5f1ac724a5a9e9ec22c49025a833e8d54cba75779c713df2c26b5d157935b236", + "8d0bad36b93158d8c77a5b4d367c857322b4cb49a5985417aca336616da3ffc1", + "3faaadcff1a7065c94dede68f096e3cecf57be938f3eab2712c449d6a798fe77", + "d349f7da795ac6d6452631ef2c259cf9c7478f58a2e9e4eec9734a3b33937873", + "c423f681c93851aa6aec411a8e7130cf9cc4f7af5d4035d0f20ecfeb83bce452", + "f04ac517845b6199744c649e0d98137bec053d2de26afe2495a0ebd2e41af521", + "663583ab73045d77aea2001c0161a21c6b03b0da610393052e7f3a0a7754f5c0", + "d9b1e99848dea194322a96e3b02e8f9cbb46f679676ccb0ea9335da4669f4d97", + "1640416ad3b96b3219e1f3e70ae6ad6a39da64b7f9c5fe3b6a567fd370a1afc8", + "aa7c573112050b55d37b2c6df57b70ab1ef195105863f37bd9483977d700970a", + "09a10f378f089d0fdfa275db2ccd4ac56aabd21d8c885ee979203232d32ed5bd", + "08b7b635a8a8cf66cf36e2dd9d9f65998681b619ea58ea1862d5141208c8c575", + "ea4d97a17288ef38a32b4e72ab103efee785ffef7fc7dfe530f087afd9501e4b", "1d33323d0eb1ce2747a89a4cd6039866fe07a5fa511e56ed427c80a7f7e857d0", - "c48db90508c027dfba9704be510a279c99b12c929406948ed11224e2949045a5", - "b4c21864550f3a686c082f87d6398e71a642e34ce16621bb1b02f6aedf757beb", - "7a62bac519a1f8aaacab0097efdbf9792322b74137a9887c4dd8f17d05e4c66d", - "148e0b15a53f9fecab764c237e4526566aa0fb35ad98650c607989536fac3065", - "d22175ded168138d822c21b38b4da1a5f4f1f1d73329944eba13da49b468dfe0", - "0a955c9017fcb2a62e904b2f09a8a5c6ec6125c3d9b45c32e5c98146c9a505fa", - "c0556c414284e88b0e660571c74c02cdbf6e1a9e33a24a240a40da35b0768a4b", - "7f520bb1bc88eb8fbb691b83101a48500343bea1e086e69104cac733c9c45c8f", + "5bbe93f64b4dae8f7e73fe913c41ba10c6c17363495213a767fe9e13a1fbfa2e", + "d34435db094425d881b6bd4d4b028db54904ecec28ceb9c2532faa04e5a815a2", + "f34f2c2b3d156492de9b1293dbc348faf9c1fc9219749659aaf3ed5678143b23", + "25838d5859d42a2f58426ca05ff96bd65fa8f1828bb7739bac19d1bbebedc793", + "6048967a366b38226f2f939ff7462758a94e577eb56287e09f9a50ee6426c1d0", + "353f98627696a66cb296e60a3aaff5c1fa80714bba486d9fea01a6ccdfc44c70", + "5e3f1db488d7ba51d6c163e37c208b1b823f2595cc76f3c66fa2a53a6ea8f391", + "39ca4bbe5f53eadb47f0e1d7a4cdb5851a8cf21ca50f02ffbb7d01111178a3e3", + "43b10ea4bbcdac086010c7c264408754aadc71318a79822f6d5875e8c99f5355", "a8d5fb8f2b584af88d620ea31dbd5c72fb5bef20e83a592a9fb736d56e6f25e0", - "ce2a28281a34bb1da7f25356fff1b3c1b524759900cc8c77a7d296be4b6b4ef7", - "f04d15788ff2fb2695ba8975d4f0ef2f2ffd3fa3db5e80a2a05d324a61c38270", - "02b9bef8e3b3daa61ed040fbacf9b6702a4ee770e9a79803de5abb7e5ae53981", - "d8d5a00968d17e92d52ffca34b58b2eedd5fc76dac027d293d8e83d72b901cce", - "a24de10495fbb58f3aa24d4904c41d53563a5035b6e8d487fec63c5f9eb76eea", - "c1d91e14c53685414c5529e718d127993dea72d4a5a10c8a222147bc5ed43aba", - "266d17021cb92bed094016ea30af38540784ecd5653297579ea386867eceffef", - "b90de98159fa65d7ae51eac931e1e14d1a8918b779c820af6465976c4d2a3f18", - "ff119dad9481dd0f5d3bed028a421dd8c2a53b7687344d6d0df2e20bfbd86b2d", - "3be98e3772788f9418a92fa64718fdd12de85021a174d94846b529ae03211353", - "c0925d4c4d6c5bc7e089d4d81ced0f05e42a1ee6b3a761ed5f7a09d30f4a9c16", + "c10445e7805f93a1594a5eb9b641f8d21e3a9769eef025b86ffed5cc0eba212e", + "fc00b777f73c54e4bb63d296e864692fcdd0c229b4f3052b20c036d736d54cff", + "b76aac81c0d37d62e1e7ed03a0537d857b8a417040484ab20dacdaa9518d6dc2", + "278b23fccce787924a033b8ef046ee7ff0461ca1e8d37745ec08fc9adaaafcfc", + "973c45cd7050cdb4848648d3158a27b3546bd63522ac35942c5006a1063b21b3", + "ffee1b435c5ce6b7e2639d88cdbe3b672e8c904ee67dae2d369d826d892e4730", + "61e945423408cd60c6e9625d56022c4b8281afa32477fac90219185743ad01ab", + "8d958d500b4b7898e2b0ea45eac8f0fa74ec7f1d472c0696a2b0da4892843418", + "8346931255448b361d5aa8582b49a62d0d7ae506f3a90adbbae9d12e621bdc44", + "1161418c147436bb88c10bd0524b1c7cb07ef9065657f583888ae7978e321ea4", + "35945700d664c70ed0aad21b586da9b999e9fa0fb1ab11d3abe8437546a1b12f", + "19777917dfda3ea377a83c014d62a16a5f9d95236e0d89f599a5a175bc08ed34", + "d1035604804c0df6b91ef5757a70b264741b099c5bf1612c3a708a2930ac03cd", + "ab28ecaefb86e9bcf3cf71918d6f58b408b1d7b3b5c2ea916f33c65c2ea47499", + "c9bb48270559b63289655d8cfcbd33c7308b436511bce2ddfdcf27e875d1498c", + "43713935cb81be095b56a77f958cb142bdfe04c2179ec408ad25b53f13d146ba", + "bf26970b714a42fe3adf075d59f57f9f5b450544de4f6beab65037f4db097d46", + "ec775c67b60ac974211e1207c011bbe7e35fef290bd9614a252f780b40d81ae8", + "356d605abf8d2aa0210a516af4d5e2ce3cdc612a3240ef62b648f2bd98ca9967", + "65d0e14ebbd6e6b17112f9aeaf826dd7ce25386f6a3124e6d3ea7ce88a224117", + "79dc26021a0ebf89ce819d7aa481c8d6383714666af2c33bc44c78b76833a291", + "4633a7eb81a4ccc4145b3f940a4814a321327f7734d5068d89695caaa1a83ae8", + "083702404de1ae39ab3c70ccaf9ba9cf8207eaa949bf5a9b0a951b7406df1205", + "613e49ecefe7bf7a6b3db5031a563e9880004658093ffcdac78bdc0bef4b6b71", + "5c6defecace46dae102e634f2cd2238dd59d079365c1a3b1c357535b567ea503", + "e3ba4319eda692fb15e8ad483f41de076da5071ff9bcccfa689ef503a09d9c97", + "013b073b47ca49697307a73abedaf8542183444b117eb548bfd1097101a2eb3f", + "d85b85a18419a9a312ea8a18c79970990fc5d8ac99dcc5170097a45dc409faf2", + "bd52a94d40f5beb8d4f2e44f8100e956f87ba8b8215412a7c10955bcd5ab41e1", + "a084ee3071b89393a059f18ce4823dc37cd35ac44c075adb061a99887f6b46da", + "cf6dba341b3b429a18811c5f10a589e6fd12169452ea370b09dd24c487e338ac", + "cd33a12a784c824ec6d9c2de090d3b769f3926d0a6e4ac898c03310d075e45a5", + "c091c43fa225df1e9253eb0fe3155409b4dad0040f9d92fd882fb67e9fbb765a", + "2fb6c6f787f9a7450919e45fc1b49258097526cfb46a2609406d50081a44c8ed", + "e36e733e044f411c6f7d1b53b1723c70f8ba715cd85cfa2d7988de603e884ea3", + "05e8ace1b645bd17fa34270cbaa7d138536d8c58dd0cdcae9b09cde2af2e0372", + "29e766938312351dc36df07c44749243c7461c3e654caa800d62d963492e9268", + "aaee4742a8b58adad238ed7034d65c4dda52c605998b8d830a9a3c8567c8f275", + "af73514aa4c4ed1b14706e035c354787b5c941e44a3be469b1bc6aad837549e6", + "f8087586d04d525df612a1bf8b5374a9e6b71c34838662391014b46ed100b8b9", + "b16e0c0c6c66b388514c37e11f8c66027fc3a86ffb6c7af41f780fcf35a73494", + "804427ed272ebad6aa06769100122cdd83bf8307b9132fe33aa0838f2e4b857a", + "6424c899b1af80c8189db7d4af0c349553434034340d5f36706dc0f29b79591c", + "9183c3184c0f2f63d9be6a5d8a19a79c0f82c9511628ac8b52c7f6bcc5e50bf0", + "b51b59f60f89ce1a9a46222fc5077b718876052b027702a5748a337a72e25e8c", + "be6c9c9463988179d4031488e830d61111ab00d7862e354df1ac6d4020a65daa", + "44562b926cb36d56d6360f3ccf2d9ec0c263af1355cb6e4ed684dc65c7f2a5d0", + "a20d49bef4a3af46928b9b5f4c6c627ec3512021f1f59ef36484b5f3fe2bc678", + "5334e93710a92956560aca0d67ebfd8cb333460a4cede9a1119e92257c600d1f", + "2f2dc960b9cab563ce897940e865f6b59c47fbcd71126bd2c544461cf1d1b927", + "655d14194899719f2ef9786f5a1a4c7b3e16a35a6f67c05312f7497e58d02dd6", + "ad690af3ce7a22ee7fb5638fffcbf7501953f56c7c3b42337867a3f8f110647c", + "06628dc8e671a7d09acc8cc525237eb67922657c853167fb3572f5287ccd8bfe", + "ff1ee3533dafaebe36b2ebb0bf761ceead930f2d0724c596707b75d52ae4bb97", + "70a5b150e81dd8bb336c036018b6cd267b1b350f60a13c3b42eb3ab6a543c2c4", + "efef3cc5cd8b5a58fe90620607c319dd854fe4f5c7770e9475d641949d031c58", + "1b8d6065989115f8ce85a634b2f07e54713f19150bd7b4177e791480fe2c0321", + "8437094abacac6112c560b0ea3ba9248960c2b8937f6533c143d2b4427738ce7", + "bf92b761d5e0d7024040c76c7686f9fc4a3250dfb6fbb83ac511a328ef2cfca9", + "aa02414e6843d5f157a29a61b62c86775bedaa112ee45a97e8823afca8d66645", + "9d4e21a10f6739794f55053b7b234bdb4c436b5851d4787ad655fb38f3b439bf", + "1cf30a5d16447141b4256ddcdec2d38501baca0f560e76314faeff314183a338", + "a8a44dce0be536f6705c737906895f0f58031e1f2c1c94f07854da26997d35ce", + "5283a2d01dafc0a378a90c8b57dc198d1b01bcbce5f368465e19b8afd64a7909", + "2acf59e72edcbac16110fc680a3958d1ab07719814c8f28cf177bb807b074933", + "72b560fd3244dd0bac4fc2225c2df852c70d916b40ff29f339a27ffb9448e35c", + "8a347165d9715df67460bd23cfcd52a41b9ab88dfb3dfe8f606f9199d9331758", + "229466a901ba7b9ac0a06b462bbcdd3a06d9e6acae177da4c0aca039133d3823", + "7353fb43813b56299c02e9a82fffb36c8f8f27596317be64ca8fdda9534282fe", + "38b1bb8e523540a4912e3ca7b7d7b00a4a8d8cf673ab4799c46138afd4d65cfa", + "9d56f2182dfbaf41cd91ab3be743f060072c0ae75dcf40c3f4c6227815114bd5", + "5610a113f5d620901cddefd3360a6547505cd4224a62f6ae3651fb26e12220f9", + "424d01778ab37086aeca3a1f288a8746d3dc94f58b15435e4ab42f0590182f3f", + "1ba79cb73e8a2fea8d604df280c66dcea3bab9fbc77a170a77f90cfca2a72bcf", + "c5a9e8762d1d5fcdb0405e4abbee11722d9a85c4ece41a9eb6fc5de2e3973619", + "592fa50c69c954d95a84c10bbcc02560640cc080603f70e357e27fe7a26f061f", + "5da3987f8b3052d7bb881c3c8454f7c228f50c73d4a8e182c32c479c2c3e1db3", + "1083abb2ca4c961062ba05f8958694689da3dfe325dfd3a0c94ce20a6d56ebb5", + "0634a4da7757bd3262a762a1d492c986ce9458d0ecfe5ea2758baebad4a7ad45", + "93b146245986dd62f11812096c23a3a76769a00f9641e72aae3d5c2e564f8a15", + "b426f56e2448461184b98eead368dd239c11d291bfb27e07a680b49299a85a47", + "a761b23a609569b4fb28f0ffaa413b7ba202e6e3956837ec53a3b326c804d318", + "8216d2326f3371c9362f92183ed8b6f05242dfac13fa532a75ecd7368b754ab9", + "7087dff68e5ae8c835cd85712b7eba37b58e2777bbcaf8a5a4f5b40dd2a0b703", + "55959779d01712cac1bc630bfa658c79e9443c41ca9fb1a7752c96760007a5ca", + "0a03510a9950608881c23cf668e4dcbbd42fbb26509499ba5a4a57c1d9c570f3", + "c6ced1e5dcab4d715755d5611eea381fd0614ac275d2af8869d9064dc98b63d5", + "4acf7ed9d33375d9b0fec69ed65f1ad4d154cc93cbf36e60af96ab04dfce0a3b", + "ebb3b69f94d8d992f5f6973324b4d9690018d112a21b9f93c810aa77e91c14f7", + "15024c2ca9ada8b956d781e6a6c0efa654d7559c69bb3c9694537bd43f595c37", + "a4ef8357deecc341996ce8495b7c2353dfb4d2963bdb0490d4d40b5db9532d94", + "8f97751f8d4d110952ab3f4dd47507df0a43f7e711c0997c63c4b151843e0743", + "06ed1c91d03c5caf3931dfd9abe5a0abf03ba28e0d863ab9b1bd6ee34e9e6f7c", + "10cbafa48956d810859c1c40fc1c948a855d77ec4fd7a4895bd35c560b29388d", + "8d340c0069c6498dc9d8702b6900860d3e391b28d81b17dfef9574b1220a3745", + "8dff97395a2fda60f2c0c4bf2989039e2bfc4a674df56ca3bd6bedc915555874", + "9d59308fb9564e1534d455fc2b715cbf982ccca9e8c80226cfd4c997004a1c78", + "48643c10a410e5f5f1d4c95beee94b1973545116d2270628b536d9ff5b2a081d", + "25bd2a219c5afcfdb15ca80d9f3d6bcb43e9b4bea51660da9c3d761214f8c80f", + "7dca47dda4a459169e508c19ca4ce7dc46a8760a80fa223543e7bb40178cd212", + "67f08408069674424b8dae2ebb13ac56e7858302912599fb2ca3a6c4064d54ae", + "66bb81a8caed2fa2af138d47bc90265db8c3aed6b3073e56184085647c7b6eb5", + "eb96f3faefa1c2553baa13faab0b053ed947132da2f85c9a03b5e0b530d6a29b", + "fb6f2330b46cb1d8028a8aca1e00ac3067dc42266469845ef9600dc0a0e86445", + "27b1b9a92fad3c4503bb624c13be5b1c1c60fee3441a91682970c8f37123d90e", + "7ca25a135de6db15fa27448d971708c285c1f628e9552c4c1bc4b6f105400d77", + "bb319dee8fd87ca70e76eb82f2332eeebcb4b71650fc15cf285829b0979f7060", + "f338ab256d1504d511aceb51cd217988efea9f8dcf1eb15ba3b3b2384d80f1bf", + "603899e497cfef19404a0b42b256445b6e29b19570c2d4311fa13190cadcfb08", + "c07aa88345ffa211044395a85b6e2c12fdec84b2e291e0380866530d02290186", + "20952135c30df27571793e4e56c3dec703b0add9f6047fc455fed9da91f5b769", + "f3e4d5cf093a553b17ee430c4f0ff5e823722890ab1e5a7db6d43e7f0cc7a383", + "eb278ab0dbf9a7650c6a270644202d519a24312e78110b510582c5e1146f8178", + "7bfee2adcbde8982b29c6a288eb42c827d96c84fa15b6b9a95245fd110412734", + "5efa4ca18752d896210f378efc46e6990ffb49bae9a284805fe4558bc54cb486", + "768c1014df9a205198d9363f028cb5b9b9e494c53dadf02c56d1f9dfa80678b3", + "a931c6237ff8568c6576c660d173db15709ecde417f1ba5e0a49703ca26a8409", + "619c5638626d03d942dd365147de2706e73739ff0423e62533f922b6eb7e6490", + "bbe81dad778db3d45b69deb677c03ba015752cfda31d796e288f78d62ea6c009", + "7a62bac519a1f8aaacab0097efdbf9792322b74137a9887c4dd8f17d05e4c66d", + "b453f669d1f621ead3de799e18d113e14e5afd4089d46248d36b60200336c881", + "cb292043dca344f7faed6c5f773a0c18c4b49eeeafc18afb32b1ff1040cd9328", + "88010b05c626a1f460f0ecd833a23152d0352c42405cb9d4dc8e480c11a2d53d", + "1465f325ecdd86ac9a4568dae6887681707fed0ac82a6611dab3646f11dbc3b6", + "d783032143ee2a061bce204b6f7cb9a3aef47f367cd1cc23a2e8c0767bb63157", + "64b17d4045540e6e1e0034378ba8ae1bd1112bd814eaccc45b6f347e34918892", + "6c4688fdab5b02b2f655be9ea8e443d4bb68457f88f782ea30cb23ea52ae197c", + "18dab881f510bb11776bec86f30538880e02e152e1f984b0743d3dca7d2a8df5", + "5edfc98f3733d44f84418d876ee4cb1b2ddd6719010820515727159de814ca2a", + "9699145c38a951b204210ce608ba7cf12ecc9a0bb5c22b7a79a0128e89fab420", + "bc06bd4589b40a01c7775c3d36d839ab5bc378f66004ce71e14ad44fcfa25046", + "2584b73bb75da954b5b374a9e70e581d0041e1611711e1aa3b3cc1e88db4579b", + "88063e9a745a8afb9c6cbe5440269b0fe6bfc8289ad02c7f4e834531720b9441", "0ef882f6fe9c82b7389b5099045298c58fe45d54dbf3e3174017184375f3bfd4", - "e6755d0e72be422b49b6bb1530faf9fc00a583b119342156148f13ce3e396cc1", - "3fd19749cd2bbd558a00c0d9d003dd7dcc232ab5a98de389ab5fd20008a772b3", - "0ec0439924ae2e2abcf64551ae6b7d2d8285ffc3f26d7a4200846077a2a8e911", - "1942ef2491dc05ff2f51528954b3c0c6551b768b47fdaec63540f6b6e0212ce1", - "1dbd3ccfb0a10b4ab28d3b39778ed20e5bd48656cd991fc9751383c6605b8855", - "316bcab85fff739864cc0e89006b29a47c2894982941eb445c05a091063ba853", - "f3c317fb53ab7d8c4721c5ef695e8cfe67511a8e4a362b557963fa22295d7e0d", - "9c5b892ce7621cfe287e75aaadbaca0c357ab9abb485b826f542da4d03c2d506", - "4717016327d53f639038544415e62c147e43e9cac1d1434c60af243967252a92", - "580ad7045357635ff56205e71e0b46e90add2e28229dc7c4dc441b1a8d4fdeca", - "8648e2eb953ac0e783419dee072103b06013745d11e836af9122c5bdeea7ccef", - "ea3eaee2085184869d32b559ff7e8d93534beef89518024dc46b540236e76eff", - "31559089a55527a08c32e19c5b811ccab0822a5fa17ab128cfaa45805d6f8c86", - "05d756513c937b0dea3e925eac2ea91ff534f0fb844ab6c0647ed47913b95de1", + "0957334f4c57f161c0dff558d05cd3bd172d5179f23e63ec8054415c390a16c9", + "60650c7ef30d75d7cd031e47e5f86b4f2b759c3a280f7da7d6ddefc39985dc6f", + "c0943374d55786ef9df4fd1836334e7b24bbcc93cc372744f2d8cd0afd31df4c", + "41021d9e17c67fed33f11d9a7d631b41d5bfbabeb735c5a8763c6761ea620e4b", + "a83af5c102b8fb1d0aabe46c870acfa603b350ee5ed8db79d2dc03182644ce53", + "9b27f4e329c380b808b97f5baf5f757b123bc583a176a4007a830848510533d9", + "b438853f810d9d4723eaa315643b7a6265a82afb19562fa658adfabdf65b8943", + "c3ef60007f71cc5c47c4e254929608b227b8a4eccf38cb5faa56227471499a9e", + "0c045465df38e08873f54b67578194d210a4414d256c27930e10e7df9064b549", + "7df575155eefcfd7191f620b11277124f7b2177f2414098e39bf4a1a1f2b28f4", + "043d50463fea7e22ce7795d47b4e13ef6ae65a7d334347ac8955f989bf80ce47", + "0bd42630ded938c64615d454d4ad047793458600ba54e64105a37378d2790271", + "0e74f22d3088aca096d16cdccb25b1f7c78d5ae56a622d2e619a4f97cdb29f30", + "634eaaa05dd7347eb94981b0c278b3f417b01f78d416643660dfde7d73d18849", + "c9bf9d1f68c7d8c13c5743a6ae556b79e983aa81089cc07cd3c952be1b4417c9", + "ddd28b9fe0abd0efb9a81dbfcfdac4f97372b9b2f47e8e67fac3c3bbd736f288", + "0fbac779f6c06eec00babaa947a88f05b62add5f7c7d5d9b508393c3d0d8e71a", + "5a6e8ee9e1b05c144d419404c973855f8bdf086246da8c8e9e65ff79c7f1fad4", + "b43f314823102a795b59dd64c6114a6d99982d4eebebc8e3d8bd1e995c1f17ea", + "b68fa54dbd6198f4d1e9a2b86b69b44af8ee51c2cfde97e8cf674d71513482bd", + "64c0645338bbf8d8bbf3b1b3ef8f865d04b8c27723e8a976f5b2a50f73fa7b13", + "4649f1f79414be60f3fc437df1fd67dd3bc355a08a9e462761f8d29a63738ca5", + "8732d8fa859081b547edd67fe5ee50eb941e3cb0ea0e32c64d60d89e85b5a096", + "f6b76849ae341265deb295f715142e22ceb6ec8fc27607ffb86aaf0be32609d1", + "4eb4cdada1301394381efaa07c58a33eff06eb5a4d166a9086399a4bff9481d2", + "c78dde2fcd06e0282c177a84c9b8a46bbe26922de39d58edf9c078f553cec51a", + "81747535640ba6063b7dec49113a5f28f003c701b8e5cfa63ca3d76ebe978191", + "44694fe00494278175ac4ba3a4c0647199096f3a990f67bf2512b7b7a488e707", + "f5b25af43644ec6be54fe6436a4e2d566adb984e50fde33f9855a519fe6ae45f", + "d4d79d246209b247dc56eeb293fbf7ee7d08e962e71e947cd6479197ae05f1f3", + "99d523cea2bc74eaff899079aab9b187591edd07d97704159a008a36af9e59d1", + "b9677d96d367edd3f05a3a6051c2a9be699c3b6c84e3dc91dde4f5cc2e9ec78d", + "a5390b541c202dc1ba7780440021a1fe4b6f2688a75b4563990d5240bb2c53f5", + "879b9d9ae9eb6595a1dfce33e1d498f9e1059133166cb23fe555648586cad434", + "04bc1f212ad5c7afe6d9a358fc6493b6029e44c87bcbc0b1bdba8ec16e3ad1b8", + "80f0042a7546c774e32430af46804ea8f4a6be679c3b3835c58e2ef2559e7ede", + "fd7ea5c44cad5f018b92744f3b27759bb3517c3eae58f4b33bc82097d1c5e3de", + "e6930e7a63ce3f8dfb3fc3ccc5736fe5e3376be7c12660f1d077f523201c4bf2", + "8296f3607cae7d947a4c0eac7e3cd862f3f5129e42d6847d0c1fbf5bbd1b4852", + "551d17b4b95e2d846f7bbcd5f348e7dc7af14216e9039835551e64bc101eb31d", + "d8fb15d264520cb0cbe5af886055a85ebbcdb076c3b7a737a0c45a151b612818", + "596719b6d1dd61f6e972dffca4dee1ac07009d9e4545fc40756adcd1bb44e557", + "8add12505a73de0692f8de26ef585efe6e10733d008d42b956a194dd80a1b07f", + "9116cd56bfb242bf926669a09992954ab9b55cea61de259310f97133f3a21f83", "0068bfddff5f6b2d605aa628634c6c4903f04263fc2d7e4cdff44c291a98dd0b", - "8f69335e110c001e688d0a4e574dc3c3690f40492fa6544aa3feffcacd5975c5", - "2cf1bb4864a26f929b8c0425a60128293fadfb9d5be5185b54db9a9dc1755ffd", - "f7c28fd00705770e892a2e408c01c65713232aaececcac9ae48605cf0dff8a08", - "3d3575c25416a3450044015d317a40bec3b381d10e01fcc47f86f62f317b71b7", - "0a7ada0cb70f5225e7a497a83300a1b63ed028a3e24532dc43385897139daaa4", - "f634b7d27199562ad09853fb43e26c4af06adc0ade7ba877d3a44f75772ca182", - "211277fe0dc6d77027721605b1e44b65f7fc143e427b93c148950a6ed71aaad0", - "f5beccd4afbab6e6b1feb93a2423d3d869745715b23079df40d63a7cf2791f46", - "86252d62a34cca6e1f62af7008c849d23f8f42781402afbfe95b0835d21babdf", - "6feca9afba385b50b1325a51044666bf61979dc958d72e3228b157939245ba4e", - "66b32eef048ea0966de711a51e1d44b7567a8f49cc33b8afab6ec59b7dd44610", - "86db02a74ddac3d4010ec9b6846dfdc0350ea8e909511f4f462b2b2750c22539", - "58b323e6759324cd9fd40181d463814e778dc3a634931f9847eae61b1f273057", - "4fc576f965c1a3e5b04737d9d92dd5977a73b33ce6bedda6d10e451c1e7e99b5", - "97d0c15b22281cc34c54f2d96b6854ea49560f04cdde6ad4f8773ee7714f97de", - "369a495ba82a9500d406e2140c4165c7808a06cb6a538b66ddc106e1f15c7a81", - "8fa49a7318abaf4d20015f904e4948f3d381549985479aa91f62ee20ef27e211", - "cb981955255db352ba4cfbd672d066b33110a16d6a0973c30b1ebe8fe845ac28", - "eb8b07f5b9fb1323838368faa2cc0fdf6e818766965052e59e1dfbcc78ad57df", - "c588f71042d935d7b1dbcbc3813d54a09ca6c271bd4087bec50251b4bd6ed2b7", - "1d9cce670f8e88484e261206a8b6cdd0b67e10df3afa83872ad2d1a7cde70664", - "0d7d61cc6ab6b5f76c7b719e8ca16bf585bcfa02c8a58412fa5fd7db5d0142c8", - "f0f85de2b4131bc16d690623325ffcbc64c0e0f3734e1b03a58db9d883abdd0d", - "07803d6dc5f28a4fb0c43f487d36a44b522d764b29796714537cc9a40a130a7a", - "470dcef4cdb1399c262b2a3c23fdea2f3833adb1b8d9e601b51e811b6e928893", - "cfeed079164625f28d6d079a0a8c89cfc8c2b6ca26b70678b8f4d167528b719c", - "683f27d992d9a61bcdcc5f5aba69fa22a347c1639b6bf813ffb5e27731475446", - "edafdbaf39496d2e7f18ee46f85dcc2716a7897e30f76831aae7c415bec8af85", - "f74976bff78e8570e59eb65c777dc1d2d3584486cbe0f01cf9a4675599fd3174", - "ef5d92d33dce5f4bc38c8eb6444bc0de96a9725b7b52efdb39060b1f502e721e", - "b85607ad5321b4c87530776a65d60d5c5c97f2c74b1d793fdc4f7fdbcf8773dc", - "68f58869a12b8ad35349a22795f2ffb9066e99571437134aa14c97ebbf47661d", - "cd72590db2b12afbabe89a48b61fcf4cf1e0d2c0c300c1123927121ae14928f8", - "e04a627f457b1de5fc11fa9e1757d223abb31f02e943d1c8f4e43292ec10533f", + "1203aba5db22a8d19b6392a0d3ae0b627541d1566c7d17a2ccdd9016f092df9c", + "22b5b8e679a5d2086bbe82f0ec0cb4fcd745f7ed8b35e5ccb63fadb71922fcd5", + "397d5dc23e51cf9897d784eeb2ada7c0cad644e078b630f60569ced4c8552837", + "c8739d0c83faa74c54b0dee06f87669fff6273a5288fe20a6f629c745dc42ed9", + "99ee208e09dfb7a989acec22bf2783b32bd2fdd3a6784e2b7a69fc12b0a43a6b", + "d5a541b6cf1b9263a6c401f53b68b989a241d2c61557ec72ccc165b10d63b2f0", + "195184fd3f4e70d799242255e87d6fa25f51a2b32d8175757e617e33aaf89944", "c4ceee829950c14efee825ae358ade649a69193a2a01251a347e8d1fc11cfd69", - "778042da82b1c0d7068b219627d1a6f3981d1ea29c31b386a6de50a0bb7ef4d6", - "2bf9495f5c5cd95a876e54bc0862b854b73a0fe2d5f2b1efc46b3792ae72c140", - "e0bfcba5a98f2371a7693f8984f0a10db4f2ed3be1dffbc09117ee1076f43404", + "fbec42df541e112ab5180402098d6fad144aead95e648e595dbfcf9ae4781cbe", + "96d836a9dae9f603178e26ddb21e95e3e826adf7a0e3fafd482182fe8f7d077b", + "126fa39a541b8b782c521f5967de918e5784620be49353ae3bf1bfe40480af38", + "8e727d288ab873c2e01d68b5fe45afd73a0b39869b9eff8fcf8e1e741ab5a83c", + "1ec10bef41ca5ee57d2a07b14d4040c8973a2012ad408c1ac8ade5dfa7f74b65", + "237770240ca0682db1a58625b670105a9667440400548731f802daad2e429796", + "f61e687275ff718e08a7c095e02cf3a7f6fddd3c93d88c6b6acb992d42df8066", + "96445c61989ae21cbee6c4943b80f223c0dfe3e350a4daec55476549f97f0304", + "b257e9ee53e7e37ed2dcda8d2e331a8828316e08176398b56e51aa37619a362d", + "0da1aa766f0a7a2a956c830901e1ecfa20df8ec56b1dfc66309d98189c859b74", "1b193092fa0d8a5a5007f1145b61f5283a35d6cc047ed3f03374d94e748eeb0a", - "8118ea27d34ccd2524e115fb4c64d177c22ca7dbadf6e819220d17143ee31901", - "f7fe49fe5d91ccb7b6dd9661a42b0e28e21ddd5488d79effe015853fa31791d7", - "09e2f72166773b1dc303b01aa56629fb86374fbcbeb195aab647460c4b9c6a91", - "15b3520aeb6f273fdb8911f4caf8ac46bb3892d93e04188877218b7a27940cfe", - "19d0b5abc624836dd4e12338e5fd5c7283b426660ad88313ed9f9c80f40e9aea", - "57d46b29ff6e84d0e81f7ded9a8ce791ca1b03ee937c4d6e6446f8115605dfa4", - "095bfb62c040ee552dc4b855b94762d53c328b7e31266b3e7368886c073a96a3", - "8ef8774d49d93a578563625af4bd02d3616da5a7c204af0bf91d6d3c040526d0", - "517224feef584301fcaa679c3546cf06b2171d4888167d106174debe6e6f863a", - "d9f91ae377a5c3466c3254296debdf9c5b3f7e962294f0e5317fee70742ac9f0", - "f1424d6cd12e02f3c48a74a2e21b08b7d85f8b7089e5bdb2a431949f93b0004e", - "d4c82a1ac7568ae9f1109a2f92ec64eb6457d59868f44a59e21f18c95df8589f", - "bca4f36210ddf5c0231b46fc504310f97db68d7d826957c404696cd3e1da6cd6", - "28ae08bab0760cca6146019d2073ae0f1fcfc5673b79cecc494cab5ab04e7415", + "7ae3335f2cb6c7c880cc355e8913bbeffa3d50e444b970793f87d221b3267809", + "19c26eba1c187f1e6cfbb911ebf2a8fadc6ee70a4fcd3f88e4705d8280bc8396", + "7009103c6ad678020691bf9ce259e8fc293d3a653867fe815cc9114fe8a1cb64", + "028bfcbeac6c3825b9a277e83cce98e6bd2c0879fc361b0c9ed87bb77122ba7d", + "e3effe13cad52bedff1a7b827bff3a0aabbaca719f53779c90ecef23df446a08", + "44d7dd395170247a966636c428ec783a50a3f14d63b8bd2e71044f349fb0fd4f", + "6cc604e40939cd7eace80e18ca6be8d6b67cc0486dc54cb9911118df2897989d", + "3d73c2e9fc04fd4a5c39d4106adf341a9f9af729a298220b33012631dcc7fb61", + "2a61bec5faff58ce47dde7bee67df78b26a17e60166d85f5dcdc779ca8956e6d", + "c159fffa56dca586ec24e87f257ccf1479295ac5a8e14e7256282cf8cd1a0445", + "53b7de44c7d2575da3355c6e29c6466709ab420c16d211111ca01373d4483371", + "dc50b4181bf8170be84ae1a1c44b9e7446f565a2327bca896833579ba6b7a075", + "fde6601cbfc434b5a8c47a64932d3a77b68f95973b87fbbb0328d4caa3b8228a", + "69a74ae7c893f7b6a4a737c4c8b6c88d539223e8db2b1ad12c3fdda83e3f14ac", + "91ba3ccc10c777583d32b279509550cb9f28bbaae08c75b57b261dbb41a50695", + "3d30fa32bfd6854097e2a1f5a015c2902564bad933cc7f9f01df573a7d6f6318", + "cca5d6fca55f09eff433cea8a9ca3873066b2e4661ee0027fc5caf64fb53aba8", + "c88cf1f2504aba7d464a36215a977a5c334329cfab4dc6ca040f1de46b2b3f58", + "0bfe5aee56b1dd3ecb8e85c532040d18244a53f5b78061e271c9e43b5cf4d5cd", + "cb18f10bcae23c7904002c0cde096afa4a7ed8851eea827834a69956777a7510", + "43a5187181c7079c1d6ce66b1f501a84fed5503b29a8097c5de14cd51f607c62", + "3fd19749cd2bbd558a00c0d9d003dd7dcc232ab5a98de389ab5fd20008a772b3", + "4eebb6d0201d75730d1c55e244811d65b1ce006d2ed4cf5b44032ece475ae083", + "40216b42961c83ec54d5e630b51dfc11790734c927971a138610388799565bd9", + "90f387fbd94ccfc87dc94871f3cc4d60a28d213dc4d03fbb2ec0698d1d1b29fa", + "4e5552190330f123788f68c09d8a4dc6305d3993e68ec2a0c458b38b668a5689", + "47f6a3afd3ad748e9d99359de6aa2ad36c9f80524c561257e0f446f1cc2e25b6", + "2bfa46be8b100da527d40a8b423157703a2e2f057e07dcf8f11e2ebd2f77650d", + "4b7c2c9dfb6d0ecac33806d2ab47bd3f28850f1b7431a29b01656375bfa27605", + "57f9b50fb98f9ee9223190028c50e02bcccddcd7bac1f484d848beacdc6d4088", + "068cc498a583b6784cc3c1d66c6edb6aee1c1d2d65c37ca5d0a2132af4cbdff4", + "df74671a527f71637644f5ecc2a15719a5a4b96691f26aa2b168b681945750e8", + "fcaa08b16a084e5c79153c5f4c7d2e8873f96145bf84af6ddb362c666d878918", + "07117d13ca628cb76702ef88f4cc29585eb5cfa8197e3f30f12bc750c6d07556", + "dd5f914539e8c5f41fd472ad099d7bad67c23eff9d7df6c30efb1e21bcd044e2", + "ce476eeceac79af0881790c44d5a27ec51f0ddb96cc286bca9f48a4b7cbd05aa", + "a2f84868671755ba207bbc41ae7975116e407ff8e332d9afea9e277fc58b0aca", + "9c7532184da5aa82a19538eda318c5a78a1ffb2ae847f9b085c124cb8dff994f", + "095013c4382ce360c1cc617fa42f98f48fd0595739051d59b690d760d710beac", + "3aed9b96472a4c516616cfc3479053ac31bfc2b4a72ae6ba930aba7420479c74", + "ada45626c61490618adec63941bdf6c064f97e0f71ec90f38526010cb685dfcc", + "318567918f4f34884331f88689f1aa2e7a69df8047d50f5f0379eee79c2e3216", + "4f83f1ee24faa4fff484fccfe98f7220813d8e3fa07bb602b144086f27428aa8", + "5bc94b73df718233acebd5227a3c66a6dcbf5a4670ac6eb6e05e73be6004620f", + "62e115a100686faa541c893bf5b8e0faa1dcaca43b157330e6b3641cfa843a9f", + "bfd3b12f98c21a905075c04586461e650250dab57147a389faae222babaa69a2", + "27d58a37f9f7448ee759a1abc50844381019844acede2228bbd264bbcefacce4", + "8a3fe2334f029b3ab49fe559296086eb0adc16176221e1b8b65787f5266399a3", + "55161f81eac7648d0a7df9e24df9cb9c28eb3bfb9de0db317f1e8c00e40fde08", + "996a30c43824012e201f2012ef8afec826ad02e47289526c457f3e50aa54f836", + "5af7e35f00e987831f21b9b4c5065a29633ba6d5bd90c3ec22e2f596c2f133cb", + "c883ffa6b336604471d5f211799ccf9ebd68248775451ed4a6e65ca963b8feee", + "5b903e9cc75881e86da0e05369f2b957669eed40731620829365a909d26d5be9", + "39b843ecadd6a7e811a7b1695352ae0258b86b740f823ba56eef875c6044303b", + "658d7307c96c62333fa853fad6818a762e08b092c1e98da450d74f9d308b247f", + "8fa49a7318abaf4d20015f904e4948f3d381549985479aa91f62ee20ef27e211", + "3e9870cc80d578e863be4aba82b2201d0ab01952ca3e49c8d4f6941f9ff7e4e8", + "c777c59f299a13990643821aded8e213b8bf14d998ebb2d2fe90f77d60d9512b", + "ea3eaee2085184869d32b559ff7e8d93534beef89518024dc46b540236e76eff", + "be8183534abe6d0efcc5cab5727f3788b7e0abea7cdb98cdaa88464a0ade08d4", + "48fd70bfcfa6662323d9195a6042079657ef7f5a45020607e94287873f73b0d0", + "09b62fd7885c9fedda6b06cce8978abdeeb548f72fb78c32d1b88de21d9ec17e", + "a3cafdadd64e579ed58114bf6efae676531105aa3a70c1050035c34715195ac6", + "cf787b9ab78428cb7c49d6bac399723407f84173c69931f9a0224fd1bc424fe1", + "bf5817fd48fafb09e9e21b04a59ae37629ee66a71680b7f33cf5a8c313b30de9", + "3c19079d6598a115b2891a5984c6140cfba5d8bb46c69c6e663db3117b6b41ed", + "95500ec0aadee45f2898a79c4280e33c361e8e40e30ca5c6f858c49b612349f1", + "e4742681f7476b2acd745a905a2c55a2dd1cf8657dc09329b0605f852b9a5225", + "b6d57354b2857aaaa05c5b3d938d29e31131960909dbcfc6794c192b40387f79", + "83c16e500bb45940293b167460241c675cde9b3448392f47054e2a88a355e161", + "2b360282b9dc662300a9f3a4da09ef2fd921b8c1451f393dd59ebc95a244114f", + "265ede8956430ef4f382d9f7add5405a47a8ca60f7509866fe3891ff40f3eb35", + "f6525b1df9614668a1cab47b1dd85c632d55a10dcc088c18c5235fa8fb56c31a", + "8524f0890b1fe5b02da21c0c2dd0090ba032fd116c6c463278ec5a6566d5b266", + "30212a64ecad629c5cd2ac67f55d8f834cbd7c0656c43d5705ae98be4e5ac07a", + "51a874fb6958a29602408c1aab307ff10b0eb1bb91a182b75e32005ebb82bacc", + "10800696395a2ed2acb3d70e1dc2323ec4449f1654c2c465a6a242773e387325", + "03fd7eed6dbc56aa0c62ea3e45d999d0e67d526d8e05a36e430ab00335b8e0e1", + "5e3643c5aeff24a1d85aea753b48520e787ca474d8f62e26e92f62840e7f37be", + "ff53b33202dfe350d5c479a15ec20e3098be6a10f0b40e6cea62d56363e19b4c", + "211277fe0dc6d77027721605b1e44b65f7fc143e427b93c148950a6ed71aaad0", + "e33517ec69a6d577f36fd6352328534ab5a50c22d731c6e904928c9f02e971da", + "2110c81571310eeaf9f2e6c80ea8c3bc02fb5b222f13181ea91741924e85a03e", + "8558e189651fd2c27e9737bbe007f319b86772622a71e228448cd2891ef6ab52", + "5d7f8be430eeee9f0c88856372468761042cd4bc4f0d58b1499178438d87e11d", + "2300dc89b535430105975eb8cf9a0acab0208eaea958b3ab434060030e1422c2", + "47394d9ee47f1a0cf66e2c269aa36210d1279df7d65bcca1ea8c6a1aa994133f", + "3f2c66974c726209c62de5278b500637e169b547d395bbde68149513149786b1", + "9e530113f6346550721a80fe90268e3ba3877eb00093397fd6f9964adab28dc6", + "ce708238ec9c7754afe0abbf2c9517d5a76ecf8fb4330591883f78deea27d818", + "7229ad59feb041afc228e5409011f54645245eed01d83089cbfefb5dd8ccf43c", + "6ab6e5c72b37ff78c4f379c23b7d0468d7b857f435e39052b970579eb4ab556d", + "3610634beb3556d9d99d4c94a3c0aa457a2e867520758d5a7474f7d45c817cf8", + "47fdb94061bf6ba64d5ef98d5c71be4c0846f908a0b2eada775d7d7aef54c128", + "469b210dc5863c1ef100b24bbc40ec1306ac7ff828bb388fbed0b95cd7488fc7", "aa549b8760bf23e3e2525323f5dead9a26022b4e7289ce0026b7736003740d7d", - "c091c43fa225df1e9253eb0fe3155409b4dad0040f9d92fd882fb67e9fbb765a", - "d7719047feea91cd981fe925fc01ddf943997881a5b28bc0bcd132303f70b61f", - "ea735522d706a23dbb9a467a8dacf70f2c48f5db4d452605f84320e1dbadf46c", - "8346931255448b361d5aa8582b49a62d0d7ae506f3a90adbbae9d12e621bdc44", - "39d1b55d940b95165671d53aefe3ee51ad618bc3868e181aa2cf8f6712b990f0", - "1d3ba773d409d98f761f446cb5f2a768b916347fba3c2ffa1c68316403636815", - "6b562f260c66685a8a4f0e772534acafed71cdcd3896c3ae00e2c89a1de44a39", - "261a8471bccf590bb0b8e9c71c25b04b48f05aa59cc8bf81835cb619e7263f81", - "9adc53b2f0b39dab4c9de8f4037921a91604d2bd626be19d3e0fbba849ea6b82", - "82a7a99f39e14c5f6533023d3ae8ac9e377a618b8c9fcafca73456e212a0424d", - "2d4f3320a3a9a9dd8f2117191d83d4d2cc10a01687c812e307b2b52c8ef8c809", - "4d13625a8629ce9976d832c4ac8f4db9413b750c862fcf88cde29f76b8c5985c", - "add43ada79c99aeab1ca3425b374cccd20bf9f76ab8885c3a0286102f1b40298", - "0b2a4d877fdc9454e854ce9dfd1bf176d58b8a289bce54506f2e1f356b7983a7", - "2184b4882b7820108abd116fa8de8d0db455a71ccbd13676d9625d19079a7477", - "5155d494f24ec55d12477f93f89ea5b209bc9978d75c55951b96187191103091", - "2c3337c0e109772f3bc4e59639a045a6aa35726b0a962a93245e6ae0fc0c5267", - "282c17d18e0b935379f494264b73dd69058bb8e3e27c51e49e45c194e5e30a1a", - "9220b0f709d0a4462245500d9da90d40303cad69f0a559ad5d4118a407fef11c", - "70e13de08d15e40fdba4d6cd7abc5848c6ea1cfa119d1de7d29ab2d94aca7ea0", - "311f8c3daf14216cb0ece44bf9ad4ca2e1b235ad2c4e77f69cdc8d2848a1b60d", - "8cdacfdbf1887e4d161f1afdc515ca80aa17929968ffd84730228788e47bc414", - "85dfda2b834f5241592ace737c86bfef353ba8d7298a4d05ba677f5056aa592e", - "2f44e645976aa7b9cb2047e99242e8e9de2b1dde989e1c90d875d11fda4b1d5d", - "8104bc37513489f037f6db97108cb69d3c438fdd883927eb78c4f6d612fd1782", - "b49f09e999021496f8322e712087778229f2f295fe3001e3efc66e7441647b67", - "16c06b53dc2460c4ba70e42fa1f02f759122cd7a5f22701a0acb42c56e9a8834", - "92f3f233cf92490b67faaede66f9c79f49e1028706fa4219cc29295f3fc14098", - "5943d4100b5ca8e9fa54fcdd13fc3b2dfea27f65d096c3faf1145de8be30c0ba", - "9b484189743f9ba866a15afd075478a907801b65ab9a13348cee07ea1a90a057", - "cf8714f877f13ad771aa098b8cddbed2aa0f5236d5e905d0d891d01c7ab49645", - "d52fb87e1ec6c2167f2615f6f286f070be6264d12301d76ed06a55e67d33cc21", - "d18640287a6212454df1998336272d3a4060505e881cf2ac29ffe5d1895eb0ad", - "b4af9eae95bb5a2de122f8cd9dc2822a730ab269a6c11cc992e2510933853d62", - "9898ab7e3efddcc6bda9eae6143df5ebbdb7357ecf77584bbad279bb45c349a5", - "2402708411c33dbd8ad2594a5d8d7f1fb3a518ceab035af539967897a982edc3", - "a3c551cda78892f451872ac53bd2b3ed26446ef12b44946dba230c7091d9b439", - "4caf1ad209e31f2b6d9e4ec75a5a5e3a6711c768f4f4108c0c70bd6e7934a933", - "1c07610d40d4e2fa6801b38e17eac90298fd3792ec2caae732161c636d8a3de7", - "46e9d3fb1c50b775524c05fdd0de3f53f9faf090c1abe34dec2360c0c681965c", - "778b656b96e0e8c16d6309a2ef7fc85761c94fe3d78d2418d7302fa3f057eb99", - "4aa7835a7ab0fb3f479ebd8e3376162f64ab324587191d49eb3d2ce1e0ec1274", - "2c60329e905ac11ca44bdcf82337d3e5409f63448b7f3e8a48927bedb9e8df0f", - "1ad6d9052872281c6e03320b29b03cbcdc99f8e5f52dd35ca938cb050086aba2", - "53d52feb7adebe4052f8cb5f3290f3e669b4f004bd4b8cf3452070dd3524af81", - "0cdf8772dd2d6439e05fbf8f5de04d1a71612e2fb2c64e70938cf815438704a3", - "4ca6bf1f1feb4ef3840680bf3925b63daa200d3919d60f091e89ff0cf6620772", - "ae6b324b4640ff1d5f711adadd03897ef3136eb27f1a3b5c338432084bcc8ef0", - "0bd42630ded938c64615d454d4ad047793458600ba54e64105a37378d2790271", - "87c9a6edd8fa3576423a761c768ddce93155d208f2fb75ba2ae1b26daf3211f1", - "627b52de262594f6f1aea23327ede39266c1b3637486e51af16329a6b0991767", - "38c7fc5d1eed0cc3ccc7d92306d8319c5c316fb073e16c4d9c8ebce84ea89452", - "553f8bafbe0683714fe8d5b6a45d96f577fef5001da7f4b8443e20d62df3b94c", - "c17b17b088d85b709511439a38026e48f94cb3cd10203c98497ba3b1256d503a", - "fc7cdab5dced8936794237e6630d2bdf6ba0876da9e0e180edbc7f1275a6db22", - "cc2b33fa3703ea8c2d2a87b3cce6b28d8c2e807dcc1c5190ba88c89c701ffda6", - "ce67dc5889ea3f32dd5135d6af797eca24e8db6658a330635e4642a0e4824846", - "6b8b826f1b266ef82be5fe3f104d21eb4b481cd9280678839ae6f73897637612", - "c4a95ac3ac8644689a5b8ce1fd6b6ebe50b0e3590f0057371ac1b43002439025", - "67f9014caaaa6cec539e99db9f72b56308797105fbd7a3b6e65e73dd3ec94bf2", - "834f98c2579bdb0d95a852de2fd4261a40e59bbd9621a6930e3a1d5fc36e4714", - "b7cfd158924083c4a96437b829659f00dcf01dc1fdb0ac34e55b898c88ce1d85", - "41021d9e17c67fed33f11d9a7d631b41d5bfbabeb735c5a8763c6761ea620e4b", - "0cfb4a321961720be24a19836ec39028db63071b1b20f452688393de17f5d6a6", - "bc64bd1327d88767db8f1dd98d374baa95bbb1b12543ec8c14cffe0db7ab237f", "170b11f9131f2155066c1e04293dc3d654bea7455c4a32cfcacf75f0de30eaaf", - "105ed1cc2d250590e3d0a982a3ed8e30e71d71208200debe784c040f37001b24", - "5943a54c5d1d8b09e5fddb05b77a5c6916194bb2f49ce9a9bb19f3bbc52b509d", - "61c04ecce9c877b3f7a9a02dff4b5176b3505c9a050a392b8d9f0e3406bd5f09", - "c753a6be718e4c97d8d312a5513c933f478219cde7b429afc5eed4f12de93fca", - "01391aa6ccfdbd4862c42a0e21a7e63f36e6c49f8e90f63c4d173a401dfda577", - "0ce2f8db431adf460e619931c4846fa4cd1d2d423a1189c18afa80c36563ac2b", - "d0e9ba139e88c3fa037da4a7032e2cd2ad9e97b6a7e04591aa5fa378c458f634", - "af73514aa4c4ed1b14706e035c354787b5c941e44a3be469b1bc6aad837549e6", - "31a2064c7a46cca5d9a31b4ecb3148fc22d41a512cf5bfbda2d764feba663f97", - "35945700d664c70ed0aad21b586da9b999e9fa0fb1ab11d3abe8437546a1b12f", - "e4f2e8c2845dfa9fe9bdcd928b4f62738716b15bd94e77eeee7ef1e694eeb288", - "1bbe1e85191480c8535e5c7806b360fab7d5f9f25d9b0865a190d2afd447a6c5", - "5889deab3e6d85c6f2969895e076ca8d2b9b58c494431774c6cb21622e13e775", - "9f1df2530972f3fe7c85671e97a3c391919d5fc15d5cf663dffe4cf3e0422ac6", - "1ec10bef41ca5ee57d2a07b14d4040c8973a2012ad408c1ac8ade5dfa7f74b65", - "aa90337da124bc8d5a6c663613b9591e3854235cd88735967ce8b10f6aba71a1", - "3afb778e3278269af010255673e5481821d61e312a1f2de6e67ab1dcf20ce45c", - "6d89669d80b6049ab68b768fe9e102f8d02a12a5cf77dc7cabad6d9350d22c09", - "741883fed4a70e4d024696e40bad9303231f27576316167444ab23b1e638c3ff", "ffda6720385e8900c5f24f8f652490ff5c7606b23b993fcb2ec2448b8cdee13d", - "816b22377df83b654e2c936403438700f3c1436d108c9528dd7fc8c2612c8a8f", - "56c5fed416363040190963b0ed68ccb0134caf7895ac3cbe9bd40bbab6a38968", + "68a6bda954ae6f4bcffcb4f4a3ac7fcf623c379464ee445cde79112d49799ed9", + "032d90eaa8e1c77657a8081fb8a1540c05580b8f4ba901f5505eb5ccd178980b", + "46e9d3fb1c50b775524c05fdd0de3f53f9faf090c1abe34dec2360c0c681965c", + "2e239ce039a3402e489d04af25f4f4ae47b5ce706905462ea750cb58dfbda3b7", + "f235941cb31f66bd842a99ba202104554c9ebdecf3048e22286092fa865c3037", + "34b0f20d0e814387817d78d9503b8c596939468fb5b0c4110f37d4ec6a74f3f0", + "6a59ba312894fc12435d81b5b9b380fca44c6c0d1475de6ba3cd337d9a92082b", + "0ec0439924ae2e2abcf64551ae6b7d2d8285ffc3f26d7a4200846077a2a8e911", + "b1e0e06466ab65719ccb67f3c69beccd8e1bbaa76f4816f1e329981f2042897d", + "e607521321e73ac2028cdc7e56c20dd41048a87dcd1c431990f813d4db06df34", + "7e02489f48165bf418e5fc3ca8818a0fc4fd8f761bd97b2857269d3938352220", + "7151e39ef77868ea68a527151c7985422102689a8f633360797bedbf9c62a84d", + "2ce4dbc4c98398bb8a2c4bac1c5bfafe7c81bf24d631f58e606124d773f5d426", + "b386b5ce57a8c1d665597c622d570af0cb30955bb701c76f30d0a0db6bd8e006", + "0f7d85f086ef35423238e069d90304978a8fe197c98218fc431975181e7617a1", + "8df9cde9a97129015545e8d02beb3dab876c99b0d94b8ac7a24f7c176435b200", + "ead1520248fdbfd42fb2a6068abee0a7bb28e96906959ecff49be57bca6a61ca", + "e8051e036c391c9b460a0ce86d651a6ebb8d65382ec42722363f934735f23e41", + "d640874d9d0b2b2861af04d1515cd9d5cb0b2985356f332e8358167f92d67491", + "8e17801d61b50bce2a8c3003c8371ad657574e58ef8d96ded613be264c51a468", + "e3930472a42f5c55cfc5cea2dd3ea0027f14a022c65a7b07983cc35dd36f8fc8", + "82bf7442462018b98a362da6df19982831dc7ef98a1cb710bc6d1b3c5a9fb9d7", + "2dfdc1307806fea7fcabaac061a6a107fcb21c97f997aff4caa8ee9389d7b840", + "9f2654b711e1f909bd67a39b7944f1e4cae7b9ab0d10f985eb54af57983a719a", + "1f1fdcdbd34ea16b3423e682d3bcc345e5f4c2e47a5ad4b1cacd9a0392d7e95f", + "fad9de4475e90fe1e1a3bf1438d1ff98b7c3cf34b5799aa8a2b23014ac9dba64", + "577e8028df88046e3e9add092c323f4e9c33316b0603b4f917bdc62e3a5acd73", + "137205e509350fb2db9b3828467fc2d55cc2bddf877e3bcbf47d45e2a92b03e1", + "088d7cdc16c527038ec77ef96c96e3f2fed15ce41f1bab52ea9372fa10f5b2d2", + "ac12669ca7ca52fc108e7136956e56bae6d3ba2390d890c543e7bec53a185abd", + "37e86744e89d5dbf399fc7f75cb8bd19514bbb61004b30b66e78b116a2bd6a17", + "372b1859e673eeda87c31dea15dac396f62bc8905a17191da0aae5fd60f48fa3", + "8104bc37513489f037f6db97108cb69d3c438fdd883927eb78c4f6d612fd1782", + "0b0ee094a49df49caa0628f8ded03e92cde0ed6479f1cc2ba47db5814e1b66c3", + "23e1283a3e83fafa8d42d34cc047cd744be570d4ff7e4d0fc9b9f0e28b8387ac", + "29f702ae855c39c29b67aff69beed0c3c5c8bdec9858088e08751452582ac357", + "b155d2383f69934a89a64b007ac32fb7cbbb5c36223411b22781651deefb2256", + "01391aa6ccfdbd4862c42a0e21a7e63f36e6c49f8e90f63c4d173a401dfda577", + "04652b84e58db3305c4e059c7ea1b078ae6764857ed3b4f19efcf12357c1320e", + "493452edc845fd670106632514dc35b8d24e58e47e4960b112e5cfa01f8a8473", + "78099da76a8cd3ba9e154b096dbd655bb85b64591ea80f64884d7d2387aaa571", + "9b10d73ad3381cda51936e6fed3d7c3eac8b79dbb7eea0731d01b82c0aaf0165", "bf0b5a827d950dd2959add0c45e983846db8b255bff42b889f7d85acfff87e43", - "c9bf9d1f68c7d8c13c5743a6ae556b79e983aa81089cc07cd3c952be1b4417c9", - "731ecf6cdfc1e454289623816f8e903bd27bb94da8b5a92398767e17be52e51c", - "7911dabef72f1a49eb416a71f6639d94fd5f1e1842d70be8628743ac7680d595", - "8dd1a4a982b0ae27d53ab09c5e4453271f869af64701abb60e714baffb7bf0ac", - "b6d7c13913e65b624f1d2df5db0bd1602098a6ee408d1cdcf0886c80f9a90bfd", - "1b8e575a242a69f80a0b2442480d8579aecdb07999527302199770114271bd17", - "9d4e21a10f6739794f55053b7b234bdb4c436b5851d4787ad655fb38f3b439bf", - "e5a7c54c80399358a8d2d57bf3f635092171567109568704209cc36e1f38df3b", - "d706b3a3df747159676fb9cf785fb8ea0acc287d56aa22b585a333ad00d24d98", - "61ca42f63205097a998c3621fd94a191f6946a471b5ab4afa602bc60f4cf8a43", - "a931c6237ff8568c6576c660d173db15709ecde417f1ba5e0a49703ca26a8409", - "4a4680f5b2a670250c8e3a17ca95aeac421842653676bb9ff35cba4a239fc6dd", - "a95272a0500ea92f26315971d84975318a20514e81bc1f4e6ef165dace91cac7", - "52a6adc52ea90a2638b94206b1c8c38a06e760d709b23019381617e4f71c893d", - "e6930e7a63ce3f8dfb3fc3ccc5736fe5e3376be7c12660f1d077f523201c4bf2", - "61ecdba01886166aa22e2a13858c1226a4c43bf31d9a41905df06fbb0a736235", - "a63e43bee92430a1b6df8a07958642940459158f1c44d539f303399c8eebadbb", - "31592fda95b494176dc595fc8451f00bbfa874e3169755718a442b11db80951b", - "77e2f5bc42c6bae9403b4d017aa17cefca1f09c4581120debede0e03808e64b3", - "07117d13ca628cb76702ef88f4cc29585eb5cfa8197e3f30f12bc750c6d07556", - "c03853f6d6847f853eabcaecb52f823533f05fe413d1b9b58d77700665b2c678", - "00bc6b9edcfc5d2dd38228fdb59f2783df348e34679fb30cca5eb71228037a05", - "1e67e17d8f08ca81624101cb2a0077793d5a8645187562c28cb59bd5047fbc5c", - "5fea956925bf5f49189ba0cae98584b7a94ee593495a2b6e8950468d8bc7d927", - "f0bb031be99dcfab3ea36275844e876d99b773061f401e77aaf74ddcc7f98305", - "62a526fb39696bd62c3157fcb85508bd11c6d6850ea188ad61089ec6f0036f76", - "511912cb54cda24bcbf6c29523dc5a7d9d4d116e6206342df72dac5ff4d0f1e7", - "36ae25d9e04f64d1af5e8f798e4a8d6dbca154c05b728acf51ae903ac897a7b8", - "22ec36a71c86a9b373a1953e5fd6af2dd3ce32a99dbcfbce719df5d3a9dc1001", - "1c9a29543cf2b9e4e6569f3e6f11b014b9b4cf84f367919aebd58eed81ce5d9f", - "ca4ac8e4b2b152ff8ba782ead6cae3b7cf43e2edbe36324bffef99e743d5b5a9", - "634eaaa05dd7347eb94981b0c278b3f417b01f78d416643660dfde7d73d18849", - "7cd0e4d7b66eb35c61b2031f4bf2adb4a9b2034452ef607e627faa74d5a5df7a", - "4bd849e2880a12764cdc1adadde14e5a8f1972baf17b7a677fc9e916695d6c95", - "f4aa103f81b50141e761d887b340dc2401c7625afc989478f5f722afa14341ce", - "1c2fa57bfd016ba8c48fa8c52168313bc3fa8352175abdd30a33f06d8ec17a09", - "d34435db094425d881b6bd4d4b028db54904ecec28ceb9c2532faa04e5a815a2", - "b06e9d237d18eb9169e0b55a3097c691c9e36092be38136d376cdd10aefa1066", - "596719b6d1dd61f6e972dffca4dee1ac07009d9e4545fc40756adcd1bb44e557", - "1c4f82be36da1aee2e218e904dd79ce1766afdfce9abbcb3f31b74b728a9e97a", - "f380169996f8818fa9a28ffcdd41fb3551bf9e3b24806e5288a355468459a6ec", - "f4629eee8d5a88f87438d097060891c9e54ea92bcd995b97fc4702342db53b56", - "25838d5859d42a2f58426ca05ff96bd65fa8f1828bb7739bac19d1bbebedc793", + "9bea0fd3bae240bbbe2ad7b138ae5b38d224671d3c2cdcde912c1ca48e8b80f6", + "2b67d59a8e18401b3f6eab95c8c97acbe6fc213e2bbabcc0c7b99597f9835b1d", + "e4bdbb4c30614fb39e42c45daa81c63deedec2a86dc9cce1f3610339fa6963d6", + "bd305d7da716c0da4db05133cff6140cb6a0db51b8e4c92ab910e7a560525a5d", + "ccfae57db663d89987988aeb5573ac4973b44b74779978c1945006fc532bc802", + "8d87dadcea92b8d1f576f6fa31a7fd7bdf588c4fe928d0ce1e458c8a4f83a56d", + "be12fe20909796cb03ac2e68de874ea994f281fafe1a2737aa817ac2498464e3", + "cf8cc2d666d2e56309ce715836d4e0f25ca04053ef974972addaaedbcecbbf3d", + "753c8b7a91a15578c7ca55d40ffe5211af55add6b8744e1e7443c6a6b78d73d6", + "110bfdf3edf15e2cfb0f763e431d0d3eb03e637ced6d6deb53ea32ab44871176", + "b824ca05bb50f1370c64b455407b9ada1495e03897325ddd82abe0b109323ea9", + "7bab251b5f780be78ebb4b1350f02d46cb19efa01546aa46f3a889de730691a7", + "072751fbea0812e36a3e532cc09dd3120c25ab8dc0170e70404ad5fc009e1693", "a2ae5587d4be646858903cd5d3f21dbcb393d45497c2c1a018f86d66d6319be9", - "cbc3aec249f6b1317bea0dbebadddc0882827350c4027727763b32bf179d7340", - "49efb83a3bb4e0b07f865136644f54c30c49f7d735e4e9bb57574d1eb903c0cf", - "64befff9b59f1e1d85cd0427abdc2929c4fa4d0f4f2b620c203491f763159c69", - "bc5b5ffad3f505c7b96d6ae39d65926d7f44e607832354f129f50161c2e21063", - "1f65adf9c1f871423da95014a5f9cfbec9ab1c542c1ad22155c5e5bf665b89c5", - "5891d3af8b90fa7de17f63cb39e5c79f34807355dc16fc821f6c4a5d7aabcaca", - "39bf96522584662672e97ced3dfdb88d5ebe0c530ab5805aa3c4250a847b0dd2", - "ffee1b435c5ce6b7e2639d88cdbe3b672e8c904ee67dae2d369d826d892e4730", - "0a5a58219ad088efd963f4e4a40ebed3d526f4f434dd4550b1d36c8085de2b14", - "fae7cb306d7918fbe864b81586b618584fc80901aeb444ac40e096deebf841d1", - "55959779d01712cac1bc630bfa658c79e9443c41ca9fb1a7752c96760007a5ca", - "c33b2f548b13f85607827fce36232aca61e5c65456434362e7c1e55a6858afa2", - "39fd416d7110d5dec887b7f8d501c1977a99361d9457c0b0d07224d79ed67df0", - "975e464727abce072b2c46639fe135523dff1d27ab6d9d1d61ce355cc2186506", - "783986ff162cfdce8eb36a7763e4a68c6231a6db19e5f4fd0a13287c921107db", - "e925967017d102582351b3c66dfa27e149b53e40c3352e71e9a18c5c9001a6c0", - "a2a0d6d8f4f1e5f4f1075e407b458448f0264e3f20308ae5079307ea16fc2032", - "5667ab68b9874896ac5c6c366b66809b271d8615d78f641a90464b0103f20c9e", - "0a7657de453d53eaf91ef4e13c58523a9503e7046bb626a69abf4937c57314cd", - "bfd3b12f98c21a905075c04586461e650250dab57147a389faae222babaa69a2", - "d607641b0a40dad0eef33a3bad27386f03d62bd7e7c9c9104863b68e34aa9798", + "9464e11fc216b8a24ea23f064714906076736582a6354eec1a7eabe096446c33", + "a00e192aaa8b79c505491003f2e450a74ed0c3a562cc58a9d0556244f1100d56", + "c9084bc1cab17631711f813124050a5c8d779c92b991251a5e24a942e7a5fb78", + "d1eedca8499de6f4d73710d6bccf76aa6db59d7731feb23ea385b73bf0648c06", + "2d825027a11f448a52d69fa12d21a37c0cc588ab7492ca1339f5eeb3db399c73", + "b75d226a6c348821da831e45ea9eb24d1be39259a5e4ff30fc39bdcdf71e7ab8", + "66e39bc3c9fccfee5735a0790386afdbfea1656f10ba45f184ce834494386312", + "e87bb690798e2b0527a53e530f6ed038d3a8a2d74ea3c5c4a1f0115ecd2eb699", + "4b2ebab52ba5ddc868bfde99d026be500ee011824477b173c0a52e43b9360f29", + "f0f85de2b4131bc16d690623325ffcbc64c0e0f3734e1b03a58db9d883abdd0d", + "b589eb1df8c19964bc27e8f0d85cb7daa1a2f3c31183388302e56c3c2450bb34", + "ec3222e07f9cfcbe2b644afffaffdb9368febc5300fe53d2cfd9ee11da60efd4", + "45ef756dbfb46ac113cac9fdc89076ef46587f96bf4a53882052392b4ac0f462", + "3e87c2aba10caac3443f27ab36ef88c386ef848b3e23018b99c8fa9b57d8231f", + "b7523cd87c0b9d22c1ab0390f5f1d6034a91c8f7101511296d7909af66e27040", + "e01e15482f14d5ce56282f95edae9af61f642ac0a37c0dcdef9826aeae1dce02", + "09909c5dc7e424c0835a82121c5b562001ef2eef330f3a2f46f4ff40a55c7226", + "d9781c3db7fe42a793fb9591dfd63334396126ca6d71c13c16a004e9e47b7954", "c14ada1928435915d8938cce0948824d088e8639a7bb7431121774f60f0f2304", - "70642f7aa2e3ddb2938cc2b1afe07e306840a45ff7e537c2466d3483bc8c898f", - "04bc1f212ad5c7afe6d9a358fc6493b6029e44c87bcbc0b1bdba8ec16e3ad1b8", - "feab5c54b74adeb7282ec7269ecbe69a75be8110ce0747b451c139c45eab96d6", - "7ee8632a9520ab2b04c303fb8b5210d4ecf51e5a48c7b71388c8200138a330b2", - "a5055090f5da1fab522eeb6959b4f5e34796455185a4294b61686f2b596faf0e", - "f920ee3b9be0a8908d02a4783cfc381fe8cd1a685216fac4ce70b6b60a51a47b", - "0e869b3ba6cddc9a0361e7eecd509f4fb1a0cd217c1e9242ae453d21c45b45e9", - "2e7bc4e078a1d993a9ffb389f8094715a152b4137d319942d7b21d84c17c99ab", - "53ecccc2528b95450bf6c9d614212f57fa5bcc79101c26b79b307b6b3ac19918", - "ceb224d4429165f4bf3cae3114a4b80060fc9be3cb5326a8aec3fb481c7ac8ca", - "da44a557e77641d6838ca632ea0df38b9f108eb72a558dde37938dec4e257571", - "dad85807906e40cbd2096846638aca272d975d111cca7da84216d4ea87471a46", - "4e256bc198df6c1cd23e9e44644e2a54b7a547aa1a3cc8e148c1eb55217412db", - "b46151359446f0c033b13a6bf646b6d7065cd5116cea1281abd618051d1d59a4", - "d13036eb72f7959b04a11f1363bf7f7f5bbf9aab41c51eba8f5d7a39f44ccd4a", - "24ce14852def4d1cf92edf6d78de0a90f8a1fd0f23a7105e67083c7d16673498", - "762b2fc5c4e101a264ecbe3a1d1ccbe9130c6778047638f5933dc6957d737d86", - "42ed1265a481bf35d2d063aa41fbabd7cab2e7acf33d3082aaa6d2d413a59a60", - "0c84af9aa1a8b353b6dd7821e45d588085898a878ebf6585dfbc5a5a136c59b6", - "72980342cf07d68e2400ecd4e245d87915fc658718871242b8cae25e7dcb34ac", - "a7c88ba9e0473a0bd7057e7b9c0d6c03663621fa2bda0411f8dcafa6f2661c34", - "5f1ac724a5a9e9ec22c49025a833e8d54cba75779c713df2c26b5d157935b236", - "23f428573944867f9f77fc8b11292fc56e455234368d34c80efddca0feb1317d", - "654b27fab2e11293e21465b7015eb208b9b1dbc42c0f5fd01a3607925c1cde61", - "81cb3a40b780b5c75e12971ceea05810c43e81899ace861175d5f4945c097c44", - "b7b22d28cb308c74b883b925518142a342b2c05f09118c90363504498063b071", - "03fd7eed6dbc56aa0c62ea3e45d999d0e67d526d8e05a36e430ab00335b8e0e1", - "d02e0e599461ad5d1190a8d83c9bfc9e0d9467cd4c87cd8ba5782bf2f23af91b", - "ab5517cee56b7ebcdaf469cbb9d8ddc9b83c2afed5d658ee9d1029c0d91289f4", - "7a5cf83833020f61bbae596a00510f5520fc676b52ddc85418840355d56ce733", - "b47ebeca524ffb7405651d70f9d58095d137219f3e85ead20ad48647067d6fad", - "19777917dfda3ea377a83c014d62a16a5f9d95236e0d89f599a5a175bc08ed34", - "b5491ede25d7c6d6daa0eb2b949bd3754b6e91c6ca6c3833943872070caafac0", + "acd13fbf5a3858ad28454b14c3d6f0d59e9f106c221ccdfc0f9e1fefcf0ec3bd", + "85f1f718d3b19a882d5d06899ff62d42918575f6490478af699b50acfbce12ed", + "9b139d300af8cc930e467f66f3f618db7c7609f2b36fcb1084479763b77f23f6", + "cf8714f877f13ad771aa098b8cddbed2aa0f5236d5e905d0d891d01c7ab49645", + "61e1829bfc1fa03eed619f2214551e91ff5fcde9ef64b7f4f3107f840f6ca073", + "04120cfb85018e8307454f8437d06028b4f7f15f4808fa40db53eee631f0bfe7", + "f6f032704222dc7c24d72dd0641e589a1ff656c8a58c0c11579bc629b74a3026", + "cba7ec048c8834dde1168a1b2619ccd2d12b1be4e7fa9b65630cdfe696142df5", + "b94d728aa4aa9617660385c53705121a13a4ab190a80b5cebf90dd7517a0ea4e", + "0200002c4c82fd2c66e23435a58b077fbbb778a14394b96149470b2788c78e90", + "9c8e5209062c1f68d997ede3d4bb1eb20a148c5e2f72f5886ac39058849ae6d7", + "b75c5dafab5954e912368ea2948da8706501624134565a6bbd8cf9a01f6b7920", + "34c394bbfa5f9fe046375c272d10c347576c29a9c637b935b11dc349513a1c6a", + "d30983c394100ccfc16516624bde55f78c40b6168266d8b27ffc98f34e4495b1", + "c75a7731c5647d3dca44bff388b843f232005022b6fd7b0546c872686551cbc7", + "bc056f1a76389b37d19bbdcf871245db1e534591c11c21c7d498328f46f2bba3", + "13495e5842fd64c43104f418f24ded110a84cee7e1960c8cd0d1aa42a8ee23cc", + "0df15c1bdd6fee342d706831997bbde6e80b6a2a559007eedfd86b8171955cc3", + "1ab1d2279d73f26d8c2c2d29d0a4d964f0c662ae49fe4888ea9f08ceb99946e9", + "96d7be8f8757a5870d9825b48aa2be9c33928bd4ce1a8cf77856ffd461b63fa6", "62f12b73084fd5a92c8a7cbd8ae65e1e2b7b82684db75cf6521ab0b27912dfbd", - "67517fe05e3a305f04c6ad9cda02cb15e0b1f17e04bd9b8deccbad9a02c3b837", - "6333fd5a77b72c623df665bf999f3a26315df8d30b24b153465d8d6612c923fb", - "b6e9c90a0f4249c5bf1eb17c64d1a90305aae8bf844d1f55d75336bf89240b3d", - "da9864d075980ad9c48e9db550c22c632edfdb389a75fa666d0cbc0851aaa2e7", - "1203aba5db22a8d19b6392a0d3ae0b627541d1566c7d17a2ccdd9016f092df9c", - "79d4e5c0163ac93a53fc65069febc7f68da89781547ced498cff730dd4b7fb41", - "18ad14fa170ee31924c97c0ded842125e4989c082f9c18158577f7346244a582", - "a0b2af082db1e7cc3dab489cb0b0d827608abe7e786f9bf2deaa07b8b37e2d5e", - "e339e5a9f9ac99c2c064da364254fd0f54c81348ec997a52084eaf322cc797a7", - "14059d1debbd74bd663bd2a7c53c8bf16a9546b45e53e019b0700e239b49e08c", - "efef3cc5cd8b5a58fe90620607c319dd854fe4f5c7770e9475d641949d031c58", - "34151dfc3028887b215cb3c605ebd8fe9b5c0e3a5958aa2a4f4d42af5c424c46", - "698c86654a4c00a29a19ca8c1a12a59e25a9eb4f1db8a3525c723d157ee4456a", - "f64ef9c33cf16e5050de83a9811a8b77fa9daa54f678083cb44f22867b20da0c", - "56fa3e42576e33eb85a27a7c016e3d843775236fb8e259d29b908528aa9e97e0", - "bd52a94d40f5beb8d4f2e44f8100e956f87ba8b8215412a7c10955bcd5ab41e1", - "1d83fd45d621911550858e7fec25aedcb5c7e90199a6911a31b5d4e93dce2ae8", - "cdbc9f553cd3137b38c3c108578bad989961f110d976bb20fa07b09e41550c19", - "f79c144d19ba0218b308e26b7e7db99639fe5b9bef2c50c46d7d9810815cfba7", - "d3338935cd57f8070292d0fc3e8c6cc3f4047c1581c4b6dd761e9a1cf96e9790", - "3940e88f43a2689c5741bbfa1aaa455ca7b51411644423a6f6b26014a365379f", - "5dd9cc50c9c73bfffa22ad8b469f66d5b0ed84d400e1b9e8e89ae5af255f0250", - "d0bbcaef3a232d96f6609f8ec93abf778c0579c9533f2f9be1c006d3787f0ab9", - "45c933f2dc291dc0ca148b87be32a49db74b3ef4c1073c22b238cca0c7d5fec3", - "948e765ca6c6f95f06892e168cec5e12bc0b7c07022622e4ccda0328db8ea640", - "7724523ae235fe3d3f12c6e4cfa5d1cd5d60d0a9d08772d7a5770433915ed915", - "369324b6b754bc2528792c507cc30b3c787acea33815c2af6c6d0002a1358600", - "1af371a8d7123a3979fa6d79a42204fa6403aca196dda141dd0abfd4b501b7b9", - "41ef383cda5132e7918425dc9a940ef6457c91b94ef74ad82b22fe33317f86d6", - "ebf7e09d56378b8c94f47cb57d93875c0a16389694cbbb67ea7ad7bae09db98d", - "d6a0303eb1a61c5a1d01e1b4c5c84ed9e823104c71cc14ce50c0a9f8387c10f6", - "8853adef982a77838cd2efc203f9d9228f0d5b6dbadfdb2ea5dc117116ef8048", - "fcc119d507f8363eb8cc2d6b5282713fe6714774e94f582ae13767521945fa5e", - "864377db1bb29fcc1d8e9b22e27060c3c25de36b8060dee98cb09eb213a97354", - "dd513918183e0d3e381e4251a01543cdbace4fa950395a85b230fcdbb00d548b", - "7bfee2adcbde8982b29c6a288eb42c827d96c84fa15b6b9a95245fd110412734", - "40b6367da9ab8c1ad45c3e6454fd880c738870b6d27c6b65b14f3ba5c301f4db", - "4377c67a005da24b3d9bfe55f2b5a424f4ae7308b3c0ed331a04148ba41e75e1", - "dc84d847ce70e2e11812c3bdf4ccfc485b397f7950431967c9ea92bdf5269d8a", - "14aec1f89dccb4e8396c28fd2d5d41ff52200ee07c2cb1961c2d48e9b81d08d4", - "390041157d7d6a05f455a71ac5621faa21f2edad7c98f282facd895b0b940906", - "69c04bc9d83b06d52a88996286fd47366668d2f3957adf6186d225911a97300a", - "7a1ed0d921f5902e69661932a925cd92ecd13e8ec313275a10ae2ed4575e70ab", - "0af6ab4ebfd934be54fc792d76516eb96a6be486c2de322ce6be0e65c595cff7", - "895cedd3661ff9039636dd29d02bc5442f5c723519f2e7bfdb8fcaf2af281f13", - "58415ac3dceeca6c566f948f483ebda679ca81be76da6873cbabf082f0023c09", - "8437094abacac6112c560b0ea3ba9248960c2b8937f6533c143d2b4427738ce7", - "5eb0f7409978198397f4122a6fdf5ed30180404f7e99c4e647c75b0f2f0f6ac7", - "67a1cab3f097758d7cae6728bc4c27087b85b8926fa58f7b383bfcb95da53e4d", - "8b88898315b58e1dc5de516f151fd440a12ad81504293012d44483b20437ae82", - "ddf13dfd6f8a7e153f5a3917a82fd108f08464f3a7cf9a0176c2ddae8bdeac02", - "2110c81571310eeaf9f2e6c80ea8c3bc02fb5b222f13181ea91741924e85a03e", - "fcaa08b16a084e5c79153c5f4c7d2e8873f96145bf84af6ddb362c666d878918", - "f77517ba3c5feb4122a22713fdd46c256860e8c85c7060b99f07e91b48b2763b", - "c864a835aab81ab605f2a8d2c32fa354ff9087d69d45ba3f580eaa1543525106", - "b0f25e7028e46288c3d2d5178ef224c494a2762db85181925d9991125964c1ec", - "24035e038e7a6a8ad12c9928880e7996d3488a311f83e69675566a4af8f8aeab", - "3777d6e91e574a0ef7461960cf4f3472f4ad08b4bc79c7670fc233b5a2c4d3e8", - "c9bb48270559b63289655d8cfcbd33c7308b436511bce2ddfdcf27e875d1498c", - "7148679b5716ab36d9ef7aa14691047e6216fe09cef333349e3986b1ba205611", - "2f4cdb495fb2b24b8e4eb1adeecadfb994abfd91d7b2771a940f1beae22f42d2", - "e607b3ba9258e6da0c69b4c244e70dc2f7ddbc8cb3159bbb7e0993b15eda5538", - "f8156c94a629763840cd98abacb3a27ca19734afb59ec8ccfe323d0f058ace37", - "4e1446791916274b787c16ae484ba49620e75aaba40d904f7015964957a30f24", - "e2c07f535a6783e44530aa13efe90dd5e596fa7210f6c2a913897a29adeab644", - "4e512e554081120a847c1c3aa7a90172aef3838137b3e5d504d6139904ccbc93", - "56e735984c497826a1feb361f7e75f5ab02d158ab836c8add1117b88d7f387c3", - "6ad1a245428130e1ed1caeea06afaf3f67aed275c589a101d399ed0e0363d7ea", - "9306ea1ba4089ad317a4004356af8ce788c285aaf838d8f280744abba711c173", - "58e04c83937631522951ad3f25d94ce83f4606d00fbfb166741da921e9e2797a", - "67fde708757d6bb5b6fee99c97b311234ed6286c80943a3477d77838bb157833", - "9c8e5209062c1f68d997ede3d4bb1eb20a148c5e2f72f5886ac39058849ae6d7", - "811708346a7e20529e5a7676de29340102b32f46ce989b81e3e510ed4e81ce10", - "082f9d4b3e056030f2f42bebf3c1e499dcc28ce8c7f04656e63ff8c9fa037811", - "d755d41fabd7f5445dbe96997ca5c33820280c6d1e474799463fbf9f276d8f9d", - "5c92d31aacc40ea6b0ce4a1f1f5c1c894322ff39538d5cf9592586ffe63b2622", - "a450e375da0f5bf91cbbd3bbd64b225fb0ab2024abc1706e753dfff0f75a0ec7", - "d72ff2ca3f139d0cd6e68dc80f2b0b6592dde8423ba951cb0caf32329eb0765a", - "44c1237a5857fa3fe39c7b6b1ea8b2a1665fde3260c854d703b0a5bc6dca7b46", - "07a74d0d6083061f4dba0a56a229edb8764bb6511b2ab501a0ee7cd3538315b8", - "45a907915f958be801a7993d3f667f80d9c8481eddfac0f4b0776233b39e9cd1", - "bd09d5cb757381daf4670ce7be30df3a9480ae9521609faefd85c1a2918b768a", - "20a342752428535522675ba57ccad285e29073dd2164e6ddc923dd51f138c62c", - "46b7695e16d7e75fbca2807707ed26275f33171f9d9dc166c4cdfc27197bae0b", - "e28f45be9f16e948aa127969d0bdfdb39ba9dc4124dbf2e3b4dc56a0b617350a", - "96445c61989ae21cbee6c4943b80f223c0dfe3e350a4daec55476549f97f0304", - "7d66345eeea44b9a8f5362a63d77d4a8a21366e28166e2349e79839e10c7bde9", - "3e1fc96d823f56ef9d3aad980980431109efc3fafc49276bcc193c478249f4d2", - "24fe7353bd96e363a3e94b09e067b7cb8f0141ba58acc84ee4132f0e3e8b7db5", - "8824cc38a3cbd18f8e0cf6d41500f134ccdc1550d3ba48857312acd9abc33c56", - "32e164cc7e83bda976b93bf063e2960c0726c684ad4d7aac84953d8d4c68e73c", - "5f45c667e6acd15b555b81a1a7df405ee1ec71ce5e35788e497191d17d07e7fd", - "324b418f179fd7969e13f4ca7e67255abcccd80bf3953a50055105a74a389d33", - "51ccb7a4d9e8a4e9c4129c6bc01bcbb944b6ee00489aed50d2efc533b8e2af85", - "73e779fffdee19cbf60b0d60c408116aeae8ae74282d77a4e28be23889f48cb7", - "67142be06ede0bea368bdc91e9ee55c8edee39b68a3a485dda0653b148e1b4a0", - "78099da76a8cd3ba9e154b096dbd655bb85b64591ea80f64884d7d2387aaa571", - "50709f5079b85ddf22abce0aa5a938f37930d44ec6f6db6cc05437e0353d5fb8", - "71e0cd324f86b739511442486b8f5a24b7c16b5878040a23f394c80d6db8aba9", - "678181cf803e41a8440a003478c5595dac86495234877d0960ea71d39558b881", - "db4bab04de0ad95abd4a71250b9922e34f2478b151c0f531899674add9ba274d", - "a61c4710c3aa9aa8a2141e4aacfd8aecefd61d306c927babb2db7e7626a08a08", - "a29863890d5d6abdac7db34a7a26e36d51a6f8201ef9800d3b199d54b14db69e", - "1948ec267cde4fe77ff11c73ad40fa9f72491857d9aee2e8ec16e27b8bb897f6", - "be479191ffad1354013d87fb918a8685dc9f55d5439ca5288745e4c5263b3566", - "a643ba14817db65acbd6402131aeb11778434c841c719b2e88eb0afa089bd3c6", - "b99c0e0e175913bc27d3fccb15010f4676f8b33d81e86e93e7f877d3a8a135dd", - "037e0ed281b3ba57402a95e94493dc598cf8a1cb401d9377f223848d96b8c5be", - "ecc1d0dfc017ea648aca65698b63fc39583acb7856cf8719919f09eb043e2826", - "3ae28db5361e59988f107164f11d88dd923801caa5cf1808fe7674ea257a00e6", - "59f45f347acd68b96a0efa69f320f5f5f421804797eec2d91710ba6e1af8758e", - "2dfefc93548e85a73d6ebe4b6a5dd66732c246f3750e95b33746c83536e4587e", - "28ef20af09a429e6e6b39a3cb8458ae6f48f89dd4b3e339cbddbc5aa176cce4c", - "2d6a249854c7aada198109caefa140ad45cc8cdfa91db7d60c89a9900869ef24", - "761ec8a202f265b387107d357377fa4e76db16f0db7c61b0faa48e15ca3b56f3", - "e5a1711b0ad0b813874e518e393efe3a7a4645b3199d2ae573577203aa51e685", - "3c5ce4f647055b0ed7e2cc7887ccf5fac2caf45a28f014d792f71304fce256b0", - "cf10405c7626e9cc8133ea4cb76c5719ee522470e2c6954f32c4e3d39efb0415", - "03d4e8119de94b6f171ce9a28652a97f20bc66de49908b6f31ae22e0b5dc3eb4", - "4ce5774949472162766df39f9d4e266e912dce551ec1c2c08ba948a663a7e707", - "9d531b8d19ad181fbc615f67e632507b69832dd09e25b01968be56d6cd6deeb6", - "42879c6a4b073a1737578d03db402f23d5b56ca9f5cece221fa54bc39087e22d", - "31d7d101b7b2f782aeceacafc745be660d8586efd82d8522221a39461aa262fc", - "0f1f06b4fe2a9219bdf3e8a2896ad05eab5635ddd87c2cde26431b4b048e50dd", - "cfe1812b981ad0958b49ed47e747fa506c117e10d245dff6942aa8e309d8f028", - "72fa2d18acc566b6c7ffb287a0a8fbb0a29af71c7fde5cacd33bd2028cf23d3a", - "91443c30f78a4d21a7ccf8b809264955d32a62a0b01f1145799418c754665628", - "a737a908f36830672600019a23ce01a758c324aab701c4ba10c118570a939320", - "4f83f1ee24faa4fff484fccfe98f7220813d8e3fa07bb602b144086f27428aa8", - "cde7761a3e48a795d0d8c52c63447e2fdd608b617cb65d2d9b2422c35f90c1c6", - "8f99ff6a214fbe73dab68aa6fc0512a7c97707ff4d186036bd95f0bd92f8c8ce", - "c5f73468613a3b1a4aaf58710488aa810cccbf6501cd7d8f922138df2d72fe78", - "8dd68177ab03811eec567cd7feeceb4094e97a4607ee5e9c7478d50b4223f9e6", - "2759c45069b977a975127a511fbb61c1bf021bf10528d797f09faf356a665a91", - "bc9894d9c54b831116d3473e264d4c4634df5b18fc0f2f3d5f4917d60197ea56", - "088d308908f7a568f64e849d32b520d9a76cd1e731fa2eea5ee29992c4fba0fd", - "0a4a76d4bebcfee1d9f08413efc186948031e3ec7d8c2707c16ad5377d4bf156", - "7e23284c0ef0d7cef34acf4801f848fbcad7964578ce112ce68b3872aa1c4dcf", - "ed0c8f26d13a24642ef8cb9496f9bf6d8e984556e918c238afa7e13f8a22d9ae", - "d1c655a2667c2b85c39a88097b98f8e1f2c2327297324373ade5d64535df387c", - "259cbe48b11253b7cf5aba89778339a97d3799e0b90e104fb5889af9a0970d1e", - "51a163d20ad2348ff0e999a874f14feece723279f1c9e70e5e43d0a352648a80", - "e8051e036c391c9b460a0ce86d651a6ebb8d65382ec42722363f934735f23e41", - "34477dcf2ee1613630def5924cc443f14d4feff5bb52657298e162cfdf23076f", - "4e3fb94d726c59f6f6e91ee837f9d87eac38eb13b4801cb54cabe3d5c5ce3650", - "8c1e1c5073bca3f08477abbf768a75643ee7bd005725d4f45d827bbc6f41c2f7", - "a761b23a609569b4fb28f0ffaa413b7ba202e6e3956837ec53a3b326c804d318", - "2b6c4b126d8d38e839d704c3cecdd71814acf09268e80e1e86b4389863898cb9", - "7ffa26936746443ed19cc59ff253aca50f5427fbe14dd72d5d5bfb32edb48843", - "2e984b96556887e21dfce3f3d7dfa090421e4200fc7b132eae5e08e34eb370ac", - "36e4c80f3d4485e7b309d5e2b777b5c9b06a2aca1cf3aa209ee714d6a2b6132b", - "3f7b329f6bbbb3a0e8acf29cbc760226324bff3b7ba04461d6b70abb8afc7859", - "e9517e120602f1b6ff636ff71bf14885482becc111955877e8685b8b88c2dc3d", - "e40e84236ad5fdc15dbbfaa437b5f468db928d5f9822b98a0a8998366da17656", - "4eb778e511547b152095f8bb72e52a3c3c197a67fedbc63b2697f0cc8b8dee77", - "4044cd00cfdc1a636864e0be25cf5025f52bd22e4731e09974e4a6835705f723", - "65f839a5d0520df781a22c6ada300604f046e60a5164b8455860ea8e449199b2", - "aa58e428c357f755357c75bcc2303d838dc99d34f726e3b9639f9a171e685caf", - "e24ca281dcc373c104020d128d0c5a00a79eee357581e14c8a5cbfd0b72221ab", - "9977b72122d25e21b4350f894f0dc2eae9a6faa598a863c095f6a27e44bb666f", - "a9b3b92cc9b17a9b95ee924b086ccca0e3c5407c8ee3cf26ba63088018109a82", - "5da3987f8b3052d7bb881c3c8454f7c228f50c73d4a8e182c32c479c2c3e1db3", - "2027a47c0b043d5f64571af35d9f1db125409d12549a971a58aa201b8f2b0488", - "ca79c6c62c6df3f3fb9a3f8dc35d6b3f0c27842696bee06674e277098d2bfee9", - "09a8b39db58f132e8f752064ab5a71be4f9fda2667786a268cdcc507ca84ff28", - "34c394bbfa5f9fe046375c272d10c347576c29a9c637b935b11dc349513a1c6a", - "2a61bec5faff58ce47dde7bee67df78b26a17e60166d85f5dcdc779ca8956e6d", - "96a57080416dd18ed122de00eb7cc53f35af876a673ad117a325f1a2bea87924", - "8da43691bb8d512ea6b3fa2d6daf492acff1c5c3a5d2cfbfcbacd04f97d16603", - "f2ea44f0231c7632c785abb5a2be0e1f6e1dfc73b6a44db0c5951bef407ebb99", - "17dab4b13f9bf61cdfc84b3f2096b39d13952a9669ca78d8e74c14794b4d280b", - "47762f470814b0b47fe5efcf364b91c9466b79f5dbddb6dc59758dc97f698534", - "7e4cec299b13140a36c584c68e6248cba694a7d11fed802cc84e86fb243c55c4", - "1023daa051b07903debbf56e09c043ac6ea8e77dc34583c1221b38e35f0260cb", - "23128742cd8f9e7ab1cbd85c4747869aff5de37f5b3d703ef348e622ee9a9a5c", - "b0b24ac45f0701762e4f93a62487f9dbf3193e00e146d1aac8ad620fdfe6dde7", - "383becbae9a0cf0011978643fcac2c21c7ed700b9ef436e901b90a1e475047b3", - "fa00cad6318fcab5deac2dfc429b053a357dc1a9fcc63720946e0c232470e80a", - "b692259c89ae47186d0b0f7dd1763fa3c7c7d175c7c476612ff439454d20be13", - "b94d728aa4aa9617660385c53705121a13a4ab190a80b5cebf90dd7517a0ea4e", - "7298df3e09cda63c352125499c50002a7a265f33f5ecbd20469544eccfb3c508", - "28e30b828f9f630833035c8a0e18e6bec763c63ea032d0dfe7af5827d8834d8f", - "2bc6f87770ffd5cf4751b4511f2385ff9f6fe69f20f395ea6eed9c527a27718c", - "64d8a8d25b623ad324bbbfe34ef88a9cbda105eecb6390f30a9a437c4afdcfbb", - "34775f725af76990ea4a07a65e1af7d2c1ce9b94fcc256bc7ee0a80b113b6193", - "56bccb3263bce9a0f2daf055b32b4bc4176676280076d9ddd5f0f9a19a41dd2c", - "93b78ffcc209f681d0cea4c4034064161a98cff0a0a57c31819f3eeb3588d41a", - "aa1e3735c56c0214f56bbd1fe9fcfada30e04cac2c39c7cf772acb109c4e63ff", - "5c301f93488523bb200d68fdd7341fe851321477afd7d67b7ed45e53051b4517", - "6aca0a7d2e459b267b713afe61c4302731e75470ce7f26cfa8ba6bd977254296", - "7e5b66b4245f0a766b531dee1e53348f38f40606bfe0d5b86d247c3233766a19", - "9331aea4b6c1ae678b8d54b9fd9154d883196777d4c679d70df6121ca0fecaa9", - "53a5661a2e9fa6f60b9e3d2299e6cf4fe0196d4a850be93fea8b007ea6142dae", - "d7ae16a40591c89535215adde775f20f39bb59e512c386b888d4596ea889b667", - "e515e25b32cd681e302172ac21ec68afd4112f290a51decd65a57be01bf877c1", - "62e04de62f85b1c39dd465c115015d506ccef38f7dcf2cd738946a7d614811e4", - "7a48d809257de9c24bf97450c5e0b5161a00c39ca6ec01374108df238e91ba9f", - "996a30c43824012e201f2012ef8afec826ad02e47289526c457f3e50aa54f836", - "c818bc6c8a6099b9f5edc5cb494e7e485c5cef225b533db1c3735fe53b4cbddd", - "c925910f91222e4b226b1ff45f7d53f0a651a7c5d813a083d5f89f24aa83607e", - "6f13516614e491042e8f16dc1f0cfc3619de73104a787e4e1f88aeefa83ff5be", - "82cd745a0c46efebade4fc550b740462b35e279d3ebd37f3da50482271081ed4", - "962bb0da7a2e86e1e2046f1a73b63321ea268d54e65a1b32479cd71519d4a63d", - "7c6309c9f90f24372b894b7d9a63b8f587f932663d564eb32664e093d22b93c5", - "32a1237cce674c7464e8e2e83cbf16bc787c97b0494cac4e81d01625b260979f", - "ba94de5f1601a50c5798528b6b2e6f82669792e2c3ea107e2ae924ca7575da69", - "f59b1a939eb835d0a79a5a19112fde369c2fd6f490b798a8abf15c0ad57473f7", - "32c614725a492ed5bde80dd94269fc090f14746346eb0b46c298119d46a7655c", - "4b0e0c2199b94cc21f8f3752126000d63840005d5ee2fa063ffd56ce4772a227", - "39ca4bbe5f53eadb47f0e1d7a4cdb5851a8cf21ca50f02ffbb7d01111178a3e3", - "1fba8731962da72c77e546241dacffa9b94fccd05c4135a231a192fb2f0f9ff3", - "a065dd412c8256d8ad73342ff610b5777169d93b9c61cba32fc8e2bb4b110346", - "50418f430137a1c0629719b855d2b35ad92fd0994a87bb5e97a95e7629f60fb5", - "eda7d0ec9359fd964fb368ae5cea8b39d73b1158e5a3c1c3e367f1173ad53d3f", - "3faf1db176e22935972d4818ed28f682a8237b8886a026cd86980d7f2d5dc139", - "38ace304486f4e038c48cb60fa9de7518282cc7ab02decac7c05d74bcce94dfa", - "b7b79fc322f099faa7e537889f7d2120f7283669fd23a80a1db8d878af302f1b", - "958be5d6b373811e6c4e202f9d06c1688fa745606d3e734b313a6661707f4c42", - "be1eb33a2cbb67e62876bbf9ddb253d6236f2b67a00de27d25153362a0de9874", - "d568b12fc8c0c741fc5f74fa30b8e9964b7ae3ff470b94fd7ec8aac25e197bee", - "daf6af6eef54b9064a2b7f33f05e3d2c0568d17858834f1c88c2b2018c1ff70c", - "a25a94e5910d167017f77af107c448403a8c9c95a0a2f691b26a2db4540bcbed", - "27359ad258787f522c5181a6c248afb363d2229cea1e57253e82623add6bdc41", - "ce31436539b5d2bc6539aab93ed9351ba0589630a5eb32fc08ce0be5a4da4011", - "594417fed321c615ebe0502de158273f420817617e8317ef542e976b04431394", - "30d323f6de86cc276241761f02abe85b2b21484446dd609067f2322033fa9975", - "1dc4d3e19ac54bf20a7bc4884a7fb2c5da766c6254a0d5ed3ceff6e972a5badc", - "61e1829bfc1fa03eed619f2214551e91ff5fcde9ef64b7f4f3107f840f6ca073", - "b1b1cf4b912be03767e072eaf4d9d7b0df14caa294113a1d71e89da1774b370a", - "3610634beb3556d9d99d4c94a3c0aa457a2e867520758d5a7474f7d45c817cf8", - "2a169049e5b39bb2b3306cc60f6cc3d1d0e5e3d6e02d1e071e875c2bade741da", - "e3e258442a4af8ab945c3445a0351a35a3797f35b437e1812ecec68e9a1c2ae3", - "8d44cbc983a0c97531ea7860f0d4c9f423555fafff9545a87219769d3b9e92f8", - "cc7a74d7667642e85ecbeb72e88204e8eb34c00c52a0770a6923b757975538f9", - "b98e5d24ed1b51b93f89f56bb1dfe3e96295b72b1b0299ffa22f2401643445bf", - "df01d242199d1b4538f83c2b68407c4149fe54be2b1573397b182faee1e3882e", - "ad3468177bd0359dea5db1f7d7eb2b9b4b12773dd6f9d6681d6221efc7fc41a3", - "cfb5edede2ebc049293d355666aad072be36227283b3984e01c6fc001a438431", - "a02469ca021d05e1c2618bcd244f50d6cfbff940e576567fd545af2943d041ba", - "6048967a366b38226f2f939ff7462758a94e577eb56287e09f9a50ee6426c1d0", - "25038ffb1fc32437c18ce2bd665526907371f65f70292238d87b7120d2078dab", - "79c855777bb71cf420492dda76b6dff0374ec0f1981bc01f3651d41ac3f2b683", - "2f12f01ec373e4e36b8efab54d0111cc6de2927976b8422967461fcc194fc73d", - "8ad6a5cf4fcffdd0367e9e03e994ecdd0e04dafd876aef142cdac89af0783b3e", - "0da1aa766f0a7a2a956c830901e1ecfa20df8ec56b1dfc66309d98189c859b74", - "1c23fb8a74470398e62da51bf8ff3f6c78ddc3441d3a50312d90a184a0273fb2", - "99834322335b1e8b9ca4785008012af3eb475a3ff26e26d675f10e2beb3a559f", - "177940af4c80edd9d240a3650e898553c909f08870b591db3a1ca3745b0c17a2", - "9e8a36bc289025a49380835c95155d57c5968d55543b36f380f6ab2f37ca5f81", - "87f155999e0078a0e864089c6cb549066fce4a8df57a8c909d3cff6738cce7e3", - "09a10f378f089d0fdfa275db2ccd4ac56aabd21d8c885ee979203232d32ed5bd", - "e4742681f7476b2acd745a905a2c55a2dd1cf8657dc09329b0605f852b9a5225", - "e6e8d83e4b4b6c4ee994b0fcc27769e763a41ff503b9bb23d7d5e10bb43aa6f4", - "7cf12979503214a4b3ba691497dfbb5ea24be71ae5138869d253b5ad69dafec3", - "3b8924d8017e97fe9554863057d272f963d9e660538bf02aa9f61bfd7f3f6610", - "2d6e07df0d1d88dc4d379b6085fbfbaa10f447ad771b5c09038bd14e32ff9503", - "0a38719d97b8168dcb539b9b6fad20a7163965272565cf429aad69e983fa7eaf", - "0595558321cd9caeb201d4b79975ca582af46fb076aa2b286a3ca38718cfc10f", - "16d4d0960b3dafae0a5b533a877f3f2c7328727801ba811154edf80c41eac4b9", - "8216d2326f3371c9362f92183ed8b6f05242dfac13fa532a75ecd7368b754ab9", - "e239fa0014bc3b770a9d56bfa65607ce2b5bf4bc9683021f2b77610c6bafa500", - "032d90eaa8e1c77657a8081fb8a1540c05580b8f4ba901f5505eb5ccd178980b", - "8296f3607cae7d947a4c0eac7e3cd862f3f5129e42d6847d0c1fbf5bbd1b4852", - "e58c4a2b4af9f4d44d689fd4fe0924b5d1df60a17831d2bf3d5f7ce5a69eb0bd", - "3ecd8fa1be87f215b0d7afa12c49742fb2e2ed907e2068caf7771853f8897c55", - "84ba931e3c3f890bcbb185866add6b77629d0d609a5801b2ed3f4a1d7c666229", - "603899e497cfef19404a0b42b256445b6e29b19570c2d4311fa13190cadcfb08", - "c24a80861e2e7ca1b82ee4aa24da7591678b6ecd63e81af56d69ca3d76bf9f83", - "d4ec529a7c44978b800c9cc0ceb4e3b69c35285c7a3812aaedf0ceafd9d79e7e", - "3eff7aaaaf4c1e790b8b58bdad6a08b8438db38021452c8f381138b4c80b5fda", - "5613651890ad767bc402b9df8b27a945e266b1d91be67db180c7324f433493eb", - "bc775a9200259561b73a03440cd29d7a4cef5ca0bb16b9b9c9426b6f89e92b6d", - "b6ab0871570e30642df6dd9b395b2b21369980de75894c9b3521b4778efd271e", - "fde6601cbfc434b5a8c47a64932d3a77b68f95973b87fbbb0328d4caa3b8228a", - "5f6c33e6ef7da331066a2fb896999c37265fe04ff6db14a06456d3c3815285b8", - "53b4f612bea7098cc6c17532693238c2c4021744e3529fddecf79f57aa041016", - "080d930064e212f23ac2ae07908a0654f0ccae6d5e76a9f26147fb3350035630", - "4edacbf396765724988dc633f72d6c61e7e2788bc86a6f3059ee2be4c1a172a4", - "bec34421966fe97582849d79076230ad9360d2a3a17af21c84a0d3092071f56d", - "19226632d93e5ca22276cde73e7d90fad4b7b9e59016112fcec1eb60173bfab0", - "5d3340804060d5777275fefecca181c42715284692f31699b62b467177ae8e0c", - "f24966797f36a0f507622e3f44c93325248867ef2d840e255be000a4a79ba24c", - "8093ab1ce6242b88f05712a25f53859be88b59b1b2beba3de7f4800d1e09264b", - "87d6252634e2388d67587f1388773e34f8e5a97e37402ea21cfd69af845d0d23", - "e8d1b780e65c084b8fb9d1fa09d5dd021ff2a4b413eb07c96d2e56b58e5d479d", - "8fcc424356bf9a4a6d8f7f9eec68281e3011d374950b55d4313b182ec06a8989", - "ab235dca27588789e90f329d603a28fa6b49d45d137305bed67af5fab361310d", - "9f447ccad53a3ed9b804d7e2b25a3c6f6d02b15f27555297694ec09bb5036642", - "6fe818a99644f329aa5ae1eba2cff8d617fd6ccd09636159f4efabd23b057b34", - "f61a6c054e21d9cee07814dc52cf398c4fb3ec7c552457bf27c2c6f817efe85e", - "a5919eaa229f07e1291f5dcaf730041f495550ac3d0bb3496e20193c98510c6d", - "09d4d9dfe5ebd5c4db03ff623ecc64e5820a6466a7998c8d5154a0d746aa686e", - "619c5638626d03d942dd365147de2706e73739ff0423e62533f922b6eb7e6490", - "24a2426762783ad73c87c319afecc9f291c1c6bc77e4a102ac6c7072711a7b45", - "f34f2c2b3d156492de9b1293dbc348faf9c1fc9219749659aaf3ed5678143b23", - "2f2dc960b9cab563ce897940e865f6b59c47fbcd71126bd2c544461cf1d1b927", - "d6f320624bd8579764c872d53ed3642d24745b19a431d697187c051bd930f214", - "d1035604804c0df6b91ef5757a70b264741b099c5bf1612c3a708a2930ac03cd", - "655ae8c318b7bcf254fec7b1e3b20319f11bfd08a675c3256bbfc95bce6668a4", - "d522044841b3770ac78c2f302fc12d95d761a05648df920239b37f4b6e74f3f4", - "8f610b8755ea8d53034bf51b1c2371746309c446f3e54aea686ed8d37516207f", - "c83c06c21a16bce522c414dc29117810f97d47e8bd174d008cb9097e6bfabceb", - "fe029c2c7da088f9d3b88043852b1a6e1fbd8c963d656866ff41fd726624e425", - "2dca1b29e5bbcb2f30701b82d176716e480d389959d9000bc90f723e6ff23e61", - "7bfc73b3012f2e4c1e0ba3146a5f9afc2a1ad9bb026715de2f00f943bce18e43", - "055c4e95dd3bc889e5a3e017d7f691f4ef8ea59da1aacf9b0fd6e134d896c692", - "849c13a2a2cb95094c43c6eee8f0fc8b4331e47759c96ea7b2696da86f046864", - "efb5b66591f7f870e952556676304181ecd886fadceaf02141cb63c443020798", - "3225deab4f62eca68dcfdc7619e9bed699e9ddb8d2969b1a6b63fc59f88056da", - "4d34cd964c2ae0c2e012390310382300e090d5a8bf973740cfeba6233f21057a", - "5edfc98f3733d44f84418d876ee4cb1b2ddd6719010820515727159de814ca2a", - "ec79f47c81ed21ca8fa435d3d78b573fb12d66c64822b68afd80d4dfae98de9c", - "7bfe28cf4668049a203d09b5e161c7e6164fb6b29521d6215906edcd6ba0d3c1", - "e315a948719a0ba26859e7e07de36456f0d19b0a82c45bf949f2c96bc26c265a", - "6a130f11dc1ed98420041a1dfa0fd3609ba6d5ea843a97c45c1dc9789ec4af94", - "e1132af43e595ae67091c0a2453fba89a219b6a5f251e3d897d4fb00f0a54656", - "c0ba489968d7252dee9a17d3063e47456cab78414e6f4d38d005d7ca776a4691", - "bd8ddbd7cfae217aefbd164ecbc77fff2f7151991fad38ec3a82ce03040d520a", - "e446a2f75258af3215ee61025e91568679f14772eb5dbcc7d9e773909b2332db", - "87185fe6d445f8f1e319f8f323ce022be6da50f6d2cc5609f9f1ad89886917e4", - "43a5187181c7079c1d6ce66b1f501a84fed5503b29a8097c5de14cd51f607c62", - "2e44adec46a035275000443c8b5823ae4fbeaea9dac5809ae7d1177e6593b72d", - "14a0e2dd045c94c9bb5064d139b62d8171a56214a65cdbda10dac3e5c24169ce", - "9113edabb02573c3a490ac7da86d1107cd6861d824385faf7c28204ac439b125", - "5ac217a450e93007e84842e9da57ff57c28e42a992c4227ba0c8e34dab328e20", - "cd99319364f5cb9a540983d0173ba20ef546ca4470c93c9b06f302d4f1d0348f", - "1f2dda64ef3b71173b60859c0309889db6e22ab0a3ad6207b8982c483b973f54", - "433c4862939e4d9dfc0492957e2ac466ff26b5d711a75eecbce89d84dfe67ce0", - "8961ae39128ae1a196646a0dca9a14307e1f453cba5ce6968435b3506435c4af", - "bc056f1a76389b37d19bbdcf871245db1e534591c11c21c7d498328f46f2bba3", - "7a7d213c82041c35d0ac663d83bc40aaae442a5974fadab8501b67ab54a48e3b", - "8587ae58cd80c9a272afed536784e6d23110824c910153b356e93e04262a558e", - "3c99bf55e3a14096739cdfaf84c606ea16840158a2fcc0bc47bcde005fffe8d0", - "5516c98432a7389adffcb5b1fd3b087630f160e293707b6226ae22c593470b56", - "19c26eba1c187f1e6cfbb911ebf2a8fadc6ee70a4fcd3f88e4705d8280bc8396", - "b837b6fac7961b76e751100afaf8a73797797b9ac168c6373125ec959271c9cb", - "b76aac81c0d37d62e1e7ed03a0537d857b8a417040484ab20dacdaa9518d6dc2", - "77cd45cd40f3649a0f8f72a7dfa8f93b2334dfe3f7a22f2a0c3235e182596401", - "46b3ad431877f63ce4f012c661f610b85af47bd20bc666a1d0fa7db6ca0c5bde", - "3b8093c2f8f2e5ee1e567486dd0a98655fb606f814b6f26a3091c3ff4984d069", - "debe49d8e682fa253ed94ac89aa38ba32dcee9808d53a9121100e1af312a48f7", - "2b360282b9dc662300a9f3a4da09ef2fd921b8c1451f393dd59ebc95a244114f", - "daadca21a382adfef7d9d9f994fd93a2dbf8b19e17bcba9aaf26bd06dcd9b70f", - "8e75adaa8b7daf136ed75212dc8ec5dd1d63110fdc3b7c852c67af847ffe5e41", - "c7dc055a94400b0a3658b63e9b5e8a861517d8e7f983fcd54f158234cea4fdd0", - "3cbebd9a44f719cf2ccf6e5650d7645573d3a84cb71e5b2bb7844e0e56627afb", - "0914cf3513684c056dfe38c810c1b977d2274b693819585e379b8101593abc5b", - "64c0645338bbf8d8bbf3b1b3ef8f865d04b8c27723e8a976f5b2a50f73fa7b13", - "22e936b614c8d34e35694de98af0a3baad02a0317238671be78489861878297b", - "3a5056340ecdc0f0ef65e385fc465fb9f0f8493b7c010202011b27f3f44f2632", - "b75d226a6c348821da831e45ea9eb24d1be39259a5e4ff30fc39bdcdf71e7ab8", - "8dbccc5d31240afad0fc93b29d26750830e490e5de42fb4bb56dfbdbea0a8384", - "2673b7a69ee3306ba3514056926bd0ac428acf1a558aa817ee0c30fae0594766", - "03cea2556ad0058a660b06d9cfc191757f3c2d752fa96e7fb3d99f6d2b973092", - "eea379683c70f889141187881ffa7174782bf4974824c43a3fa15d78aac8c0bb", - "3e7b57887303f21b1815a03dc8a0983f5440744454fb5c4be77b442237ed6abb", - "21b4ee1cdd299b64951df77a4456622b81dcd879b515d597a8b57fa177e3828c", - "b1964cd68a531b032ecdb8275dc8356969e771d0c01ccf3d6518942b75b9ccf2", - "74dc064b4d8e0d092ce832662e7cc0baf3e861d00b82f0cd951df40f5d296ca4", - "4ef8e8676f51cbcffb10f1cdd6774a6f14f737ae61f32680c2983d912370e019", - "454c658aaee51baa13d3ef441466709b55b9807119fa0a74ad6353b2a893f485", - "ce708238ec9c7754afe0abbf2c9517d5a76ecf8fb4330591883f78deea27d818", - "eda4ed4cec5f178b33ec92045b82c82026187d8cb4dedcf534c26c9e8b10eaa3", - "0da77b5fc72a1390cbff36d44ceaf1fe8bee0e23e36631df50405a05d9f5545e", - "ee80ee56f7e81e957a3991b4bd52813fd7ff92a55611ba97793889e4c0d1932c", - "33c4028e9d4aba6c3cc6bddf98722984a72a0bc939aecb51cc59c8ec68600ba9", - "0957334f4c57f161c0dff558d05cd3bd172d5179f23e63ec8054415c390a16c9", - "51e658f27af66cf59b12dc9258be90286655575ed3fcb5a844943511947398cc", - "ce0486410cfb76699b53727fab459cd6f53a9d1a666aaa55d479cc15cc2c2326", - "c88cf1f2504aba7d464a36215a977a5c334329cfab4dc6ca040f1de46b2b3f58", - "be07edcf381a02e96a788c07a0082ffaef17efcde16b4f4e74214b1da21cbe09", - "db0997d11b15ff7bc43e58ab6ba7dddc54fc92cd9d409f6d804b1a57c0769a02", - "b833ec6afb29002f0c148b2e979400e653bc767662b33cc6251275c80b5a3261", - "442b417fd1a4ab20a2d153914b2aba297ebb4a13e896f9ee171e346c68d92f3a", - "1544fba9a0221c66ebcb1208d7ab2bb23a269b365b3b09b080c0aece4fabf280", - "f87e304596519dd26cf10c919c0f0e9665b9758d63296c7f28d373ea5e7f2a0f", - "d7a84cc931f1fd93f03878533c8ecbc861bf53a9c12e673829a7d806e3e85679", - "a24812caff45e02718c9bd97793f5ec0e0c79951a3acfba39af5285ce9b6aa5c", - "20f1a5d684922bb06b43713d5f774814d3c15bab06b76f1bce77a2a2164ce4b1", - "f0aac579629094dc3521a4022271373ce4724115420b7f57a0870b9ccadc4fa3", - "5fe122aa5c7deac202a9e8dce11e8b669d8ad041f267a72a34c303489e193596", - "c7d38f702d68f1f1629a648f4ce3151d4d871cc966ed00a749d83a974a6bab39", - "630eb8946b983e7b612b7a44bb4acc9cac0c5107545d75bf0a93577be2f03166", - "476d211ab315d7bc9b9ed4272a4d899c1856866105c7c8cf3c4d4aeb43da6491", - "433ed6154a9f5fda63c29437b11c394a6d7bdca9fde4805f552d764a4f87dc07", - "fbf754f098f38f8330abccbb52d08dca633b5563dd2d3cfc92d2e035cb24042e", - "ed78a3ab6e4659ce869a2c1b25a5f6402e4ab37aad3ee12a745fea82e9ca8d8d", - "f43027efb18a5c003ced16623bd096f351c6e4be5219cfeb27e5a1fba58b7093", - "d13b8b4c841597bf456f8519c9dd9cc19fcff9fcc3b7050025608d4e0cacb19e", - "613e49ecefe7bf7a6b3db5031a563e9880004658093ffcdac78bdc0bef4b6b71", - "1f46b209b97cf056eb59e28141ee89d27bdb2d9d2184ebd7f1a4dd62f03702a3", - "36d5d6b512a48cb6861ee89ed6ca17322e31b0e7b660e484081525f287728c41", - "f4625440c0f6e654425113e4033b2a215d2d613194ade44add9feda73b3c04b3", - "f647c7260457770201c73cf9f08928c9606264940d9940ef2d93652dd4329478", - "5dd61289065df45692b1c945d07fc51172616f6fca1316db259e0387fb417198", - "188827c44ae53568c16e2a581b0b1223666042b7b4cca5ffaaab04c50f762c0e", - "defefb71baca3e7bb4650c1b948ea8e436b4938d57aea82d021704fadb47954e", - "42cae7e85919e12e830420a27cdf8e43c30666b8953498f31de449e7f6f49cf9", - "f083127cc1a17649818f3f8f4e8d926a0adb78d895f027d61647f6296b29ec53", - "ec7f727c426561755b75e1566fb40627d3970eb433b56b05a2fc47e0486c2d83", - "5b3e809f5146142fb1a7341165dd72a87f5e4ec0b3f807addb51c62a15423924", - "a1e4f9daa5f124269240f48f55619c03317aa64472fad391ae6ca3f6df46d8a0", - "8a41589d32e869d30f462cc28e9f508345a100f04d29734c0bbaf0b4558853cf", - "6b03e96995b645d7501d4bba8796ae09cd9e4f806637fb767e11124af149769e", - "eb3de0f368ebbb76e3b9ae99b831e4afe947a21f1a3fcad21fe49e7c58c35532", - "1b81839f989d1ad1c1abf4d9a1ce05df14c6c53ec4734b1dc45b6722b1d591b4", - "ff56db96baf76fa5e4258c24198fbde4212ed8cd3fde183daa3d3f43a7107948", - "804427ed272ebad6aa06769100122cdd83bf8307b9132fe33aa0838f2e4b857a", - "dcfc8c0ec469929a9f18cb860f5c276ae49b51e7cda6f2c7882e118b896017fc", - "98ad7e24193de32c9b75cc8ab0d51730fb4324b01e295e33338166351dacb16e", - "71607d518dc03cdeecc015c72932bf1d23fbed8b76d641cdd67e08e45730b24d", - "1cf30a5d16447141b4256ddcdec2d38501baca0f560e76314faeff314183a338", - "d2a78b0e18a728d846ec108563d52e6497617a4da86da84a9d114f84a3a4e0a6", - "7c72206dc7d17cc99ad35d11119ece90ccb60852ee6f5193ea85d5d42114794a", - "87507067d7d571be5d8a78ff3be612427df987962f027de60be8cea9b44852ed", - "4967084d8e28e78f52aebd1033c3e351052d1060cb7f267676f6134ea0c5cb63", - "5f55db44957979693dd5e3ae3bca79bee76d9c7d9f88bc61162087d3dcf34ff2", - "7e4c743a9282dc57cdc6358a082ca6cb1ba19ab6e049ab8c4ec9ca8b146ce09a", - "ddeb93e5a1a3192e838702667a3d0e7f6c5cf6ea3f19f474378098f4ae5735e3", - "c74f0e4dc031536000fd072fef6cd6643401ad4f6d618ff6c4a7600f2a94967f", - "fc8a74259aa7c317c42141f18dd50bfe09675f09309c90595d56e186ed1ba8e7", - "6799104a090d2a71965105696c4b455ba75e3f138d796157b8236049fce252f8", - "39c22e3d79c3b6f3bb6740ff59d9429875d6dfcd2213cb5411c381495227777a", - "4d6e362d91e22d18cb5c5ab2c5b6f7f3068fb7efbf17c45cce38892bac597bec", - "1495f544b3a91e2746e306bc06be5094a49a898d67835c59c8ba6d1ac6058c5c", - "5fbd8e7eb0b18d1eec920142d1c8258f74b35f7d5b59388bfc5d7fee133f45f9", - "32448b27a414496f9790d1b5e2ec62b2c1935bd730a19bb73d41a9ebf4641bc7", - "046ff529546644352a9ff73c77271b23938ff18e0a0705a05643b9826127b3ad", - "888a136b76913622cec8578907ecb2b5681f290781468682f4194f216f9534c1", - "d5e7867914543068456e84c6149597f22dcba3350660ecd44d273945cd73b259", - "dcad9463e92e68fbc97f841048376eb8616fcb7d0a38bcdc8a75fc0b780dbca8", - "4a825a26b9fcddb6316a2cc8377378f5e54e23de6afc4d11f69a62b558944114", - "903605a3cde2c6cdec81af71d06d684171b98b0dc1dd65bc1d5cb76961af4267", - "d8067a2d643dbf09439d2e9e213e6ab6b5f243d8102b7f39fa7845bcc39d6fad", - "06ed1c91d03c5caf3931dfd9abe5a0abf03ba28e0d863ab9b1bd6ee34e9e6f7c", - "d2e6fe4b0b9c806cdb5679c210c307d750ef49c352cd23b2573f84c452024e86", - "0ee730a4c614e415a783eb03f783e4ec9f7d2d33dcd310173144ae2fd106ce93", - "d4327184759d0c2f6e7cb6c8c18661d2908fd63d24790e956ecddc2efc1d6d43", - "30212a64ecad629c5cd2ac67f55d8f834cbd7c0656c43d5705ae98be4e5ac07a", - "0892ee129e02a4fd1a32c7fe5b9bb87cb961708830d58e451486c6e939023367", - "1ff928c103c3af1c1d0ce613e9d4f10f4b34b9330abf84b2f1c0a6cb1873e766", - "337dad054e31d487e2549dab6ee7673b75be14d481190eb22017fb02a529292a", - "1b64312ec849ab2d646860a52a83bc7b05f05d25a814dedbffdc875e5b5067e9", - "61863a9961a0a008d8c9ed713eb8cc96a5e4bd090608b00d6924d746c90c3f28", - "2ed0d26d2507cb94cedca1ecc27ada33ee08acaea9f1ffc59b60864916c07e0d", - "232cf6e971d05ed8ffcc87e2a4f2b8392468b156fd22f75a5f4ccdddae97c51c", - "0701e951cc16057b03b098274f97ceee6c308437a380a6458b1caee3fe86d255", - "ebb3b69f94d8d992f5f6973324b4d9690018d112a21b9f93c810aa77e91c14f7", - "b1a84abf768d28aaea594d9a74e42916c12932d71bf705788bf15ca47f0fd93b", - "02a0ae3625a77b4ec7b8400be9aebd56d4182ca7d6988d450d9bf7ad1c03a4e3", - "0d568af745d9490c9ab5937097b6546eb54595197a198166dbf61b0df35554e1", - "f95c2dd57c3bc3106097a9992c39bc384cfdbaf123b8abfa3d124be6a21a26a9", - "e54e7002e271a46adb933c4cc1e42b4e47aa9957d45dc3be5e83a3c54c7dfe76", - "fc404c72cc2f03d2b4524393b369f867e947a899d9885f67428593ffd9b044cc", - "82ccacaa2072fda1615fe48192c1edabb68783299bf232130e13c2dbe828ba1c", - "af4d1ef5466927d71d93cf8ddc0938117c92f1e835f33792cf4b2f7b8f84e55c", - "fe5b10cb7f9e0c5d292fb2eae8c960e63c41e669c863d01979b52dfc00829dcf", - "bbd089dfdd0f8685ac11317abd20b75e21c7fbd6517790ab325b84e146f00e0f", - "29f702ae855c39c29b67aff69beed0c3c5c8bdec9858088e08751452582ac357", - "cfeb56a010d1e86441034d42505b1156ad38aaf2d440001b2e248bbb8c406960", - "787af078f05ef25108ae779a0be0c05e69f8c05536aa9e0efd0eecdac8c67e45", - "a067d424516b2b2214987648b480094cd6d91896f8cbb655684bfed59a84a488", - "a460bdf143bc86aba013a568c677e9fa14585652d5aa2ad7e28ec50baa083ebf", - "dd71c1a3b76a66de4ade8456ba27b567920314fe19dbced3f4fa6b9e2298a9a9", - "cb292043dca344f7faed6c5f773a0c18c4b49eeeafc18afb32b1ff1040cd9328", - "0d348e334184e62a36ff4e1dca807b194915b5cac48b492d0ead69a2f42d68cb", - "197fda1e43c2756a088923e60eecdffaff248745855c7a0a41d640ded62a312a", - "7da8ec7654339933a14c1f368ec73be1e736eacd7ac0dde43f45a16ccd2b1712", - "491560e40bca9284eac2a115c7bd0254525c9d98583bbb5948cea74f2d0a392d", - "bc94e241c0c0425a43b423fd923bed2b5a2c8c4a98d8a6a6ee08bb54ed4a4eb8", - "311c5091dab1cbde86a593f4e30b7172d814ff38785488c3e611b8ed35a657d4", - "254bdc0c051a40f6433ecd43f0cf9d01a5a1d61f0f79fff5dd91f0e35f329a42", - "2bb962fe6113c7710edb7bd18ae21236c293a9a8c58d35acef61fab99923774f", - "ecc472ba194eb6377a332d642ee021e7242cd0b471675aa341438bbeae08bb37", - "1739d7646412fa36612cb5d8559ba3134910194def287b612aa658a10563c480", - "544f0c8a2c6b5ff8bcb9cf521726115d6dcb59c843e00394c75fff6192a2564b", - "af2876c6c618aa981b7aed761c5fdb6aefee9edb5ccbfb1b80d6f964804d2595", - "5e7be3b1c0c4cb48b3d2f35303494fbb22d882dae21a2419b2c525aeae5932a6", - "d7608f4fed5dce08f5a124b4731587e73279baafe5da8217569f244d3589d233", - "e17ef9a7212c5fa9dd8cd42b169d6fa99399de4cb7e871a19f8d7b0af8f9d25b", - "cc4bd0540882b4270b3b8917e020fd79e2f61706728be31af5fa2154aa061160", - "0b0ee094a49df49caa0628f8ded03e92cde0ed6479f1cc2ba47db5814e1b66c3", - "992ed538d39eaeee3939d584ba7f0e8bc8b12b3033ec7fe5710d668eaba27508", - "1a44e4e6fa6a2136406344e1642296f19435f79f0115218188740e928de38040", - "bc0fd8ffc008eb190ccc402e4f385a6a3d6693c1102d9bfd3b0d9fff43733cbe", - "394e5e62ffc9485b9a03c062f118553cbe987d73753a0733574c6e256443a66b", - "0b2a4b4379f022b973c7afdd5c25453daa3a89d89d5b58bdadd80a3bb3f364ec", - "24c5e2455f75c13cd97ede49ef88a8adbb0f96845b9a4d884f3c1a33dcdb2a2a", - "276e8f7b5a81d2ce52ff0f92de4c25345e934b5bd2eaf676f38a8d688f4420eb", - "3b057fbbaaf319ad4a03b3c443db293d2ccdf3d7dce331d9bb89189dffc93044", - "b644706c96ac25d44975832822dbebcb2009c17af933eb2957eff48225675ea7", - "d5091e30c4c5918e802adcaac8109ec415eefe19292a65680f309a79b7df25cc", - "a9c92cd819151652d8dff0ae55878f2c785922f21945894a2b5ad89d2a503862", + "f12e397189f140b2bf2aa87700291939c7d3ace6d23241970793f2979b05a8e0", + "7cf12979503214a4b3ba691497dfbb5ea24be71ae5138869d253b5ad69dafec3", + "8961ae39128ae1a196646a0dca9a14307e1f453cba5ce6968435b3506435c4af", + "285533dc53705b18add207fa292c33d6394992be68777c076ff49e279aa04e96", + "9106b155e53a4917fcfa7c2cc6f7b457d2260bdf633840372306c2c5c7de4a16", + "3e5a4ec994cdffa569df546c8b401f87999c2076b77a0cf22e86e1a6975adf52", + "f1424d6cd12e02f3c48a74a2e21b08b7d85f8b7089e5bdb2a431949f93b0004e", + "78df08c773d9df76a333b05b0a68c62611b903bee73ea67db762901453ac29a9", + "92f3f233cf92490b67faaede66f9c79f49e1028706fa4219cc29295f3fc14098", + "25c00f7a8e3165259df18c523d2a51e39bfba6c973f81deca2cd71254dd21485", + "8ea07ceae4da2020dd0a5ef4f85b37217a5f7d130ce1d9bf82a27149c4ec47bb", + "ce64adc45eb67208d1a421e62859b6cd735e1feecd0d30fbd1ded7ae58d1a18c", + "037f739c101b1ccbb99edf283dde8336a1ab50926f9043bff5dd4cc2f1c4cf30", + "d91bc4c3f65ff54756c1deacba13c89813911cb0d5d703253cb44c25d4d629c0", + "1c2fa57bfd016ba8c48fa8c52168313bc3fa8352175abdd30a33f06d8ec17a09", + "90292970fa5d193e009c27e3258a1e89bbbe91c744507843798c65da147fce92", + "cdbb9025a55f231df41f57f389c77e18f5e05db4cc71569e91969d28babefcde", + "911b3d34fc12b2dc33d6aba0d91e145e3df4234acc8c4c030948cc988fa26302", + "3765770ce777d224115660b875d62cf4343960a6f06d4a60001fd6ccd16c385a", + "daba4df0331421616bbd335859554290941b437e5434d4afe3f21ad9757fb093", + "71777c70f635001fbee6a3e620ce7597a027b3e5e314bf433e66a7e247d3df17", + "558e7ec6166ba3de5ad4c052e2fb3dd2871e8b494c8810aa4224dd86dcb795a9", + "eb71ddc62dbcfdda8a7ba80b762a258e6381729fa04b42fca2a0aa499f8dc955", + "953f68833d337194d2856b345ab97cb49be652b8a6310f0035882f1ddc920244", + "45c933f2dc291dc0ca148b87be32a49db74b3ef4c1073c22b238cca0c7d5fec3", + "93fc0ae77b64af788cb87a196412cc6ae1847e4bb4336e2d34bfc7426ae118ba", + "eda7d0ec9359fd964fb368ae5cea8b39d73b1158e5a3c1c3e367f1173ad53d3f", + "090b31052c10e7d81ece797794e34ae952a7f483c5d1015f8cceaaf18472fe8e", + "b466c512f05a4cafcef5163ae4967f0ff43c89109c631084b36e296be8d2a321", + "5052350677b4d13daabadfc0794eb0af5ea5f677289cad5cdfe75e0a035d6893", + "52cf9d4b619fa67118ccf1bf73914a246a368a82c6eaf4e355ab7dd13687b77c", "5fbea90eb54a91a05009a8563ed442b7957fac8cd69cd52f6296368200d3cdc9", - "2e323b51f60924de3d4e9a269535b274d7c3913ad43ac6c281735e29378fb380", - "e87bb690798e2b0527a53e530f6ed038d3a8a2d74ea3c5c4a1f0115ecd2eb699", - "a09489f82eecbc72817209cd2b33748163d645a49a91227bcadad26b1817ca14", - "52fae125a52d3ae21aa2b663f28946b52004e86a00aca2f44a567de1bc9db5d4", - "6b9c3a77dfe9f64f0996bd8c279881cc6f78090394d8406896fda77492f8e3ee", - "4b11fd79e527bb16827c2cf0d28343ff6d857dc1821e15d41540e599d7971600", - "e5fd0cd1a0d3b066e3b9d9b43609f271c39dd44e33225a37454fb5226608d7a9", - "947dd30e53ed74f7247970746360495e7ddd8ff03fcd5e2654b53963816c8680", - "269c413dd3e17f29e6c008ff85775d98ab15f8ad7f2445f068ed9d028ddafc2b", - "a2b23772ab5da9d1f7f43510816383b2cda1c9031c98708b5ce4a8e9ce952ebd", - "192d54f1535058d04ae39a455994d04e31145d6daf61444841694c98694e8591", - "320e22a5079bf644d6835d6986f7b1ae805b1e05015d877d43a83a68e6834944", - "d248eb941ce44abdd4f32480851bb6ee3fdcb698bd21677f10bf6e609a9e764a", - "2e87c5c7b82723045f41ff885029afd0c4acd5bc66e2be2587b80086b7ddcfa5", - "4be93702a6a0d2b8a1a5904dec4b34e5d689a2a2ee0fef5323813e5daa63546a", - "0bf371b173846476b2b1cb6402bc2eee57f600c789a2a521259d8dee589c02cd", - "f48cad285fa3891f85ecd223a9a89316b1ffbb52b68ef61a80bd816572e51466", - "1f29c5da553aab54a6c55937f7aca8fe052cf998f907acba63ee7dd5c597be34", - "cbcc3dd763e2e9c022950d5fb41888f3b83a3d9354ff533ab33a82a068e2a746", - "03f933e9581d15ac3449fe7574bf79598737d7f7863e8570213027ad5ae5a92a", - "e7fa76e7747a041a702d7ba25d7bbd6d1a83ba5c35879c90c793193f223f36e7", - "64e582b0e76b17b63eb31d0b452f7ab4da003d7a33b5741308518424e4e8bbd0", - "43c96f25677f8b74dd80c9d2b654ca1066ab0a1863961687470239c99931efe0", - "c65e576a2593258682d5cb50517571fd250b256b7b80f9be4682f79e0287c3dd", - "09f3b1262a75b7aaa4f77a0047ac5160446cd0e2820a061859753afd9d875717", - "5d8d421462f8aaae31787f49bfde9d56116cae7cc1271731fff8385d040fea73", - "b43f314823102a795b59dd64c6114a6d99982d4eebebc8e3d8bd1e995c1f17ea", - "5fe3762a7d73fd2894465206c01e33205bcbe437aac7d6d15d3b1cd3c80e039e", - "8c6e3c684d4ba9d93fc08d2262182438d348a318e77a262a01d0ecad14e2faff", - "2e4a7d5047ee2e964d32cfebfc588f035e060d9bb4daa31316b9e6691c22f9a1", - "2e239ce039a3402e489d04af25f4f4ae47b5ce706905462ea750cb58dfbda3b7", - "55161f81eac7648d0a7df9e24df9cb9c28eb3bfb9de0db317f1e8c00e40fde08", - "ccb931cfc8c5167f98bfe5d180893cae3f0b0d5f1281ab3083e0874f15730fff", + "0d8aa48f7250abf3e832edc40723a6dcd003e763031dc355b9792a1323f3153a", + "e6773493b9f104a339aa88ab390aaf5c799dfadf43a8133acd6b64dea72ff942", + "edfc0900d70f8f4821b28097b76990c9842c745ec4add474705947db6022f8b1", + "d835844a3afba263ea1ca33370a97e79df9cc294a66b16e13fc8d01547a50f85", + "96e1a8260adfe9cb466468bef0d3031cd0eb16ea7b362d9e7c0fb358df7e2fec", + "5ac217a450e93007e84842e9da57ff57c28e42a992c4227ba0c8e34dab328e20", + "5c411fd02b3c2739094525b9ae48c59dbd8d8609650624ca8bab49309bd5a9f4", + "e4d8dd529995d36c900de792f9a4c6386ab18eb257e701ced18ea02b92e7feb7", + "b31169d49159ffa1e07e824c92e753cd080881cca1dde1da4b1b01e30114daa9", + "81e031fbd395a70fe565bc8805fb90f075c761a84225ea42ae966561f78f2a09", + "d14f7bff6296b9ec708914ad3904416edd516f68b9de839e633442303f7b0d1c", + "5e7be3b1c0c4cb48b3d2f35303494fbb22d882dae21a2419b2c525aeae5932a6", + "28d7cac6320be948a5dbeca575b46c5348bcf7d3e8b72edab05509b4d234c099", + "1d347a00dc5177be97b3c60149bfb6f29b8ad57da3d5af1d9bd917d9e39beec7", + "8dfbbe18ac92b0d714408527a8d21fae0130a2e12dd83c6e2a4f868102f896b4", + "a1c1fd50ae03023b1b91165a0273cf7fd6ba9b48698c927cb8566f0b0e15a920", + "bc94e241c0c0425a43b423fd923bed2b5a2c8c4a98d8a6a6ee08bb54ed4a4eb8", + "3c406be14b5f0fb3ee090a9a9c86ec98ba42f66ce1dad636ced0608fca68b323", + "9d3a3facafd3453ba4c35f42831357792b248a9137a43a79237d2d2cfbee1df6", + "53c72e7f899b09d847ca0aa1320e4de49268cd04d8149797748320ed88cd0d50", + "5f0c81f1aa723baa303b5c8c6580d2cbd8dedd1ffc6b5c9bc857e4cc7faaa1f9", + "e9ef5d04b1adc2cba1f345d80444d44d076c3130d74047da5ef884fefd7d40f8", + "0143be881ffec936a22957c75d1b38317ca3bfaf4ed7a7623992b6dbb68e5170", + "14127471f6e2d62f053da35f034b8f16b0a926c96ff118fdeaeec691501e7ad9", + "09e776bc0dbc98191f01ed6eaccec0a75fb46ba3658c57b664d0f8a9a3b158b8", + "09a8b39db58f132e8f752064ab5a71be4f9fda2667786a268cdcc507ca84ff28", + "8592b8cbe096d4f6fc9d08c438639973738a32da6f14194fad37c7b7328d952e", + "8c61320a2963d24be9f8bac774a4096bbc720d645e2d7a8fa9983ff03bc23ec2", + "47e468d213a5da7ebc682a9f7949446e728f02d4ef8ef9b9323120f6a080cf28", + "842218334f64ead4a5eb115d1fb29ad96f0de37957477e0e1c0b69d498d350cc", + "14059d1debbd74bd663bd2a7c53c8bf16a9546b45e53e019b0700e239b49e08c", + "8093ab1ce6242b88f05712a25f53859be88b59b1b2beba3de7f4800d1e09264b", + "843376038a735f7fd70dfc246e94ae7b4f8f6da9d9c98342c15d2397a4613aba", + "e6661fbd1f2cd35fea90c9632daa37f75835dd0f0520bbb87deae3c479a3049f", + "a4f052898aaae4f1094daf70e8a430f4b86bbb78a49b4b86fc702c00aa670ec0", + "a0b1530dc7a0fcce4bd7b2ced3454260dceb6f4338e2b5141ff86bb24fc0a1f3", + "5fe122aa5c7deac202a9e8dce11e8b669d8ad041f267a72a34c303489e193596", + "dd85ff6165f4fb41f7f752b680c87f28032f68d5f79e09de39208674e52392c1", + "10c384bb0e26cdb410773380694fdd1128bcdc2b275655086b3afaa79483de7b", + "b7da49a791a1a1871499ed364c84e0a8747305e4fba6b36a5df56a5aeb46232c", + "d568b12fc8c0c741fc5f74fa30b8e9964b7ae3ff470b94fd7ec8aac25e197bee", + "4da0a18f30f35cb7fbc8324a5e02fe5bd229665e434fb16a26f59ef15d18f23b", + "154c86e1d1a6ca39a08994ba4f927a18c34cbd88390a251893468970991fc788", + "099bf1165d5f4c909f3538a11dd77927771f1c5bf95152deeca05e01a9f5e672", + "c7d38f702d68f1f1629a648f4ce3151d4d871cc966ed00a749d83a974a6bab39", + "c9bd4048f833b45d9e2d02c54541d0e5672a1193bcbe52a5ff16b047e12efe04", + "41ef383cda5132e7918425dc9a940ef6457c91b94ef74ad82b22fe33317f86d6", + "1bae704a7481e6a8742fce18f16df78dadb1377c7162836b8d29eb213d2a9a52", + "f0bb031be99dcfab3ea36275844e876d99b773061f401e77aaf74ddcc7f98305", + "b853e08669ba02e5cd6f00b4aff92bf4906c07e676ff22d3f75bcf5deeda44a6", + "9abeaeec791a06a597a73f23756b6e2062fde7c4d1ec54445f7e4f25abc8824b", + "a5cfeab1c751c17a545cb3c42c5ccc325ff497b684ed7c9a982188758b95cba8", + "aa6825e75f7e99d5610f3038d86760078dcc37337d4b12ef104e8353944da8f7", + "82ccacaa2072fda1615fe48192c1edabb68783299bf232130e13c2dbe828ba1c", + "37b62dabaee02e962a034100918ebd2a97a0007cb5f7f94630848141d959eaa0", + "d7ae16a40591c89535215adde775f20f39bb59e512c386b888d4596ea889b667", + "4a793e37e45e9a77854762e08a8e8b8ab4d5ebfa5b5aca8a7b39aa88b8432f5b", + "585e374dd4118fd7b02a54b043655eb26dfd3b04481ebf6396402b47fe90b87c", + "f6c64acf511defc6a29d9adf9427745cbc3e3ecd2bde82d705081fd1d375fd9a", + "a2dae733af442a9e9cde155fcda19a624972d290f95aa7d75f42dd6539ec5ac7", + "f8e4f7b84b26a4566967d6d12b75f73362f7e15b0e15706d1e0295e7bae2cb3f", + "4ca6bf1f1feb4ef3840680bf3925b63daa200d3919d60f091e89ff0cf6620772", + "9cfcd21e0447f153a93f38809741275c85d6b261e7a99f00b772a10c1213d67a", + "177d0b90c3c9f91f0c1ce183bc331198b7fc74799a1a47ff071b78d6242f38dc", + "cb5bc1062c17c9a2433cc6656d37d756e354690b3c0d1921d506321b4af0c3ba", + "b8cda795804b0e1efb249c3247c032fe135d881159a202a5a4e23e91da1ad607", + "9fc741c374d68d4fcbe77bb71d5de6cbbd96104dd2f35342ed0b8a0dc3b6f7e8", + "5adee8ee13ac134fa27bead5824f005393ae54caaff28b7696272ef89e28e581", + "cefd3a19221d97be04d9195ff327f5fe964723ea183db2c75a1ec9acd80525aa", + "c297d8f1c8337bfc40ac1c55e258cbb93cc5885671277801e17ff80839dddaad", + "24355cd05bc96611fb524e0df11a86035c0c0dba1c0f9685636759070f4538c4", + "ac1ab1b9cc08f6cc27f1c8a202f92effa9e330cb15848684d54af697d9a39a74", + "2867b050b4f3f89d0b9f75abc39bdc5ea6e3083fe82184b9c37ad5492cc71ce9", + "57329127183155b6a8237ec48dce5c03f1fbe46474a01e5b71cb75ad467f9e93", + "a29863890d5d6abdac7db34a7a26e36d51a6f8201ef9800d3b199d54b14db69e", + "8774827fe8f2f9787f330cc540ec463744e4b473dca742650024a2def27d4c3d", + "70fc4fa0931d3e8e531954665788051012f8fb3fcae8634136c2f9d0b39162dd", + "340a54e456ce8b6d195d05145dd0260c737d4c24009828e768719418b03b0d08", + "73570ecc3b135cb879d56d811eaa9f2bc4727de8340ff8e71eab771445ce34ce", + "31bcd13472c505c703d7a923d28574e5014cb3560bac3a48edae0766042952fd", "7667496a85da4f54671325ae869817142d9431ff7cbbc170572d717ace0b0324", - "bc582e292ae61ba2faa31ac0cb726caafa23c5e110ef2d25199197184cf29e49", - "995da27708904b84ee67d3554528c04c821bc4ee6837a4e438d29d03ec84f2f7", - "9f61f89454c88e32dec5212bf9ad909a7ae10edbbe3388341195beccd3a01ef9", - "a5debf34d05aee6504163271562d5b855ee0f509cf8878726e7ebc87b0172036", - "9cb517d90e222bb973b1d7cd23b5462bc934064c33763faa713f8a506ee66536", - "c591eb7c6530c6b96fc3ce10421310b9255165ca245a4cf954021e0dbfc110f9", - "59ca71eac048736b6270f5cc907a6b95640eb880c18ed9da48a448053e3e7314", - "ae94ae374e2126d2873fcf558cf2ff26645762eca3f15649cfa5a1fa53afd98f", - "73acd232e2700eaeaacc5eeea64fa00e170436ea8581698b1fa03999817adf1c", - "0634a4da7757bd3262a762a1d492c986ce9458d0ecfe5ea2758baebad4a7ad45", - "893ddaddfe23f0f126e2daeddefb8d92adf221d34ed272acaa1677170279207c", - "195184fd3f4e70d799242255e87d6fa25f51a2b32d8175757e617e33aaf89944", - "fdd3a3c62b38f6e0c617946706cb5726827eaf6c9d95afe2b70bad0f730a1b91", - "66759e61754d96d07d89e05f7b5deb18884e8e4bd8fb02f5803a855683b71bfe", - "56db578a2a4f1dc279d9d57e07fdab8ae790cd4463ac9bc1411507ace8e7450e", - "7f971de9506f14731026520dd286eed1feb08640d2a07e1ec372567d7cf5957c", - "48ea12d5f1ee0be6e44df2c6fe7b186f5d7ef3237c5aa21364e66f2ad40c3c70", - "38038d28d6414f90d1d51c574b47c1571c6f9119ed645852f455cc88a7f585e4", - "8b8d36e46bcdc2b694646ba3b2d66fb8c24c7ff3179256acb141790b37f107f1", - "ee082b9a4a5bb337118c39e4c1f7a747e1067f347e997a7cf4a42d8d60eed0b9", - "b6e440eb99c6a4f5f43ae780d813ca4ecb4fb5cac0146319034191e1f37b36bc", - "690cedd12c61d4dce66079bf779b7fcf4e334861b88eb55c1f5c24e38fa65b2a", - "42aa3ab02dfb5ceb0819252a405b41f53c8e8b64fc9eee87bdad13cf7c16ce58", - "3e748c871be69d4fd2fc6126610cb8f3ac91698fedddc225d6e04fe25ef5c81d", - "2a17f6a337fb9a6adf6fa99b4694284bdb67478110b2df86920e77980a489129", - "819a86d606476241c6ade26d2c1a99a1c4666c6207752fd46c6de99675f88f0b", - "e3effe13cad52bedff1a7b827bff3a0aabbaca719f53779c90ecef23df446a08", - "b155d2383f69934a89a64b007ac32fb7cbbb5c36223411b22781651deefb2256", - "4a3440a95434bc6c776d0f1e9ac0eb1aa83f916e556325bc731b9a44fc62dc7d", - "947d594617273e679458cccc554d0fc644f4f0e27818000d4a647877e210d764", - "ef53950285ce45af4de0bef5452ea1fb271102aab977a413acaa2117e15d570d", - "9bea0fd3bae240bbbe2ad7b138ae5b38d224671d3c2cdcde912c1ca48e8b80f6", - "66705e69637139cea90f83cc83cd1adf31b3ba5732e66340b5d637f86ec94b43", - "7a69a5a8578ac03134eb4371f85ecb13e6109d1dea024075df29b7fd0c6c44ae", - "fe92919026a76f460f8727991c6b7d215c4f00557d5e18c853513dcc62c9ec57", - "3c19079d6598a115b2891a5984c6140cfba5d8bb46c69c6e663db3117b6b41ed", - "7a711a00c3bfa9f8aef7df1b498ea93e425b902cce497c3859af647ec78a9ce1", - "3d5b86dbadee4c09b217edc6c4d2633770d49e378916b96d24b6aa7f70479109", - "c75a7731c5647d3dca44bff388b843f232005022b6fd7b0546c872686551cbc7", - "6c927a204bbd07db8486416799d9fcd4547b537c4f887007a76b29514e071129", - "d6cd2bc022cc04d4db1d440ba5a3a4ce90abe69783e18bf7a2ed20b0e60e9d38", - "f29c772d6d0d4d2b1c64a5209ecfd289fe9af8509a460a84f3c2553f5f4c942e", - "ef506d334367785f9d9fa9119d81206d2b8f9579a4ad2d821497d9a61ac07bd6", - "a2afd267f5ec45153806c2f80c6097b2d09ee4f9dc16f617264832729201479d", - "9aa37b1db395401f895b655753e66284b35a898721c7b356aa2de3685d9c5fbf", - "a392d90630e8e3241a1f877e82b4157844a8156b8d484e0086c5989abd36ab79", - "12c0ba74381b4cfe1501b58adfb009044e7ae62884a1cedfd0d20e3bac122e2e", - "e7536be50a3bc3b1550f27619d3f4d6f396910729755bb24318c3197f2a9f2ce", + "57a3349eeeab3dac7026ddfbaa8c2cf0b5c6760d6c51ab3634020a4c37643108", + "edf04f9e4fe592e3b4116cebbc637abe98297782a1da139443217c6de9670f63", + "baf5419638cd797784047f7e1b852fbf37f6e6535b34441378f2a3dc6e4df247", "7b769fb1e889b3d4e4d6c165bb070f96545d81dc41397941606fa08dad32c113", - "ac60db48653203440d229890cc62e3842a6bcd09676161f841c352e30055ea1a", - "e8182f6531e650072095fed068a0da58b36446f9cc9c9f4a126037b7a4ae519b", - "e5b358fa4823e644b12a994925bcfe366f08e35e4a3426fbbd877f3063bf405b", - "61e945423408cd60c6e9625d56022c4b8281afa32477fac90219185743ad01ab", + "7d7af99c881b40d3204c7c23693962eeac25a1408782f6c3372b94b15552b764", + "cb981955255db352ba4cfbd672d066b33110a16d6a0973c30b1ebe8fe845ac28", + "a0699e76ef27c3226752b29ab0e40756877ae0529dfcdad81adb4963f7108ab3", + "ac918c4415019c14baaba4568c6d8af5230582c7c9b5d9043903712e1dc3b611", "584b7f1d580336c785f430485ef53d4b732c0fc27ccc76ac4d1021495d8dd41c", - "5c4e481352ca86b49924eba14cc1a4ea3b61bd2f4c7ed0ca7a9de07573bc25ed", - "5f0c81f1aa723baa303b5c8c6580d2cbd8dedd1ffc6b5c9bc857e4cc7faaa1f9", - "add0088fb425c80af5b9196b7c3d9e3ee80288e640a27193fb31cedce085edea", - "a2b6902890b012b3b29ba29e3361325f88efd8d908f8b77167515f38b583bb7e", + "8186653366c08b06aa1bfe72777ed7976fe47ce0096efa24b17507ef4bb0bacf", + "14b6b16fd0932e63b24ef0a1f5c9607123451e17fb672d93fda1c4d737f91f84", + "b6dd93b959440833e900996a52d74f5d3121c9cd3a1c6c5d18a6f1e0cd4e3cd8", + "8d41cfd5eb21a178bb4484d2efffe7f9e512f40e70759078c024ce35d3e6a9e0", + "992ed538d39eaeee3939d584ba7f0e8bc8b12b3033ec7fe5710d668eaba27508", + "bed0c141eeba88b7b5533563deec380a6ecb093ddca590019429aea3c96ed05c", + "03c80b354d30fad2c1f0b0c866ddc22e3e3854e659876b31b3c5b4243437a786", + "5516c98432a7389adffcb5b1fd3b087630f160e293707b6226ae22c593470b56", + "3f75739f7413f110a5a575e00d1290b3fb45c1387de0866e411f196e275f67e1", + "7eb04c4a243759c8fc195a33b15c01c015c375268e2410462ff412e835c991bd", + "f03425ac0078e98a99050fbb9cc66472056980d87dbade7524ed8eaa50ded4ce", + "442dc6b61727f20b65a217446728821dbb37d6c70ebe91a3c8148d7d2c1ae7db", "2f7ac351eb23c73efa41785bb7d1ece986f74198a92e71ae99f81d292f083448", - "5f171936519a31898c0b39af8a2d59a8608c1fc19ad840ea7d1400cf57705250", - "b9a11cc0f706fa907e42e113cdd944b0f4dfc09ea8a6a084aad7a15ca8b1daf7", - "8ea1aaf7a82ccc13ef2ddd08c69967a845d36dd66cf3dac1536bc77d47e910ff", - "979240a9568dd20af4b0b62f737a5c1069e01092e8589f150127387b6b766f9e", - "40c3a92c134512ecc49f9209b14762a21f34f4eeda745e528ff5a98c9089dc4d", - "6db68b3719a1bb2a4e1b3ba7f774ca9f94fb7b40b5a8e31fdc38bbe27c44f381", - "ca28fdd42c96c91190ddc7019cf0382e8004535aa984c348d07be81e716c6a5b", - "e6a0eb169df990f91e695797a7d11e10a09415d2feaed02169d192c9d3d33000", - "c40aa61773640b62ae7e007c54fd2e7518a6fb07fac79a48c659549e97ccf164", - "f35813b067f5859e38027977eb0451ee57cbc2ae157f82c0214e3b5443b7c8ca", - "2eab2090651df33c6c816d5cc6b1ec07b87f124e74cc78cb161db459d2d60df1", - "4725f7df694af17733dd18345b1493380904f9ddfc11dbcd0eed079b50cedf1f", - "3f4bfc805e6ceda302cd58a02059daeb8ec094f5b542c0877634c397dbf387d3", - "24bddf5068a2c60ce480ca5baa93870136a1c2fea4c23fffab947834497a5371", - "4244eb9d12e828b51dbcf50cadac2cde5221fad5fd86f77a1506e02b0b1a71ed", - "cf2016b3b214fde71cdd4ccb372070a6c55a91e3d012edf5cde5e805f5b1f0b9", - "ae809b52127d274a6c061f2fb8a7ef684456705e57360148b17f2b84d12ab38d", - "99912737b5f67f7251a13cdd970137f49d0341e2e9a2a4a89238c7e9b1f5a323", - "6f5619de3a61dd62a77e04ae2506b5cdb883a1a214c9834163c8df19a2bd7202", - "8df9cde9a97129015545e8d02beb3dab876c99b0d94b8ac7a24f7c176435b200", - "ebba4f8c20215f999459499db1ab87f800ac19edb6501e797146dce7f89e18ca", - "c03599aac0bd448e042cb4db8114753936d49ae79217877ed84f189710c5d69d", - "db29f286e12550ee682af6a9660e58a0aed160a12927dc1994ba43eb05c87e33", - "be3acb3349499300150b38aad3345a31e57fb656455711d4b7ab3979ea8e7740", - "1f1fdcdbd34ea16b3423e682d3bcc345e5f4c2e47a5ad4b1cacd9a0392d7e95f", - "ef1b12aa1f14b9a34bfd83ecca713cf4918e8c43fe97dbd5bbfc329be5d60331", - "2573125cd429cf558be813f24633352fff1a93da905501321bc2a2cd951d71df", - "fd14536acc9de50f2adbffba76bb0b28f14183b78f7a65432a147a40afcc187a", - "f633a849403e05a09d450a1c4454cef816dbfcd99303cee920c6aed8e9f92a24", - "9204a1a2684e7c68843fdf7ae1c60c8be6a84c60658badc4056b273dcee9f25c", - "ae197ac0a8b9dc2e1ff61767833b737ef8ede20e5549e0dc326d22e928650617", - "13495e5842fd64c43104f418f24ded110a84cee7e1960c8cd0d1aa42a8ee23cc", - "f0d05efe08a21c1001a9faaabe4764363d0c481f718693ba7e2aa80d0ae1fd16", - "328f36b5f14d0a808e5c0d989fdaeb2adcc5a664cb295b6813184aa1fb0cf041", - "adf877bb3f627379c6aaf8717f2baf3bc4d419685a9ad335a34f8dec53867979", - "8d62267b76ddc017aab7e2eb32f1d86e34596c8437d59048f344a4bf883c7873", - "edd7c79797d883a9da5317b5a7453d1142f128975c852ecdc8c66758ed57181d", - "23556216252d494a550ae0b7d34519219848f5b0989e46a6d8498d7c1ac9b4e8", - "f7c46353a077138c0f0a252c9c83a7bd776ea1b2dd3fee8f02cc3033ee3556ab", + "00d2921f3ccb30c11aac24b1c5271f48e01c5018404ad0b62232dded0f961749", + "9ab581880cffa1066abbcd47ac0d8de8bccb5ed6fc8f076717de34b20f20a3d8", + "2bf9495f5c5cd95a876e54bc0862b854b73a0fe2d5f2b1efc46b3792ae72c140", + "6f5815b08d89d62be2754630d5388459b4053459997c3d44c544367c292da491", + "ed161d485b46ed8efe1b3efd3b84b166510fb4cd4d8ba43ff8f2eb4ec58a2593", + "bb1682b729d25d7221dd13e1be6bd653a642b04186319edaf746f85fd85132fe", + "f0a45668562e33479aeff819b6963d5029b60f258a1fc77d4c7b78c29e3a64f1", + "7dd52c2adef2ade5e3edf002cadcf68430dae7798b77a31a724ae4b6f95fd2f4", + "343ba95eb76fa9432273fb52aee8c70b89297b64aeafe9b4a47ccdc575fd2bd4", + "cf2016b3b214fde71cdd4ccb372070a6c55a91e3d012edf5cde5e805f5b1f0b9", + "995da27708904b84ee67d3554528c04c821bc4ee6837a4e438d29d03ec84f2f7", + "471a7aa80a88c3ab714e41e51a59275a236f2e8cb96f71743307db6e77174a0c", + "51ccb7a4d9e8a4e9c4129c6bc01bcbb944b6ee00489aed50d2efc533b8e2af85", + "1a44e4e6fa6a2136406344e1642296f19435f79f0115218188740e928de38040", "81245d015309a49ec72ac1853b26e67068a0e41d5dc2235857de98b86eac36ca", - "8d87dadcea92b8d1f576f6fa31a7fd7bdf588c4fe928d0ce1e458c8a4f83a56d", - "f04ac517845b6199744c649e0d98137bec053d2de26afe2495a0ebd2e41af521", - "fcf8dbc0c4b216087d1b788230ae5383ee6c2bbd23a9a53ae273b6424ae3cc14", - "308a50e336c70a594233aeb9b2a33e4af00e6b0f0237e28518f9cef095707d68", - "532cc76c50e353b9fafdc96f930b53457b5455c51703399b3d589a66b667d14c", - "66525a545fc7a350d6be70293ffb2926c97ffb054547c3a21b0325df1186eebb", - "768c1014df9a205198d9363f028cb5b9b9e494c53dadf02c56d1f9dfa80678b3", - "a589e87ff165ba50f1feb140e8b57fa95979abd55f8cdcfadc31d6cf48aab4cc", - "544410736c4027621f051a5ec1a61c8b70422b9e9bd17371f65423b7174c20a6", - "036ee345f2e4b61b8fef1b5b64fe6a84c03fe4ebe1639004c29f2cb86731812d", - "e06860afab4651e6f7122e41abc639676e8b10c9b4c2590afcc2f437dcf1ee37", - "9f1b2bad3f4b50d21ac79d8d25602bc3439ac0fe20256bcbbe2e68ae0c0179cf", - "ab8854ddfaaffc972531640564c400c46b49441c58c9580cd0db54d59a83bffb", - "8653a6ea6538c714be35c4e0d5fb48fdd61e8c87623221e0cc54c0cb2d965a65", - "9f2654b711e1f909bd67a39b7944f1e4cae7b9ab0d10f985eb54af57983a719a", - "c434a192f776ff53d854b1655e96f621596418d0f4ed5264fb495c32740b44a3", - "15024c2ca9ada8b956d781e6a6c0efa654d7559c69bb3c9694537bd43f595c37", - "1f12d1f8eb38bcb0f695a74b87140b9fb201adae1f2b4ee76b5a04b06902cdf6", - "f0a27bd84f139163a87b6f3401350363321faa23cdaaa3039c4a4ce468ca773d", + "305edef1b0cf670e81ab1e1479cb9c4b05cc82561b15b7b6b84e25140c88f808", + "2759c45069b977a975127a511fbb61c1bf021bf10528d797f09faf356a665a91", + "7489b60bf93725ea7299a4b7c94c71ecf1a4714bb9c7794674dee8480ddeb9ea", + "5f92acab79d0676c0d742e5919000a81c5da9dfc017515729ff9c1bde351bb8c", + "0b2a4d877fdc9454e854ce9dfd1bf176d58b8a289bce54506f2e1f356b7983a7", + "c86eebe536ea56aeccd915f4d77449d55f5d359a87f2143977f818947a526f07", + "3906c90445cd5d9ade57ccdda629d4663ba35dadebb246c256273f43872986c9", + "a083be7d39ae59f6712c05eafd809f00315c9a8a3dea4b0386beb3531bae44a0", + "7bfe28cf4668049a203d09b5e161c7e6164fb6b29521d6215906edcd6ba0d3c1", + "9cffe50e10dd0bcd4ec60fc0bee505d32358298a33e4c70b86942c915cc0c385", + "055c4e95dd3bc889e5a3e017d7f691f4ef8ea59da1aacf9b0fd6e134d896c692", + "8dd1a4a982b0ae27d53ab09c5e4453271f869af64701abb60e714baffb7bf0ac", + "c3cd47ee3ac4e0768d669e1ae591bfd90b9a5b0444f730a7c9c2e64525eac82b", + "ca4972f5ec72682d5a49ca5a0de6c33c243884d46e9a9492689b473d16a2402f", + "215181cde7ecb17371b00097d9550cc8dd2ffdef182a2a4af48b8be4b30e615b", + "22e936b614c8d34e35694de98af0a3baad02a0317238671be78489861878297b", + "f7ffb61e5ba728494c54819b3950ac0805bfdeaf4107cf420cec7a1a3f56c15a", + "2184b4882b7820108abd116fa8de8d0db455a71ccbd13676d9625d19079a7477", + "dcfc8c0ec469929a9f18cb860f5c276ae49b51e7cda6f2c7882e118b896017fc", + "05f89a7ce9d59307d1c60c8a66fc1c4cf2d219b48efbe07308cc6e7871669c48", "dfd9e06956dfc398641513b60a6292ab7d1dcd9f169b02252cb1af661aff9e6b", - "50530d3a557637ea436a0f46c383767e349fe524e38e9c84bd00fbf23d18724d", - "7bb565cefe89e6a6724a067b1f17f6c26e1b4b95da625b62384a1b0a7b61b942", - "b890b8ee51df828f19ada7be80d7e8cfc4dbebb39744f82722309b5ca68bb763", - "53b7de44c7d2575da3355c6e29c6466709ab420c16d211111ca01373d4483371", - "a77b5966193cf794f22c76af5b773aba88e4f2d9b88ff8441dbd5abc52fdd813", - "953f68833d337194d2856b345ab97cb49be652b8a6310f0035882f1ddc920244", - "c38df1b688b6d5dd3adb5ddea0826d01f74a4db3b14ddc6a46a2188dc999863a", - "a20d49bef4a3af46928b9b5f4c6c627ec3512021f1f59ef36484b5f3fe2bc678", - "d4b36bec03f928025119d1907d7669377f31d0276337106fabf8bbd6dc49aa2c", - "eee478bd5dedcc2a39a4bce31a256720f2c54e618ebd733521c06f851acf55e3", - "ebdee3a0f32c343a2b80f3be5cf5b513300606051c15b2b4ad6d9093cceddcd5", - "bfc47380877fd8324a435d859330b2762eba165727d16cd86ac7c88e7e601d6b", - "eda971db9cd248cbad38627a20aa1e0899a65936397729f65e55ccb064e44bff", - "b23aa79b96e574b83fe4db6edd58eda7555555721ffa5043921fbadcefc21b20", - "c5b257b83bf6ae009b1c77d3946da338edf7bb55b3ba51fcccccf838ef07e2a0", - "3aad767c3a3fda7d15c386f8568fddeb249eb27777306cc5bf7c66a08e66a558", - "8fa01e48295823021bf089e15a39bc188e8200ba9670650e834a4bb6132484a8", - "b484f89b8e6e2d830e17d612707666e1b8cdc7e9aa6af796c66ea0d83385250a", - "deabfeea7b009e723a1f55d3303f581abaffa881b6ea373c5c242e56da63aef6", - "bb5c4fb583ac6a26fc2d6a5bc8bb0f663c04e8bfe25bf863265bfc3ae63ce22f", - "81bc30640f4a434e0dd72ebd38c9ec9acd0fb36da1ade4416545d67d1b827600", - "99ee208e09dfb7a989acec22bf2783b32bd2fdd3a6784e2b7a69fc12b0a43a6b", - "b557c096bef4cf6234bc3b7aaabae69cdddaa8733edd2c495a6767c5c0438cc2", + "5f45c667e6acd15b555b81a1a7df405ee1ec71ce5e35788e497191d17d07e7fd", + "c3b886fb35de3547c2fefb1e2cfbcb43c182d0e35a692e28f8e5250bf3af9dab", + "893ddaddfe23f0f126e2daeddefb8d92adf221d34ed272acaa1677170279207c", + "869e0f1a3d7ac86e8bd34b63c2f4a6b23bf6e987c9f8d7f9e345ffbc6619076f", + "8c0be5d24009fc9c88bc0834eb07ad455587e59dac0e94a85b3bb802e24a53c1", + "cf990d9e60a8d08b0ebaa4fe0f4fb49719b0504f516978330fbf2a7b7f48584d", + "e7d229ce6e362b4c333b0f3ae3c38607191df12362c2b4bcd890b6a0de9aefe1", + "34775f725af76990ea4a07a65e1af7d2c1ce9b94fcc256bc7ee0a80b113b6193", + "568c1427e54f35bd9f0e302b7c43a3ca2f9dcc1beeb80da805364230abd17652", + "d5310baecadca51a64e47e92854fdbe124fb7a0ab75a1ed13f73b4bb5827da03", + "34166d197118c56843f1a665b68e54f24d3e112b96cfd1c2d2319cdf23d2df56", + "6d89669d80b6049ab68b768fe9e102f8d02a12a5cf77dc7cabad6d9350d22c09", + "3bcef10e3622b1acf69906112127dbf3e7db074f6df337c48e6a441fbff44f35", + "1c236e184ee62b55382f3930a94a1d54ea3c237acd3ed526e32aae9640cdd5db", + "095bfb62c040ee552dc4b855b94762d53c328b7e31266b3e7368886c073a96a3", + "e64fdcb1a85584cd9438cf9682e8384c2d79e6226248c6de56ca2c10e98cb0a8", + "c753a6be718e4c97d8d312a5513c933f478219cde7b429afc5eed4f12de93fca", + "6ae4e867942743ceeb557170419e8862cfd7de41ad8f931b7b2aff653c8b0678", + "698c86654a4c00a29a19ca8c1a12a59e25a9eb4f1db8a3525c723d157ee4456a", + "c5f73468613a3b1a4aaf58710488aa810cccbf6501cd7d8f922138df2d72fe78", + "a2646b2b95431ca846d17770a466b8fdbf7b5fc69c0963e863e9785250845d59", + "b7fc9298fd50b3213e43cfb85d74bfa7dda139b6ba355fb770fe72a48e748104", + "c9e8b02d4dae6994caade8da033cbb5674e40210b151e39d63dc145dc51ea5fc", + "0b271f5d791927ee7decaba2cc38690bac3d23978a8e4f6ab0cc85fdbf8bffca", + "d20aad5d3f07e011f8e1c94aec14aea53c500bba99e97dd002f33a42d5b7b01d", + "20f1a5d684922bb06b43713d5f774814d3c15bab06b76f1bce77a2a2164ce4b1", + "7b21c6477b003da97b17f74fea7fe4cf1965714d5583652ce67fa2421e16181b", + "5b2662c1afbdc0c25b63dadc2a0feecf16b0891c106fa29a98ba95ead98e01bd", + "4d086b50e9b913a4ed9a035f304f52c54de6b5d6c761bc51da61b987c454d564", + "58415ac3dceeca6c566f948f483ebda679ca81be76da6873cbabf082f0023c09", + "905da64c60aaa2a4ee99f785144c06728a96406248714242df945993c39f13bf", + "2a17f6a337fb9a6adf6fa99b4694284bdb67478110b2df86920e77980a489129", + "25fe769bc1472f65f4f2fb40dad84f4d18f6221546adbb32d22e38b745de1703", + "9cca312d228aa6c5c940ecd194a0bb693cca7e4493cd5267dcc802e341f1c74b", + "1cfa9e7e779c66ea15037e3d73b0a22bc94de0d1c39928988dacee5b087ab7d0", + "085edb182ec63c17e768942b5e659c3e8f6c8b847481dcbbcc198dc7f9e0dd3d", + "08026191e6237e6ff130ecc2ebfa1ddada090cf5011c80d371a1aac4bb6eef9a", + "575dbc53b7541580404f954210700681563ad508038479afc9abb65f6b148777", + "939dadef80f8e60fdeb525dbaf377b5367ebc9f035bf028f9acc1361c3f281be", + "9ca495bdf1461f96e5abeabd379ee3b75c6d2b94c5d8da4fcfccb61fe9596e7d", + "b2e04a9c81b6c3446eb215875138c1848203999e8fde9e2392bf8443f77ac9cb", + "383becbae9a0cf0011978643fcac2c21c7ed700b9ef436e901b90a1e475047b3", + "d84ba38899472aeda4ec230310b8b0ddb8f4e6f64fb732b68069ac14c86160e2", + "eea379683c70f889141187881ffa7174782bf4974824c43a3fa15d78aac8c0bb", + "b98e5d24ed1b51b93f89f56bb1dfe3e96295b72b1b0299ffa22f2401643445bf", + "818a7d4cdc3956864f1777aaa7a18cf92d504e3c78b113f1112e558fc1c95833", + "e9968449b9216639cb7283b2bc1723733edb2f5b1929915055505e1012e4bd1a", + "14121018af44727f6edec16ace31a0caf6779c5c853b051e1b1b995d2e13125b", + "7a4ba636764ee16b89757ed75416c1b13648fa494d5f5b1212cbb57146541047", + "98c7b82001e6804b5270ef585523aefb308a49e3cc3b731153f98660aa276992", + "14aec1f89dccb4e8396c28fd2d5d41ff52200ee07c2cb1961c2d48e9b81d08d4", + "a02277e9a820346f3a3cb34baf0acdd0e339d874ffc694e4b4f5eb6fd1fbdca4", + "a14ec9f5521e5e49fb017124de80fbc461edf16c006de1358ec16ca752924988", + "0eb299cdfd01dabe4b7a2dd51810a8f0f4b9482b400526086ae95e29766b3b8c", + "ef506d334367785f9d9fa9119d81206d2b8f9579a4ad2d821497d9a61ac07bd6", + "bcf034c962c3b7b85bbe228a6a49d4bf58a22323986ad23a4a88c0d6779f7527", + "657ba00436ff2d34ba3c1a220d9cde58e52ec68e90183cdd1447c21a49d515e9", + "e1758efa1553af17117f5156aed4699fa4f9869651b1ec6a82b282c78a295af2", + "f6b0c8e156b00d7d2645325a7288b12de04e93b9e7978b64fe28d03e359d7a0e", + "7d96f7a77ed3226952017d0937d2e190387500a39c84118a13bfaf43fca36cff", + "a7b6e5e0611d5852e097bd32f2463c71b8abd413b6ea6cf61ef27be7d60919e2", + "88c20eb16c6e6ad5004a9fd1d06860d6965d2dd4ff86b6f64b06b88d7c2f9e24", + "2b6c4b126d8d38e839d704c3cecdd71814acf09268e80e1e86b4389863898cb9", + "1d4cdda632dcc31f58493d60c817ff78b8ff850fdd02d3890d99b06eb72288ae", + "5ae5730e537a1d636da11015493ed2dc1aa4dee6d43771a4e97b63f911def969", + "7dc846a7395606c43ccac91e131027fe7380ec48d680ac487a03a05f79d4304e", + "8a18af0ad5c83320675ca176f7ae1f33666662485b6abf1d402df859d76318f3", + "bd79a196b1e2e23987ede513b3986fe4b37b3075c31c37da0aac2bc2cfb8ba87", + "702448520a4a7f0bf04af188e496a14c7e01ca17e7a6d92df018a9c245454256", + "c26e9152672ec234414d370d250255e970e7a0f11e5004421d76d805e99ccbf1", + "9dd9ccd6e5d34ce9a55bb1c7bf1406ac3bd296c894cd73d66bac4294f7a313ff", + "614c4be2522624c064e957fcbb372e9ebf51ada45827da2bb52deddcbb51c61b", + "4b0e0c2199b94cc21f8f3752126000d63840005d5ee2fa063ffd56ce4772a227", + "db2a8bfb9e1c3f65785804a3035a9f169d4eb70fc4224fac6c447cb63936a336", + "e542a90741d7cc806f8d5ba7f160d5015bfafafec76187e858e7f5b4523b82f0", + "3ed1a99213e2e1be6884a8631eb26fd0d72f4a681497e2ad7e06712f902873f2", + "5ba68fa319f669e6db9dbf7ae7c5a8a59528dbe5e41840843058eb7a294347cf", + "b0f42d23efb8a05087d8a8432d425b091dfc0b924e783e0c74a9d6e43ac4d41a", + "0f1f06b4fe2a9219bdf3e8a2896ad05eab5635ddd87c2cde26431b4b048e50dd", + "b0d59fac1af486e6b3d0f79111cb06dc8cbca521b969df6045f287fcdbff0e4a", + "60616b8594d7316b7d9ce6031f75c4430724b1b23b65c530ef08ebe487b5ab20", + "f76e3fcd1ef9977f972d19d5d73c954a98b2fa827c5300adb9c2a6a60d00adba", + "627b52de262594f6f1aea23327ede39266c1b3637486e51af16329a6b0991767", + "0af6ab4ebfd934be54fc792d76516eb96a6be486c2de322ce6be0e65c595cff7", + "ce31436539b5d2bc6539aab93ed9351ba0589630a5eb32fc08ce0be5a4da4011", + "8fc9269fad11347e381bf3dbf0dc91af0a6861e1c81e623ee128e284aa8a2935", + "09d9be85b8e9cd071c5b0c92f58add7a4450af99db26a6933b8c2ecc3a46dd99", + "bc939407a6147f3fa7e8edd1c571bbc68dcf1bf20916f6122482df51c28f44a7", + "44a036b4bf5da8d1eb569355b5d8175fd1c153d549c9db2ed83b83fb2a6889c0", + "5175702e720ebfa2eabcb858a27bb8cf9c6122d3566745782b67fa46ed27b946", + "4933f477ee38ccb597bbc512eec2751ef24d190df6d1415f166df39c761bef4c", + "207f416b25bd3f9932343040fe7a32548f67e99d2f07d5aebc36388189b85730", + "afdbd3166b358303cef5fd8fbe4b7ce2e4052ce8a8ce623b4b8eb39a12287d7d", + "098a067e3762286a19d8d7d0edecb304bf19cc36d36d4fc03066305ca3f9ae40", + "520eeb7860d134c364a8b71c757a85c6a0232f08fe927c604a8b102cd405a6ea", + "bbc97928005adff5810b44978bc8e7c9376af022dc5440da268eae4912811c48", + "2e87c5c7b82723045f41ff885029afd0c4acd5bc66e2be2587b80086b7ddcfa5", + "87507067d7d571be5d8a78ff3be612427df987962f027de60be8cea9b44852ed", + "8be159654672b3ef67385dfd0ee2d34acbe3b591ee3fbd80c60899ef30e474e7", + "2ddde07daf289634170168550c44c1855e25d595ddb37b3c1b1810870c852a10", + "3bc76f989bf85bda4bbca90ea42a25305d7a3068af1c5efd1754b9ee8a966f85", + "e8a55372cf638f92ebc1351046c6099b91bba6d5178633043bab8538fba8fe7c", + "15b3520aeb6f273fdb8911f4caf8ac46bb3892d93e04188877218b7a27940cfe", + "93113bb61aa61515008ed794629edf0d7a0e6186f3651891ec0c429817593d64", + "12c0ba74381b4cfe1501b58adfb009044e7ae62884a1cedfd0d20e3bac122e2e", + "b2f2cbf8e19602c009c5d1323d6bd02a638a99eeb87170e4c880344f3168c56e", + "3b8924d8017e97fe9554863057d272f963d9e660538bf02aa9f61bfd7f3f6610", + "db0997d11b15ff7bc43e58ab6ba7dddc54fc92cd9d409f6d804b1a57c0769a02", + "bb3127491700fb5618363a6083f8328bb2f9040e6cc42a7fb026998077fbf839", + "778b656b96e0e8c16d6309a2ef7fc85761c94fe3d78d2418d7302fa3f057eb99", + "a5eacdbb5b31296d6ec210aace8a17f9dd4069262f706b29ec8aa0c07c6da36d", + "f6d863079c811bf0f1d1c2eac05b4f89f1ea63cc051d7d2285dd4808707bfe53", + "7140f67282ab54316be60a25ecbb8bccda0b5b21f7282c81d18a555d005ad2a2", "efbfe78bbad41db70c37c4840c34f7408ad222edef9e8ab737f197facecc67b9", - "d35e42224b7ec69f20fa429c814a7b82e1d84d591bdc2386953275c4094e3218", - "b9677d96d367edd3f05a3a6051c2a9be699c3b6c84e3dc91dde4f5cc2e9ec78d", - "d8fb15d264520cb0cbe5af886055a85ebbcdb076c3b7a737a0c45a151b612818", - "19100463a955b0f84c03ab40c8fbcafb74369ba3cc0b5b438c2a19b3d9c5ea6f", - "0586e29c7a4fe60ad3a0cddf5bb459408227d8a5b3b3995a8fcd2a5295673c2c", - "532909a31a3e88c339d3514a8abbf823c3e44f7da73bbfda6f45174a4ba7a974", - "db4d1876ac38e9d1982d13fa68d895731d12b41452dec832618db485577dfe17", - "3be9ec5d4b5b4eecb463f2b8cd5dfce4537e221e1aa501d4d28dbcde41858d16", - "f68d644cd8f46420f36cb6d70cb83756c6935a85949caad3092fd1f4501b67e5", - "e2e4dccf0a6e6ac01f8e258f508fd533db632b2f66ee14de5b54291b4ac15bd8", - "1855aee933b7f613a5981d0258eab97247dd8bd611540051b1c6becc355b337c", - "2bfa46be8b100da527d40a8b423157703a2e2f057e07dcf8f11e2ebd2f77650d", - "d61e4989090de493bead4770b40bfd33667953fd7722e8d7efbe85c6bf4bae60", - "a083be7d39ae59f6712c05eafd809f00315c9a8a3dea4b0386beb3531bae44a0", - "ead1520248fdbfd42fb2a6068abee0a7bb28e96906959ecff49be57bca6a61ca", - "fe475b13bfb6aad5e0527a23e3c16f89240418ae585458d105a257848eaa5a71", - "b01d8e1be3240cd6a4955d234f98ffc818f6e17ff1971e0c534bd673cea79adc", - "64e3fb7cecb3f198fe952b2a19d82c4f01edb3e48a2f73f1eede49f146dfa0ff", - "908df8f772daf6d1cf68cadee6eb23e6fc671de8315e8e7bcfa9536862c9b45e", - "ebf22e696710572eca779fbbb856627e31889ef1b0414a8d19dc4a90862ec25d", - "9b3477830b9a7170b550750f26eaf475d10fcf9a3bf08d4f9343c725d2efcf06", - "8f62c96296f59440bb64dcc810a4804468b79130cc550fd8e7247b0facf7ff44", - "5e00868b54a1dbdd25aa06133a10067ca8563798d55e23b4be0ded2846e1ac05", - "a28421bdbb6bc6c24cdeeef7180a91ae1432b4fcda0e2b2e7d31c584e36cb785", - "71b422ee5e2288817b37bd8f2966b57ceb11f27fd2ec184bc9b5543792b5a3dc", - "ae0bcc01927cc130f8b82e87259d0fe87c76255e2841921fcd761e64037b4d0d", - "8b781522f3e233efa32373ced643d616c11c401b879ddd784fb69a5d5e23283f", - "3f37d264e4777e2a3f6ecbfaaad2179dae1d12b619aa3ef4eddc11c5dc5b9678", - "a4ef8357deecc341996ce8495b7c2353dfb4d2963bdb0490d4d40b5db9532d94", - "2d8975a767cc0d5ab650f12a0e702240415915c564935f2d9331ad57d74ed714", - "398bcf4d923de672e2ad62c0c04fc3e7e7a8625778bed39580638a2095f61204", - "278b23fccce787924a033b8ef046ee7ff0461ca1e8d37745ec08fc9adaaafcfc", - "7103bd66d9797bbe9a989834165898c334430a226295084ab41371583491516e", - "80e559e017488f1ef447bc74d92ac893a1d1ffb0ac8927cedb382f9adb90d758", - "c9e693b999a8634ffe000cf8faf465f03d52607e09000eb9600412b75b5002f6", - "7c30f28c5e93e92b04a787f1b4e7de815525ed9c5c2e40a124767fde7a932346", - "4fc022bff2995e4347fa523551af54ca5679d619a38a4d4b05402d899628d446", - "bc32cbd8f13b4ca8c10d9bc7844c45ea2cbe43c77683c5835032ae985b8c4416", - "bfe759dd43de35f3c3c068528541e531e32307d80d5781c661c2e04f1e2159d2", - "06b7b945365b9bee6a62876d1da4f9d299d242b2e9fd3d35a4b8e086502469ec", - "47fdb94061bf6ba64d5ef98d5c71be4c0846f908a0b2eada775d7d7aef54c128", - "bbbd9a94dfbb944a68d93fb2562da241a39eb4e5448000989d1a671610e087da", - "ef684b13df1c74e09c7ed0f46c7e2bd0aa796f495b0cd2f13a421a6d4c67cc0d", - "56f5950529dd6b96cdc18b4647b5d2f89e0f0b70624fef4adc2f8a6bdd91b7ee", - "5bc94b73df718233acebd5227a3c66a6dcbf5a4670ac6eb6e05e73be6004620f", - "c24e68df557394381f2e33f766f676a0c737a3c6a9d6b44548e8416bf7ae923c", - "51d7ddb52464d3821ce010d13c63d689fa6a58184f4a75afa1187f144c217fcb", - "8625361ec0d404519dd589fefcf011742e283eef93728753fcb44e80ac6583fc", - "4e5552190330f123788f68c09d8a4dc6305d3993e68ec2a0c458b38b668a5689", - "54ec705e69ef9e37a01c545622ebc87b9e32141e58c06bcafd3837008b08331f", - "61f4a203bf75640c0c016ce7e759c1873d5fa4c7e34f4285b2ce1af2791a14db", - "7ded86877ea618c8b1dadfd4e9dcc8b9331bb0a92ae0ff161d1751dc0254986e", - "d214aeae374fa16b2bca5c091dfdef85c3ab2ad3e484dbbd53b0ba3af4475959", - "d292175bbc55e5dab1843fb8a70bdef02d44e468d6233a1989c4bc956ed47d87", - "25a5374ba2e9f7034b2aa9d86e2d9fb14443e61995583a1cb9c51ea85df02dd5", + "77cb9078862ecbb67c370c7edb7bbb3c5bbc36843717ba0d249b6d4d8ed3b331", + "da9864d075980ad9c48e9db550c22c632edfdb389a75fa666d0cbc0851aaa2e7", + "54720b87ffe63ab38bb4a639b323fa2d151985368bc34a224d846d5fc51c6e3e", + "b25195aee1456589dd06d14bcbdac17fd3c96fce614113dfd1664649bba83063", + "e66493ab25df931bd427038c8cdc7165f839fb2a8a31200f0e45ae35edd80712", + "8412078e2370fbecaf64079e2f23a49df95166d079863033a5b0125e3c8205cc", + "8f7c4398ff9b80ab3762be84a1016f42d8d54429b6ba485219bb1c333049e94d", + "e2abf0f8925ad2e73575577fbe13cf98644f8e51695c048e788117f071c357f3", + "1611522dbb2627efd575794a8e57ccfe4ce5805add9420cd58d7f738f8021404", + "ebca36dd1e243a6a1d11141c370101e75cb3f6bf2d8dada824102268decd6ed5", + "b0f25e7028e46288c3d2d5178ef224c494a2762db85181925d9991125964c1ec", + "59f45f347acd68b96a0efa69f320f5f5f421804797eec2d91710ba6e1af8758e", + "8ca75a99ade68215389686096656de30c2432406d9f0e0bd8d868df3429f367b", + "4c431a599b292ef9cfc93b0af49ab27ebe81c5bc2092da6d964547a3c0e81a17", + "23556216252d494a550ae0b7d34519219848f5b0989e46a6d8498d7c1ac9b4e8", + "5e9fae8e410801ab00657d5d72ea593ae95c94591dcbdb2d236d3e7f3519195e", + "d8067a2d643dbf09439d2e9e213e6ab6b5f243d8102b7f39fa7845bcc39d6fad", + "0b9e4842c8f864ac865772a4f51f7be4687eb147dfe75616ed6a83bcbe708cfa", + "1f4c1ac006cd05324ebe2ce8e15724a073ae1045fed68ddf0eacaf8fe6bd5b02", + "fb7c52f0b82ce796899ad6695bb2f7619711fe559fed3b6c48445162fbfb9bec", + "c9e665f7bd5a35f874ae1dcd909290ef6efebc93316663c2bbb23cb7abba5960", + "70fa9fdb0d6b1b49bf64aa7379036715d387ef74af51a7f5b7a65b5ad2c11f85", + "cc8e592585edb4b3d2d10a4d22725fb688c888945ad1cebfd91465aa8094c9be", + "9d2aec493640c33ab01830e698c10dd8e7ecedcebaa0c46b0fd67dfe9f0406b9", + "52f5bf7dd1ebcdc0b758db050a6d3b423a3abcc8397ad57ae4d1f61fd3dcb1e5", + "1c9a29543cf2b9e4e6569f3e6f11b014b9b4cf84f367919aebd58eed81ce5d9f", + "a2b23772ab5da9d1f7f43510816383b2cda1c9031c98708b5ce4a8e9ce952ebd", + "a552aba1998ac4d574bc68840d17a7d341d88bae0372a3bd55f9bb541537f3dc", + "b5491ede25d7c6d6daa0eb2b949bd3754b6e91c6ca6c3833943872070caafac0", + "e925967017d102582351b3c66dfa27e149b53e40c3352e71e9a18c5c9001a6c0", + "237f6c155009780b1fc8cb06fd208e3efa87eaf2e2082f7b56e9ab2066b06f70", + "e9ec3f22fed374b911630664100825c7f815cac34ecd4d4ba2d6f67429b3be15", + "7e5b66b4245f0a766b531dee1e53348f38f40606bfe0d5b86d247c3233766a19", + "b14c78061145114202675987493eb77832f68fd0658492e32401850befa27cea", + "920fdb74b84284b0772f94629d08e63348484f68fa073ec3f1b4c3beff5ea5af", + "67f0cd28c63a7f77f0301133eb5383ef9c1848165d0ae33a0a1a6e93463d6f88", + "c5c190961615db6a111c16386b9b0a04e1d514d930b378bf3a0903281cec2e86", + "61863a9961a0a008d8c9ed713eb8cc96a5e4bd090608b00d6924d746c90c3f28", + "ca808711bd6f54c4f05986287c563a54bc73427f282067fabd35e6b6b9bf9f73", + "a34aa8d4770a68ce986fc6e47efd58fd3a866b6060aba390f130d8ccccde698b", + "cc2b33fa3703ea8c2d2a87b3cce6b28d8c2e807dcc1c5190ba88c89c701ffda6", + "1c3c0c8352028976de7e823ba0ea95ce9b6f6917e3744a34a85e96b26e75d8cf", + "2402708411c33dbd8ad2594a5d8d7f1fb3a518ceab035af539967897a982edc3", + "016f0e037c1955faf676ee05c3f439767052bb627ad12baa5c77f1e7dafffa8e", + "1b063dd8ee6c0a10ec8b32e0df805a90fafa8cdc5627d9c5afa6676fde1176fd", + "6b62a53a9ad9a289d11567b3fb525e9a55cb0fb2f6a7bec08948f3daa324cc37", + "a0b2af082db1e7cc3dab489cb0b0d827608abe7e786f9bf2deaa07b8b37e2d5e", + "e85d9fdca576181bf2dede813da7d18668270d98bf805f379d17a3b84b7388d7", + "e24b2c8c460363001e0e732777c3ee6b14ce3d50bccb667b6efa998a76d85184", + "45cf930852797dbea9d3f91224cf0f87a1b1cb6bb4d5ab91836d80c416893574", + "d138ee34e1926441f905daa58aece2045f4cabffd35f645edfa6101f191a1d1e", + "a520e3e50e5604df0a00070623a4e0c851f67aea913aa51d381277675d708632", + "b890b8ee51df828f19ada7be80d7e8cfc4dbebb39744f82722309b5ca68bb763", + "cb21e7a422861321ba3b312db4b69437dc73e30cea350ed4663b14c39750658a", + "17d90cb91b9ffe6659a56ac6b4c9766ccdf71b39874d61e91e7866edec4bd89b", + "5346c8333c41611f3f3a3a596a4cf02c67c572544ad91b522d806da7314016d6", + "834762685e8ac1d43bf98b2a9f6aa1fa05d4c1d61e3e7394f27c0d01fd485d85", + "a09489f82eecbc72817209cd2b33748163d645a49a91227bcadad26b1817ca14", + "e3b6296587ebe8e840dfbaff8911ddb1e1a0da0cdca0a33923a08dc440fba62a", + "a95c959d1b5eb0213fe6ace31fc401c16f637b22443a5b1e27812b773d723772", + "f4df831c03307bb7938b51c5769863cd013d1769d2b82671e86d2ae10d3d7e79", + "67517fe05e3a305f04c6ad9cda02cb15e0b1f17e04bd9b8deccbad9a02c3b837", + "73b283f23ec174e2265869423064417be3142f84a8a7a67dd5e369bd2f54c051", + "ec8519a4abeaad9522903ee9c5de0eba64c12733a5be0b12c8a0c983161b19ff", + "daadca21a382adfef7d9d9f994fd93a2dbf8b19e17bcba9aaf26bd06dcd9b70f", + "ef7e69d7a92520c4290d475cea73bdd7229b867d8557625b826af9f1f8946fef", + "12017fee4171b14e805640acc647935be4a93090e07b18851ab1336c2d7fd909", + "b94f9ae9f9e622e7380693fc52236666d15030142edab535abc2ec4eef875c29", + "01f04bfdd9dee747eb77fb142a3b7d03191a6962784a9bee0c97c537498b567c", + "90f4cdb40b881fd0c2ab0e93cf336c8bc4ac0c522fb7aa3be41e1dcb006f1ac6", + "149d2536df467fe27f4d43cb049ffa135d65def147eac860f9919e1feb1fde33", + "1b53c152e77ce5a60f9127fe54927c562059b4c15ea27f489a3924a210725564", + "a5f1abd4326bc73cf2bb25fd6b26568093400319ba12a83d60f27391366b2aa0", + "c864a835aab81ab605f2a8d2c32fa354ff9087d69d45ba3f580eaa1543525106", + "aaa0a4fae5c64db8077d2b796f8143f18a1dbb10353ca3ba963253cb73437b3b", + "7e36513dee1f1e045b70a2afce6e8da36d72e52251dd7bf3c68d57f9ade3c276", + "486e0462f176cbd01c8cd332c01ca614f1e7cc4a0954028e2122fdc95b4248b5", + "288e096c53f853e152e6d49973021715744ad0dc89bc003aaed7948d2f69fc79", + "e5a1711b0ad0b813874e518e393efe3a7a4645b3199d2ae573577203aa51e685", + "9b0c9838f646ce09a1cfff7c465d7f96d9ce36dd296a20d90d0ba2e99cd81a20", + "7a69a5a8578ac03134eb4371f85ecb13e6109d1dea024075df29b7fd0c6c44ae", + "9a201bf22ed6bf46abc0b0e9e9c77b181ca8b720f95717eae48328bd67c19471", + "e8b7e0335fc09335befad33be4a41c5d1fd563df95e8b7273dcd4000cd167bde", + "16ddd9595926add7d9e61259f1a19f16345ad9c4349b5caca2af7be4deffc3b3", + "f805962a4ca947fff4530beadc5ed924a4775263e3495fa0ab1730b3569f9701", + "82cd745a0c46efebade4fc550b740462b35e279d3ebd37f3da50482271081ed4", + "e2831d46aeef97112cd38c940ec25fa5a1aa6a151c4993e5cc56445ed97b7423", + "942742548ccc8adf9baf254025196990ca5f81bcacd66be3b16c76cae22a1782", + "a8308d3a8d1a2745518583ee52120ed806cf4748f96fc1656bb9b0c5aaea82d9", + "ba1273caa5e0bae2d8f0d0eab56d884836d03d462f7f9c6cd803276efa1fc6f8", + "9548aea75972eb2bd4fe7c80efa25917fb104659b0ae1c0889717792d6403712", + "68b9bb9d9307d50ce3f91a544976fadb6c4d3924b02c2185a8c6cf74e85e2507", + "30cc8d21303caff76188bdb08c6220c027378fa4820b1339ed11e1634df39a5e", "546ca32bca24eb4829da2b9c5bb66140385c9e8aaa315c24bcb8178907f4fc9e", - "7221c34aae27fb314ff46562c717f8b3dcf01a4789fefd77c9c68a39635d9fa3", - "609d06d9827c058eb94e6f81c493d62401599202d25ef4b02bceb0048367e335", - "8c768cc31637de25dc58238e825bb776d31579d4759afc0ecce073beaaa40159", - "25592e02d01a4f7ff933cb9f13cd390e4efe1d1bc17ec774c319c70629feff76", - "8bdba2d3c56c07b0a3314e7520ec32ceae575537702a347b170c868caf63de80", - "8faa885c67c4652bc808923f6b417a3e9c96b6dbb4c26ff8e14c35655d5fe923", - "d4fd0ca6f861193e1fb632e1ac185b5158ceac3b761b48e5a0411e2821fa744f", - "7f39cfda7c7d587c8ab6fbb83604dde471951d283b86e37accb11d8734bfd490", - "8917db1239064b966706abd82d308f2d35177c007a5c034b2be52cdf0b4431f0", - "f712bd37162800b64fe157503d8f6bbc30e08dd6d192115a3bd65c0c205bf92c", - "ab1eba49759df3b1ac1a13c307641d20aa92f349aba9e926c45de9d32b413b43", - "c153d46d57b456089a2493392c474e372f0e75aca75867476ba7a1bc79dce901", - "ed1852c2974de8094880749b38632999430f696b4644494fa5bfd69c3d86fd74", - "453c6adcd7477f31c4ae4597364b83cb7c602295446af7e4f1f3105eed4f7388", - "6b867854b75fb66d75a40d5d7de7787fdd9627a37447bc4e71c951e2a5f7dbaa", - "a93e7e4b2e61536add921a5ee47c79ba46008cf07ac70bd448415b848e6e280e", - "bbc97928005adff5810b44978bc8e7c9376af022dc5440da268eae4912811c48", - "99c51b1a3f423a90051e6f6716750b9400d6c29acc0f5ab52d30b35eaa7bfd37", + "c17725fb208c3ef0c7eab4672f27fa56215f6fa92c0dd21c2e9ce57ced74847d", + "ca8a04b709ac76f433242ef4677280e83b5a7146b7366017bdbf8196a48387a7", + "e7ec19fc4c4873496e38d1400c6ccd7f8ab8a07cd0971c1e96b89bfb5e98a127", + "4044cd00cfdc1a636864e0be25cf5025f52bd22e4731e09974e4a6835705f723", + "c521a2e4218927689674e523400a2deb00e01964c1850c257f43f601d3cb5662", + "757be0a7b1bdc02716cf1a4efb4662d0d83fe26043d8d8b52ea398e0291f19c4", + "83a18a933e5ad42e7d1cddf01afc66538b48042db7f9cc1b40a957f8d0d55a86", + "d5091e30c4c5918e802adcaac8109ec415eefe19292a65680f309a79b7df25cc", + "19032838d0dbb239e84272dfc23a3aaebc35b669dce6f505531943ee85e274f7", + "47b54a959963c3b83b4f7e4f75e7b6eca7f61d2576757d8439e81909f7424e5f", + "4de25d5b32af0ed51d35726ba88ced016980686b3a85e3694955e2ae5a8bb1bb", + "b333599d87719e60b65fc030011fdc396ca31f82e3d34a9b35975b0a806a6988", + "25cf4445d7ea45f80fa8771cb5a7149144a02a7104002ada2ed460bff8c7c48f", + "3ae8dea85749ed673af5e3f8aa5dbdf989308e7a29f064fd25678ad00effbddf", + "bce4b2bfff513a10ad20c95ba79583f4a1525bca7df9ae07381b15a34d0ca666", + "43315568d5f993e4f60bd393af20ed1103f0ce251f08945ec2aae839d3b8802f", + "c7dc055a94400b0a3658b63e9b5e8a861517d8e7f983fcd54f158234cea4fdd0", + "e84353c4d69bb3aecd49c8f0bc74b87cea78907548040b5be318a293bf185658", + "462edae7ec8ad8090390e7e01cb356b63f4d1cc1900dfbec4c139089501b474e", + "d32b97e9d34ddcaf047309ed6f5ecb532b22859496739d023ba3b16f7b56c0de", + "717fbc32231a050de665db2a4ced442c78511fb2385eb1ed7a1b6f011d926896", + "1495f544b3a91e2746e306bc06be5094a49a898d67835c59c8ba6d1ac6058c5c", + "23f79b84ba2a56e28bfdb2b9b7e94e07bdfd17d075c1fa61b08be3d1279e6811", + "0ce2f8db431adf460e619931c4846fa4cd1d2d423a1189c18afa80c36563ac2b", + "6b8b826f1b266ef82be5fe3f104d21eb4b481cd9280678839ae6f73897637612", + "a70693501e2c6a7f4f78d0c8cb3274c3e02e90d3b651263aeb813fd8ec4368c2", + "888a136b76913622cec8578907ecb2b5681f290781468682f4194f216f9534c1", + "b4ba75f722cc951314150de348a3456329fb29ed4b46ac753f6332e2b4d732f6", + "da0e166c86288f7980986f0cc30e3539e1792dd02e31f22204d253cd35e14817", + "66759e61754d96d07d89e05f7b5deb18884e8e4bd8fb02f5803a855683b71bfe", + "0a955c9017fcb2a62e904b2f09a8a5c6ec6125c3d9b45c32e5c98146c9a505fa", + "51d7ddb52464d3821ce010d13c63d689fa6a58184f4a75afa1187f144c217fcb", + "8ff1934cfe3ff0e33bac475e1b9facf73846a6d889e34a448e1f738436f11085", + "89958987ac19c33534a9e6efea3b8784ff87a28d07d66b5c875757248c8b59ab", + "71b422ee5e2288817b37bd8f2966b57ceb11f27fd2ec184bc9b5543792b5a3dc", + "cb6dc2d09d2c3abc293cc66c2cc5e77dd4d3180c432018e958146dc4627cd06f", + "c4353a7787e428954abfeaefafa4194e0986f643bc64973f89ce6ef0091412eb", + "b27816eb9f177beec69e1cff770c33aa0ca64ef713cff23e2900eaf9d694515d", + "4eb778e511547b152095f8bb72e52a3c3c197a67fedbc63b2697f0cc8b8dee77", + "aef42798c232737063f9faf97d1504b00616079e224770a733c3517f4dff1d39", + "ebf22e696710572eca779fbbb856627e31889ef1b0414a8d19dc4a90862ec25d", + "80e559e017488f1ef447bc74d92ac893a1d1ffb0ac8927cedb382f9adb90d758", + "af4517dd8daa80524dc69a9b3dc57a937b5537fd7c082a47db6731d21c74bc1a", + "179906c1f26a59c5a6b787b3e150628b8eaaa18785f3d7996d4d0f64da66699f", + "85a83a832e4bc5c4341c64374382d609ea1825085f3a295c704db9f1edf9ee9a", + "9176630fcbdd467b2d9b0e274bdfb8d700acd8a1a8bf9d17062991318792e08c", + "78bf714ca5238a877d8c09a563c3977f138d1e2fdc9e21d06d776f375adceff2", + "a0802fe8dc8d8d46cdea8e2f82c519e92fd2220ee1f62aeca31e153134a80931", "106702b7195ce8acc8ed1fd89515ddaea988786dfcc3ff12034b4ad340dab372", - "76aed0b5cfdff3e715b9f99587223e5b5645aaaa328b3d59169ad1cacbc0be41", - "1ba79cb73e8a2fea8d604df280c66dcea3bab9fbc77a170a77f90cfca2a72bcf", + "5e1216b1a8ea2cf354491d7b9ff065abd4f42dde7caecca6fc2926b23d1f5bd3", + "a5f2faca68c606c692d48a95c43bf0524d80f274a722ae78c6660308b24a3a91", + "1102296e4f840502510c4c0b990a86d135e1a3d5f41eb5cad98be8d1f5aa9eb7", + "8ef4c856776d38c553bf55823905ccbabddd95844683badef73de3993c4391e9", + "c65e576a2593258682d5cb50517571fd250b256b7b80f9be4682f79e0287c3dd", + "f66b3571315d21b44a92d86b849a20a77a75436e816a0334da2a267f54c7edc8", + "747756d0718699379e2fedc2481e6db571f9280db6de1ff5d1bd53bf83d399f2", + "3aad767c3a3fda7d15c386f8568fddeb249eb27777306cc5bf7c66a08e66a558", + "22ec36a71c86a9b373a1953e5fd6af2dd3ce32a99dbcfbce719df5d3a9dc1001", + "32422297ce58c0d60671eb9e3ae9a3fc7b95c027d13d7923126c702616ba1b41", + "b4c3d705c00bea6955ed3e3e8e46ab428514d579a2bcda74cf24dfe175ffb841", + "dd6cf0f7e8ce0891f50f4797d7e157e9b136bd3ec6b2e107fd890331794b423e", + "4568c6d1817c9e0f57e377b2643cf4decaca3b305c3f139ca08e083862f992db", + "b245a710e4c1400cccf09a22a5274fd36208016a952be5a53bb419a52dcb83bf", + "7cd0e4d7b66eb35c61b2031f4bf2adb4a9b2034452ef607e627faa74d5a5df7a", + "b644706c96ac25d44975832822dbebcb2009c17af933eb2957eff48225675ea7", + "f0b330ea4d667cc13d851c0db4582c8cfdc18c6e128f8de2c57a94dc50328a15", + "dfd6ea35c98845a9f0db8b59ee05f126d3d622016a5b88685fd7d4c87fd2f0f4", + "d265b8908da0a87bff2064d03c82625b5c8ddc18f4dde16769b337319526a6fc", + "df15ec254fefa4f6c0cac19be71fb4c952b2786b7365f0a835f8fe74a7ab7994", + "b46151359446f0c033b13a6bf646b6d7065cd5116cea1281abd618051d1d59a4", + "e4d38a3277367bd291c8a4aaa5437153423f6c129ac9a210097845bd50744f21", + "5fbd8e7eb0b18d1eec920142d1c8258f74b35f7d5b59388bfc5d7fee133f45f9", + "e02f26a7c281c8e50b5b0bf580dc8fbaa24e12fed5c92568474d00564f0d268e", + "8e37db751324f399fb0434df128fc5f4fb6b352c549160bd27eb1e8f7e61cdd6", + "9bea63e6d8a7d06f6e7883f9597563e6c3f83255f650b7d8e36589c087155886", + "21b4ee1cdd299b64951df77a4456622b81dcd879b515d597a8b57fa177e3828c", + "8609ddf74ebab4ce881a453ea3143b515c30cf10c76665de47dee8a11e1ab00a", + "0a14eb3dc47a389d5b3d0ec9cadcda07acb91e6c221d86c84b327b24f9993649", + "606c66677ac9838409ab69ba792a89efac76486badb418c0f6925c13f5d71e62", + "b47ebeca524ffb7405651d70f9d58095d137219f3e85ead20ad48647067d6fad", + "2027a47c0b043d5f64571af35d9f1db125409d12549a971a58aa201b8f2b0488", + "169ed5d7fa4bbf2a77c0bf33d384dedeebf1f095970ac880cf87397614817dae", + "3a5056340ecdc0f0ef65e385fc465fb9f0f8493b7c010202011b27f3f44f2632", + "e3997c18af039483d270d8973b2079b3578b2a35d7e8cd181a6e096e8d8ac694", + "e5050726bb9542d12672bdb921c8c228c5167b0c43f4a132ba93d8dbffe973f6", + "18f05fb4071e6c5c527dab37c8b1eeae1ab8f2a557510f9237a926d39549c0ce", + "4d3f821d6c71530e81eb605626a026d7c3f7554fca2ada2a7b3ed8b3eec8bd22", + "245ad4d3c66e9d84eb6f6dcb6e19b7a4145f61b46436c8799aa5cf259949ffc9", + "3683846350dee74a31be88137b7b385722830f310abb4053dbb5ca153e0cdfa1", + "65c8b3935525b39c80f28c136e037fc357252b0f3b23d748d92e3ca514baab8b", + "dc84d847ce70e2e11812c3bdf4ccfc485b397f7950431967c9ea92bdf5269d8a", + "fd14536acc9de50f2adbffba76bb0b28f14183b78f7a65432a147a40afcc187a", + "ea735522d706a23dbb9a467a8dacf70f2c48f5db4d452605f84320e1dbadf46c", + "7ca27e92c855ef83e0be8f1c5ad8916390c83ebfda84fb761e85a21c1abbf5e8", + "76706ac9cef835a5c648a470f0d207a5e832262ce1a46d57441b0ca51a69d915", + "0c5279cac3a91211413eb96f51628174442204e615173ec7053fed0b78059fb2", + "588c717290353ab882e846246f225c0f12af40eeb15b18289542add9168406b0", + "8518affcc6c2bf79746e315d37cc0f354f986f615e6981392dd39f8411679a7a", + "6c42c9b166e40ddac5a0709f448917cc83194982fcd59f7fa8aa67c9f49bfc00", + "4e0a4b3a68e0ccc04f730a1dae51cfa553505fc238db300a137fcdb068c556a7", + "1e84c9f6dba78950be66ae99bab3cd89f8c947252b0de776a4e58a8a59f59922", + "e9436a6d77cc7c089a345f2b5b04c29b2f1a1b3aec800ae17278585df668a780", + "afcc51d81b0055ce38041ae64ed366d0c0863f0237c9ef216e37860223210baf", + "57534e8d960ed593d9428141deb135405c2bf3b78797e5cb9c50521a5faf8662", + "c153d46d57b456089a2493392c474e372f0e75aca75867476ba7a1bc79dce901", + "a41290b70ee6d5ea9b50e7b94a3b227dcf7dc24a6b309e7890c0129f7e7ef270", + "14bd4256a12a89c61a9fc367bcb287fec2efdf603221a480e87891d352f251f2", + "20a513c12cdeb7df3ae8a1c1f7821751332464062f86487f1ad299b75a512e2b", + "e339e5a9f9ac99c2c064da364254fd0f54c81348ec997a52084eaf322cc797a7", + "f920ee3b9be0a8908d02a4783cfc381fe8cd1a685216fac4ce70b6b60a51a47b", + "64d562493a5bc0868e7110e84611d9eac0ca4e3965f1a9164fded66192d48f98", + "50530d3a557637ea436a0f46c383767e349fe524e38e9c84bd00fbf23d18724d", + "a9c92cd819151652d8dff0ae55878f2c785922f21945894a2b5ad89d2a503862", + "01415318ec54b0ed51ecded88310347fda415bfa01fe951985c328b7a75c587a", + "56083ed54aca6b0c56b16885f1355b3a7205948bd8d8d4cc7780845656596091", + "a1fe8fa11ff1f0fc05c84a63d6315fb1583e39e7c6bd2c76991d4754d79d3512", + "ab5517cee56b7ebcdaf469cbb9d8ddc9b83c2afed5d658ee9d1029c0d91289f4", + "43a9f6227e0bc5d47974bfb1fca3a5bad87f883ce4ec19a08075faae23fe4983", + "559a78ee1b672f56627d376ab726294aaccbe218dbc435c2201a813493bb1e83", + "fa40c5c696398716c5e736647a8f1c943c156b7fd131547a487f0a5c7ad610e1", + "444b5601daa54ce66083790665a7779f9f825d6a071ad0edfa6e2b58618dcef1", + "5091307167d2751088c334ce636584cd13a8b010ea14dded33e7b3686de1d0ed", + "3f9b05941242f2061840c298e0a1d7a88db27579d57b561eb8e1c837e82d9a11", + "a21bb69adf78b0d5eddb3397ede7615168c8b292f14ab3f8dee7496f502ee174", + "aeee321ab82ce0fcceb4befbd797409ece36f43afadc994db1962d111f7bd76b", + "1011245303ea0fc2caf6630d14255d950063e7459610df5eb21eb912a8c3ea8f", + "b831c855dbac2708703d8a5e8666d2cdffb8a16a9f1e4805b7b8d56169e0b6e8", + "d3a6faba672815a2aa7f7385bdbb2464ff64dcc5110dc3ecaabd6ab97f91f70e", + "23b09ed0fa27598b3823947543ecc92b6004cb7308b3a54cc741454758e70718", + "2e8479b737020b6d74be6983f2eb03f4ba112aa9e155c3d3a0658ba2e08a75fd", + "e685ab92dfd122ff9db0646d5812db3cd7f085a2cb0420323b426db678405bdc", + "3156f7c45dcea336099fecb5efcf9622f2b74cea066e8d64e9ecd87a106c4062", + "6ba4692ae0ddd19b0b9780ff8c998f610b92b12af2e89267eb2bc96ad678ecf5", + "015f31e83c3b708ea4a087371d90d2d527769c349cf4ab7d981e797777295e63", + "2bc66397125950c67a5ae0561a3ee9c603044f907b1a3e66ba0fb1849b4910d4", + "086017404783cf0687f40db7483744ebd443eea3484cdbfbe36304c0db31df56", + "e2417eb41be1131e4abd7d8e8f1991b72e9052ad5d210c417f3a94bd0744e999", + "a7c88ba9e0473a0bd7057e7b9c0d6c03663621fa2bda0411f8dcafa6f2661c34", + "7ac9c0377d107fff8dfda71c548584252c4c4138af0c7026036a59e8ec2c14ca", + "084b9a8077db8f47311ae8c6b14a049f6d9701726460e690117f3833821ee501", + "097bdc0471fb5969c560bdbb9e7bddcddbb6699ca218fe8ad294385d39d910c1", + "9a5d69870862b5ac25342918ef1e416f08d0fad1a46d4e70d690cf953a20dcd9", + "6ac8eae2dd9f03d9ffb84657c4c666ca7589563d6482678856f5e89f5ec11547", + "84dd701b8c03b4af12d563d6dc62bc06290a1b3a6144f89dadd6cad67a677a20", + "d970a351d2deb302be3cec1f35f6224d08e9a73e4122d7afc02964d7f6d0fd27", + "789e1f4ec64d7426c7cf43ba3709e0884c2b899ac916df749bbf7e4b90b70164", + "d18640287a6212454df1998336272d3a4060505e881cf2ac29ffe5d1895eb0ad", + "198d156dbcb993f5716d3fbc3f7c5f3dbd8c07bfce9d15e6dc007673cad8f836", + "52c14edcf774db2b71f62c5670f96132f27b6b765904c2279292a5301c13473a", + "ecc472ba194eb6377a332d642ee021e7242cd0b471675aa341438bbeae08bb37", + "9988ffb2b38e94d97f1667cb422d2932271990e1093c1fdb34e127e7095043fd", + "e446a2f75258af3215ee61025e91568679f14772eb5dbcc7d9e773909b2332db", + "3c406ac84c68db46767e393ebb47d61459f837e518e07f30dbc7843c09854697", + "ef97c8b3aec6e35ee9e712d75e0021e51c2397fadfebe707f4ca1145e5bfd1b9", + "64fed80b2da32ccd08d0eda28a4a06127c5c642a27470c8dba7be63ee573ed33", + "d0e9ba139e88c3fa037da4a7032e2cd2ad9e97b6a7e04591aa5fa378c458f634", + "0f57350393499cd61d1149b395ddc67e9e0e7bf4cd1cb97765c26efffec7e8db", + "66c3bfaef000ad461b476ae17ce059795e9a394e42c681211d059fba80fdf161", + "7a48d809257de9c24bf97450c5e0b5161a00c39ca6ec01374108df238e91ba9f", + "ef5d92d33dce5f4bc38c8eb6444bc0de96a9725b7b52efdb39060b1f502e721e", + "8171cde1e4d42e4e7f2b1e8e0e0640c7e44781732f8852de6dca729c4a5a7447", + "1af061217381331a71680595e898b222009cbea8e4a4176efbf33b3cbe229c42", + "e239fa0014bc3b770a9d56bfa65607ce2b5bf4bc9683021f2b77610c6bafa500", + "2dd0f8f31371c05a4fa22056928e0297d839fdc187d005b2066eaed6cc036b90", + "cf2607d315e58b856fc02c5f7023a2e789b3fcc17938cd3b7ba7ff5bef77ac9e", + "460ebfa05a1b85e0afee310654883d2b5bdeea045d8f49b51c94b020a205c17d", + "36d5d6b512a48cb6861ee89ed6ca17322e31b0e7b660e484081525f287728c41", + "f64ef9c33cf16e5050de83a9811a8b77fa9daa54f678083cb44f22867b20da0c", + "31559089a55527a08c32e19c5b811ccab0822a5fa17ab128cfaa45805d6f8c86", + "160f1b31ce1cf10e5a711076853929159b4ecba1552d2e171c9ffa21a6d62099", + "94b5afacc3c429f9775ef6910912602cc07e4338a0f3a0098db02bf87fc9f43a", + "3b057fbbaaf319ad4a03b3c443db293d2ccdf3d7dce331d9bb89189dffc93044", + "9914b2fb2c3c40ff9c5ace482d99a186549c0605e918ca47c2e8968dbc6ce64a", + "80ac6948646738c7d905c376396e05801f22ec008a01e829b5e53ff5996052e0", + "5f4ad714b40296af36143b2a8c47b2bcb5bb1e10a5af7b841844f27dbfed22b6", + "bd472094a95ea6eb71679de4851185708aec21adec64c4209a1949738acd8f7d", + "a2afd267f5ec45153806c2f80c6097b2d09ee4f9dc16f617264832729201479d", + "fdf19cd9b066c634722e1023b0e8df814c3edb71c8d6afc4e98f05d19707d92e", + "6c927a204bbd07db8486416799d9fcd4547b537c4f887007a76b29514e071129", + "4e512e554081120a847c1c3aa7a90172aef3838137b3e5d504d6139904ccbc93", + "171f268e10982f79fd02c3300b1edde19ef199b37f4448e714d0580c3390c1e4", + "cefbfcd49c77728362445f72be24991b60d172b52353cf1f95ad3f9cd18bd010", + "452e7e02ed478bab0048e9dccde232cedf0a7b6ad532b88021f072d43fa08ed3", + "605d19a96c5e06ac5f83edc01259636d3a2c052774f652b780012dc8e19c9c47", + "a2026f06546791f072c50e30685a75b27f9e9cbd86983a22375fd5b65e740553", + "34162a412be1624ee18a8c060b33f6270120390d1845bc0ff047e90ef237d4fe", + "5e68df099637a88180e86a75aec4cb66a08f00c807c0d1b133627eafae3be4d8", + "08d99ed181770b4cd5ac1848565237bfbd41697cc4ea281cdbe55c4af3d6f23b", + "537bcd0e5e14cdd1bbd9a069ffc81af440d3ef71721e76937de01e019f4a3a9a", + "50395add7f8cbddae7579c7eb488aa6514e45bf5b1cfe018f2994f2df8e7ee0f", + "522416a57e4ca041ad116ae85504f71751b120f53b6c04576b5ff995317f0328", + "6f5619de3a61dd62a77e04ae2506b5cdb883a1a214c9834163c8df19a2bd7202", + "f94e1cb972511e6bfc3833c56179c7d4a7926ab1f7ef8d752fbc8f228f611ae8", + "39d1b55d940b95165671d53aefe3ee51ad618bc3868e181aa2cf8f6712b990f0", + "f73f6998468713ead95afe7d0e6d9d5578717c71695c6b0faa683efe1d940c11", + "141dbb05ebf8bba7476374676c8bf7afe903fd92158a3bb19563666f2c4693e1", + "59fcb30c94859bec90415da3f56481f89883f2ab593fdbb49846881aa372de18", + "d8e85ad51e9009e83fa719067ddf5b9eebec11a155fe3c194810f2e98c8cd6e8", + "da268efd3b15f198b9ad05158fd288422c4c9835f8a71a479ca3e8419ebf62a4", + "2c6cbb3bf3b58c641be2741eba400852b3bad444a54e0b5cdf08737dab1e2f33", + "470a17bc38cfd8b6c1c9e6ff99b7052b531062f351780423431eab8021d5870b", + "433ed6154a9f5fda63c29437b11c394a6d7bdca9fde4805f552d764a4f87dc07", + "076d2a0bc19c06e8d6108980222c30e5e1b39b85fa63272eadd71d2322d76e94", + "d4f650598bfb8ffb01b600a343e8131f10140b4acd4209635e73429b5e700618", "ba28704b65619a3b706ca8216b6d4d3f6bcc4db3e8ad94388a8917e3c8b4dbc6", - "e50864c736823a3ad129cbd3d49a910e2325c3eb1244288beab72f061f2a0c83", - "ecc1c968bf2b0a43b64ea133a763a9db4ad4fff647e890a8932d8e17a6bb2774", - "f52d0bef08a19736c83aba0bae8679867d860a5161de3e3b4684e1e8698f7920", - "471a7aa80a88c3ab714e41e51a59275a236f2e8cb96f71743307db6e77174a0c", - "27c3b632fb5a97d1ad19e820c72f574e39c9302a2576b23965c8b13ddec212e8", - "00b9163319af184bab4c9de88695b44498af12c8a400cadeac6c115c8cebdd63", - "154c86e1d1a6ca39a08994ba4f927a18c34cbd88390a251893468970991fc788", - "665d2d8fbc57173098a368787972734adfbe322ddc0c863306c92135d7dfe8f2", - "bff640c4c2b8c334b7b847e2f240510508cad4efa7b4a6ee27bd123618ea4c34", - "e8bd7b2cb99813e5c57bb6b08696db71a01092fa100ffec609b83db535f0def6", - "ac1ab1b9cc08f6cc27f1c8a202f92effa9e330cb15848684d54af697d9a39a74", - "97ca76dc6b46a2fcb1863974d47e99216c9f66af44b5962e92fe34ffd0a370cd", - "bc6d2467200af8d3c98e88fc2d42efbbf4ead3cc1fcbf973ef28b111cc9591a6", - "58b6ff871ce9c78f17cb7198182f3b75cd00f2b3e7c4b6aebf3b881638e8e0b7", - "10800696395a2ed2acb3d70e1dc2323ec4449f1654c2c465a6a242773e387325", - "84daea976640a1d3cc62dc708aa9d33afc684e68955952c810c2d8ac87037cd1", - "b75c5dafab5954e912368ea2948da8706501624134565a6bbd8cf9a01f6b7920", - "43669141d0f139e9176152bf54e73f947c553423a234c272ac4a98f0c63113d5", - "b31169d49159ffa1e07e824c92e753cd080881cca1dde1da4b1b01e30114daa9", - "b5126d508db599a49320e5de5046ae4b1c98f520a3b755a095d733ee8c56c969", - "2eb4e5ee892e15e659e6eeca9f91ede795ff14a41535bbb66c334f29425e0a34", + "1ce22da6f0fc356471baf75da17467f09b0a66e45464c92cf057110564e2d573", + "3dbc9a96ba131b0086d7653100f13ff39cb1b11f4f6c6799f70991a2fbce0c5b", + "6b8219210c12ce210e40e92505a6ef0a6b870204cf8e43d2c0a3c3dc50fc4d45", + "1701a02d6b58de458ac4ec1c1c6d2865ac64fe1c1f62e4cdeb38f9198912bd80", + "1b0dd831ecb7354f0d69ecc4f123a3de68a2e27134d9f8d9a8915f89378f696f", + "2c4aba41674b7cb9d116cd2215c66b705fec94343ef5184929dcf0203df3e43a", + "f24966797f36a0f507622e3f44c93325248867ef2d840e255be000a4a79ba24c", + "38c7fc5d1eed0cc3ccc7d92306d8319c5c316fb073e16c4d9c8ebce84ea89452", + "a24812caff45e02718c9bd97793f5ec0e0c79951a3acfba39af5285ce9b6aa5c", + "2920aa9ba6785635f88e1af0fd9f022100a610115cdf14e243e61b15a47eb5a8", + "2f4917c70c3de04280ff21a9d052c6234faa0a962605d450f591a949722ef37e", + "811708346a7e20529e5a7676de29340102b32f46ce989b81e3e510ed4e81ce10", + "bd9b91071aed1382c97ef285185fb2f06a61b0245422fa85182d0e4c583dfafd", + "985aa385d1dea699025b3ce93edd09e994179db489432bc3c81ec2f3474a78ed", + "143ad97b0976ad8029319d24464b5d4f9f102cf2717bd565437c74f51071d3f2", + "c0925d4c4d6c5bc7e089d4d81ced0f05e42a1ee6b3a761ed5f7a09d30f4a9c16", + "0ca8f1775f5f4b8d6d49bb851846c603e558cb00444bde84705a7d587986e9ec", "3c33a2531497783f59444977923342b0bcde1a5b0e71c41cae2577a7975d704a", - "558e7ec6166ba3de5ad4c052e2fb3dd2871e8b494c8810aa4224dd86dcb795a9", - "1e3e9bbd462fd13db31c0b715d3b5d42f897aaaabbc9ce83e9bf19f89ea96738", - "57534e8d960ed593d9428141deb135405c2bf3b78797e5cb9c50521a5faf8662", - "869e0f1a3d7ac86e8bd34b63c2f4a6b23bf6e987c9f8d7f9e345ffbc6619076f", - "6fc7ffeed5a3b6affd0e510098f302ba2714a7ab5a8214d6c73016b14bc2874c", + "76aed0b5cfdff3e715b9f99587223e5b5645aaaa328b3d59169ad1cacbc0be41", + "2e323b51f60924de3d4e9a269535b274d7c3913ad43ac6c281735e29378fb380", + "0aae4109d5125577e83739f21e31c553ad2d29c5f532bd5a766600d7fee92891", + "765477ac7ba370939b22a843ea6c5d53c6aadfdc197acbebdc313d083dc98f98", + "883a6439ac1926a86e59304959d779910c05104cec7b31e9789397fb27d564a2", + "4fa446ae59bec8b314d983fb6136add7d4137dc3c9cd06a8da3e901a5d13d938", + "bc582e292ae61ba2faa31ac0cb726caafa23c5e110ef2d25199197184cf29e49", + "7f9d8418bfb1c9da84f3891262a63be339d7daefde41e67bb75200c351a8a85c", + "7f56a79c3824cf8a9c42c74c20bf9ae9afe86bf39cf1d63c6f867d57a805b3ae", + "f44d0b649f1da5446d7177e1f8c4236fd07a651c1f3720e2b8202e63aa6b20d8", + "0c7bde34e8e13fe24d2868bb970a5d32acfc958f77a2c08302bb70c7733dcce1", + "f7fe49fe5d91ccb7b6dd9661a42b0e28e21ddd5488d79effe015853fa31791d7", + "9b7701fd448d20451adfd5e0153bd2194da8ddc38e2a82b52c0787311be5f88c", + "0609f0ef1914ee138bbab061bfc111a66b636f7634bc6b3953be340b84934708", + "debb06fc0b2c65678cb37c8db341aa554b3b800a03a002e8300997169212f89b", + "2bb962fe6113c7710edb7bd18ae21236c293a9a8c58d35acef61fab99923774f", + "5dfbaeb2fee7a98fa3c3fef64c2e53119030d41836b694b2b422fb6a9f9cbe84", + "0a7657de453d53eaf91ef4e13c58523a9503e7046bb626a69abf4937c57314cd", + "5e56fda6567c32ffc28ad84e3b31191754af9321f3c3de248d8f677f07e6d79b", + "3c8474da48ed14810198a82c7ffd47c53bd203f5ecb650785b6a75bb77ae4a50", + "3949e6395c10f4ebb907b17904036a5eb84bb1d62074654231b38402be163875", + "cde265c702f9da5cf086a57d3f08aaeb99d3bb4b1272ce915df8881cb705cb0b", + "86087175c1cac386cb0dec67a01612e186cbcbbc710d76698660d042239bfa19", + "4e815ed71a283efb172aada5961c440588d9f5063001cc3ed03fa2e9ac77615b", + "b94359d90b4479e5e0898440b2d61572082646ad450777fcb10e0bd730a66358", + "8d0d14ad13eef4a99a45518e8a33a8e634c60c1a489c61945a564d9e6186f27c", + "289acff231d06790801d10ddb2c6c2166898b761d17a0af893cc95c65dc634b7", + "77e8b5c4900e4186ebd07ab0e329d5b4f85bb7d4b995b82deb8f894c9961eb3b", + "d2c60ddd81aecb03dc2fdc525b633ad8f028aaf6c90fb11219c2deccb08c0f85", + "895e2c3a76dd56815b717974a4b1ca0cf4b64731a78ffa12ce35d624c7449b0a", + "49373f6ada17ae48ddb16b3c9c56b994a99a50a0fa1d877b87c5e255fd0baefa", + "a15c3d0e61b260ee4e088a680812288b044698148df1359a1c0c595cdd26526b", + "f647c7260457770201c73cf9f08928c9606264940d9940ef2d93652dd4329478", + "3c4e83c5897e4f461d5fbf862047c960a9c52243548459bd507332751038cb5b", + "d6cd2bc022cc04d4db1d440ba5a3a4ce90abe69783e18bf7a2ed20b0e60e9d38", + "d40b4f93454caa1281deeb70394c7fd368448e27b2e7f61b4390643c93745f84", + "2dfefc93548e85a73d6ebe4b6a5dd66732c246f3750e95b33746c83536e4587e", + "e404a1dbfac8eee7db7276fd265eaa7256b3ba886e8198a2b46fab3a234663b8", + "a065dd412c8256d8ad73342ff610b5777169d93b9c61cba32fc8e2bb4b110346", + "2c0fa378d049f6e1a46108a4d6b8e3837a2e0dd82369fa6e5c0cdbbd332d775f", + "abe75b3ffbbe9c5212da90a9d667fd6ed8df9e956faac0482e2120bf28fe9a18", + "39fd416d7110d5dec887b7f8d501c1977a99361d9457c0b0d07224d79ed67df0", + "3047f8bad3da8c7f68791a0819786f166277e5149182b1205c80c2a4c0b897b5", + "54fbfc50f3716048ebd2af901756f2b80f43c8e65e770ea606cc2150b43a2a8d", + "2b6fd0a6a2232372bd3dfc38066ade9169ca58c575002c461a6263f0d5e2f0e5", + "9c4f9016dd88d842682a24109d96c521123432790230b1485d481d4b92cee096", + "e0671c2ddfd96f3d9c7a6350ba7d910c92ead0213ccd7b8bb0613846022ce271", + "7d849cb1fee19e70c20dc42f9c222803f63d02d272e26ace65495edd487f18c0", + "1c9841e711f8fb661529c71692a0b784db4a82d9dad6525ccd6ad1e230947c84", + "b730d3d317a246451e4a7d2d1e59afb5e9996b972619c5a24f7b78efe19b8b2b", + "cc7a74d7667642e85ecbeb72e88204e8eb34c00c52a0770a6923b757975538f9", + "ae6b57296037e98943107617cbb572d31c8d8d5c06b0117aad7ab81de5af781c", + "99c44d1b7c5dc70ac9c6c238048b11b35a4beee65f9490d9edac3e69d858a270", + "0a7ada0cb70f5225e7a497a83300a1b63ed028a3e24532dc43385897139daaa4", + "a02469ca021d05e1c2618bcd244f50d6cfbff940e576567fd545af2943d041ba", + "841d6fdc0d4695c7061c23516e272fab40746def305ff72fac7006666d43f0b8", + "bfc47380877fd8324a435d859330b2762eba165727d16cd86ac7c88e7e601d6b", + "5bfcfbce6e4e6ece2cc8acc4418ce5efe39fb6355a32ac00fcfb745c6fb32c81", + "2f44e645976aa7b9cb2047e99242e8e9de2b1dde989e1c90d875d11fda4b1d5d", + "e6a0eb169df990f91e695797a7d11e10a09415d2feaed02169d192c9d3d33000", + "9bd24f31f4e39aacb59d266fca3b82972d866d0123a68b108dca9ba76ea8b7b0", + "6b562f260c66685a8a4f0e772534acafed71cdcd3896c3ae00e2c89a1de44a39", + "4351cf4a09aea5fa1db0c4222651fdd7ffd6eadece755e6ebb95df4e002f8fc9", + "b6309a78f92e2addb12a38e1d55c094424ede54dc783a74e33abb55146559e7d", + "f31479f6de4dd3051b3b9f5974461c5f7d22d068e54a3e7306d6872305abd075", + "c3b43da197e879d672d29f73dce7af20ed42bc0eb362babd0718cbc29afbcba1", + "5ddddfaa99abaf8d08efd28d6efcf7e33c54fcec6350121b16e9a9c11b774737", + "3f7b329f6bbbb3a0e8acf29cbc760226324bff3b7ba04461d6b70abb8afc7859", + "fcb12276e0b896cfce624abacd3ef907461b26d58cf39729bf29192ee2a8e881", + "864b9bf0de11b20df6904ff958ee0e46ae8e4596774e9ab22f60d4c69e1e380b", + "683f27d992d9a61bcdcc5f5aba69fa22a347c1639b6bf813ffb5e27731475446", + "e17bc8f65e7e8812645d7bb8167407389e49dff928d4620f6348087c2835f7ab", + "7977e2123aa5d6f1cbad46d79351e1f7e1957dd9b530814d30fd82949a54f4cc", + "17e677a402295baf52295e5951779f0e35b8798645a2d0a90c78f6cee5df1543", + "90e2ad59a736d6d31968a987fdb90683f11bc14b350ec6149a3708fef0b32975", + "420ab4588446d418b78af045f6da21153136bbd3a7e8d62096a49dafd46d787c", + "9dad54cdc27ffd06093579882522eeb83b57eeb6d424ea6519c741b3bfa3fde3", + "e8d6c1d9bcdb741f86cfe80ed9f7466c7fb752cd94e1879104bc568c1bcf370c", + "1544fba9a0221c66ebcb1208d7ab2bb23a269b365b3b09b080c0aece4fabf280", + "8670d9c9b9bf77820c9df3ea78c9ab0999104316a078172fe9b94ca6461610db", + "0cfb4a321961720be24a19836ec39028db63071b1b20f452688393de17f5d6a6", + "ce60f53830213acd71b9282155f4dbc03b6305498da9a701aa911a0d1fcc4856", "0522ff635c8979c867fd5f18b04cce7cf102bf6c2fb9a027d0f6b56e8af46ed9", - "5334e93710a92956560aca0d67ebfd8cb333460a4cede9a1119e92257c600d1f", - "3c6e5958d9646c0153b87024e77a63e2d08f2048da38421f5c716709d9de22c1", - "42a5ea15833276714eb3e6ee398b857b57883812b7bcc602d8e5b1df27f98990", - "9bc99fbe9f43c2e25ce7946b14dc1f7f463f3ace56726e03385b34e6ba1739c7", - "9f7600277d0324fad0ee967e92699485c13c22334f8fe4131517f7f1282dfb3c", - "2dfdc1307806fea7fcabaac061a6a107fcb21c97f997aff4caa8ee9389d7b840", - "3f9087dd241945afebef97430633924f98bdd5d84cc3f4a10585049d0f4e3e9e", - "b4ba75f722cc951314150de348a3456329fb29ed4b46ac753f6332e2b4d732f6", - "4245d700f81a1b406c0d702817f36b18ac7ed2a9b6a0715fd7320f51f8599531", - "d4efc188f1625389997de9c51e685a4c693984565dda684f22b0a29ae9bf8c93", + "d943e39a2341d53ff50e7da0ea302fff536f86516b83c188ab059953321e58d7", + "d2a78b0e18a728d846ec108563d52e6497617a4da86da84a9d114f84a3a4e0a6", + "09d4d9dfe5ebd5c4db03ff623ecc64e5820a6466a7998c8d5154a0d746aa686e", + "b2a13e1fb792d1307e4e7d7cd6b0c2edb34be98bde72fd988a9c2367a0be71c6", + "6d3200631d0960c0ca633c2f4d251c42a8cf7aceb4ba6d0dc573ca9f79099f85", + "aef8c79063db1d8e8a3e036fc2c9d533c3b08c70bb19e3804c1d69f0cc6fe667", + "403aabfcef757c3cd4f9c751ac72f13c909daf0c9bd4d2e86659055f69befaf4", + "d5e4f83d7db68b2a1c27bdb36acf5fbeeecfdb5987f63ae5b826f6385bf3be8c", + "dd513918183e0d3e381e4251a01543cdbace4fa950395a85b230fcdbb00d548b", + "6f98e4cb0c198372af51d07d6117f8af063522b237085d841e601511ddd62a48", + "0b86a4aa5d73226c443b43fa2a500eb6c05fbecfdcf398d9383196b85b96c71d", + "59e828b83f93ea22f120c2fd0169f712e1f76b8e590fcde286240bf9c76b74ef", + "0a101da1955724e24b463ff02f8e079fed699e736f2ff4258a468ddd5e6a7dfb", + "bf474d216295d7a0e0aaa0566aedee17830eed06a07b80ff6bfcdce9db4cd966", + "6d1627561944fbff5edb853c3c1382ff208a9b6c50668dadca496faf991eac80", + "2e397708a07d2bbc60f9a8e759b516371a3128a4dcdf534b1927ddc53ddbda1d", + "b460bb8adf4ee0269355e084bdb8a6e3ec9c1134d8c811ed5765f329666f7323", + "3cf88037c74a1411e046d27da9fed419e7d9e4037f59f1c22fa8319129e3b1c5", + "d22175ded168138d822c21b38b4da1a5f4f1f1d73329944eba13da49b468dfe0", + "a4e0ba2297ef95913bf905dd08b158be80b41bd093b3ffbbec6c5843dab3230e", + "34151dfc3028887b215cb3c605ebd8fe9b5c0e3a5958aa2a4f4d42af5c424c46", + "51a163d20ad2348ff0e999a874f14feece723279f1c9e70e5e43d0a352648a80", + "f953c7e80de755684aa71e95e746b8cb560f814487f786ecff687135e49fa445", + "a3152c36617db91dc1cba4961e24c2aaff7737861c7d5f8b0515724c7fcc26d6", + "67f51aadf2bdc0a8c7d86038bd31b630844747c51a6f67658f038d6cff8009f6", + "ff683b0ff83d356451df2083c4471b0430022b9ff287f3283b62994863e49c48", + "7f971de9506f14731026520dd286eed1feb08640d2a07e1ec372567d7cf5957c", + "227b48ad41f251b74d64078c34bbc4548c42e8c0bd91469471d2adef1923e72e", + "d73590217d420d55a18f10f4349f4240aa192aeda4957ba2ed7a801f7a0da9a7", + "f3142a84bb03641f0cbeae85caf0db80425e6f91cbb9c3eecf2fc344c9d1a1c5", + "22fe6ca47aa572af04d51c9b863d107126a00853f3d88b79b8edb6e2c8e693a0", + "b7cfd158924083c4a96437b829659f00dcf01dc1fdb0ac34e55b898c88ce1d85", + "3c547f2c8ed864709ec3721ec9cdb613c9cc15b1c305e1a4c5fdf5b7ea263eaf", + "d5b3b450c5dd1fdcf2f71f81d472fc97bcf294ed056230260a29ba4810f54267", + "9f447ccad53a3ed9b804d7e2b25a3c6f6d02b15f27555297694ec09bb5036642", + "cc1039be449817530d86188c46adf4af733c54d8ceee07a9ee1289fe628587c1", + "29bc24f2032263b73ec3a9ca7fc9490c11f177f443d65b9680202f446942131f", + "a4fe167f6caaa74842449fbcf59a7f58007280c80500e87597d80003a805cede", + "876ded4c80860a64d7c325b9a6fda5b3488acc8d4fc2cba3a1f4ee37aa25300c", + "41b503a13906700668f775e39cc3407b72c96b3e631c9862db8c7ffad640fa25", + "11b5ec117f2802009d39d6fc085cfce9cb832ba52cd7c722f4ba3ec91f2bbe86", + "8853adef982a77838cd2efc203f9d9228f0d5b6dbadfdb2ea5dc117116ef8048", + "17e2a3b566f77fc01664ec2d517c11f8992455ebda82de719087b2b926daa4e6", + "46913cf17aca2de33f2fd0988ad5328ec18f7b69e89e978a3aac43b0d9361e14", + "2a3ac96e05a890e7d490435ee548f8f165425051f81cb4ed97bda6fab2a60900", + "b450dccaff4ac76295f1d7e7bf1a25aa672c30ef72001500f343b053d525da8c", + "e7f0bb9566bdbe7719b821d72a7892770f29d93ee7280f247d627dd193fbdb74", + "7e4c743a9282dc57cdc6358a082ca6cb1ba19ab6e049ab8c4ec9ca8b146ce09a", + "910b4ce6df3c7a2fbb21b0ce2f858d91a7fca539c8f02b2f737e98952468f681", + "240e0dc1f2f462e04f60707362ab4561c5544c9c7d668be19f530aa6e61c5780", + "05316b97794bae84c49e6302f27417cc00381ca4b8b2c5eeee63116477e6d418", + "58b32639f3b1095d417d64b879be7873672c60137451e26412314d39fbd622aa", + "d52fb87e1ec6c2167f2615f6f286f070be6264d12301d76ed06a55e67d33cc21", + "5b02154b74ad955c6c76eabbabc8c8ffab621cff346f4d510d8c06d684f1db6c", + "36e4c80f3d4485e7b309d5e2b777b5c9b06a2aca1cf3aa209ee714d6a2b6132b", + "e43b186612e605ca89f72613905a5a2fe3cc197c74ed8804d36416e4cc0c840d", + "c422068f23a91fe0163de122403ebe0de04ae106baa09be347694b23c306ca23", + "e8182f6531e650072095fed068a0da58b36446f9cc9c9f4a126037b7a4ae519b", + "b8bfeaa751993a74d971d2e11e84587100f0e12606f1924b3732cb9a627e3415", + "70e13de08d15e40fdba4d6cd7abc5848c6ea1cfa119d1de7d29ab2d94aca7ea0", + "0914cf3513684c056dfe38c810c1b977d2274b693819585e379b8101593abc5b", + "8c768cc31637de25dc58238e825bb776d31579d4759afc0ecce073beaaa40159", + "bb99bee9b64af3e956efd954c596138ef143dd24a5b304fb68af140b1e118673", + "cfeb56a010d1e86441034d42505b1156ad38aaf2d440001b2e248bbb8c406960", + "ee80ee56f7e81e957a3991b4bd52813fd7ff92a55611ba97793889e4c0d1932c", + "1472cae85e4123cc12c48bbd279287812528ac443ff1612991217320c3ac1df8", + "1b8e575a242a69f80a0b2442480d8579aecdb07999527302199770114271bd17", + "5a86dd960ec90b8eed72eca4768614ce00db47b023f62edc03d4e9753958ea02", + "2f0f3de4b5d9bc7fd99a46b01cac408dfe42416ea50cc4af0bb8ce3db606e116", + "177940af4c80edd9d240a3650e898553c909f08870b591db3a1ca3745b0c17a2", + "3e7b57887303f21b1815a03dc8a0983f5440744454fb5c4be77b442237ed6abb", + "84e751f158b77e5b1432728ba0e9346aaf417edb6509a61db27cf70b85b1a571", + "476d211ab315d7bc9b9ed4272a4d899c1856866105c7c8cf3c4d4aeb43da6491", + "ee80d9650b00301d451e3277c5e9187cc227bde4de69524aa22a1e8ab30ff82a", + "328dae3f53cdfc82061f9e8ec9387787faa307d057691a2929599bd514d6dcc7", "9a15e0bc3978ca4cd52a8fecbaaefef5ce10a51c9c5aba4ba49a7fa6bf56acb6", - "4b0facab98b030c856d39e4393eb7da8f43d8d7b40d5a49bf7edc611211ae317", - "be6c9c9463988179d4031488e830d61111ab00d7862e354df1ac6d4020a65daa", - "98f086d00082455f2aee8c399858210b6fee5bd282cb27015cccd21dbcd89a5c", - "b2723a0e07d80757846a8e76413b8dd43b62d88797f247fac8552e2b9ee2dcb0", - "285cd964d0145692d1ec4f5bec22077fb7a532b2d3bab4875b3c58190bf69b1e", - "0c9df54d3eaf207cfa508c496b02851d47a12f859b9500ef3833266cc9324f4a", - "4649f1f79414be60f3fc437df1fd67dd3bc355a08a9e462761f8d29a63738ca5", - "40aecfcc85b5179ac7d390b5ba76ab01757ce9cbaf70639f60262441aaeed75c", - "6261d7b53e178b20150a71c3378fda95808a2e673441e96151f9fc5a41a06b92", - "8cfab564572292bd524fe0b68519ec89be8b2fb658397596cab875ca04e9a3f4", - "5cd105f7c8d29ecc06c61e605cf214878eee6358f1f6bd2404e77ec1f077423b", - "46fdee6d4a543ab65000f0fe7bf223324c6011732b038f18c0bc96546f04db88", - "dfc771d6cc9194a8adad2fa493b23481d359e55efc7a3d0e6381de76fbcc0551", - "3d5093a661a6b7435182aa766dc1d488b5ef6b7756c7bb566b168361f810fbb6", - "b98c2c16d04680916406cfadd8fb63d67ef620f728e2e7488d58906025c24e19", - "d062089d90f337968fdace7e0a374218468824b5110ed379785181696a09b4df", - "047c3e6e224e592cebf9a54407b6d3c84b9ba7fb0b0c801f3742d65bfbc6dc83", - "09db6e42619c27988c97f877ad989096d1f0faafa5aa2a663a0e41d60b16b4d4", - "7ca6e855934dfc04023316e5b8acdc4b8a575e2c47d660ee32d9aef0e13749ea", - "1ab1d2279d73f26d8c2c2d29d0a4d964f0c662ae49fe4888ea9f08ceb99946e9", + "e7b1e9461b5ae31fa28b9a165db1893b95b01d0d14fe86f114889f4b6053d6fa", + "c839942a1ab82bd684d5cc3974e99e32703a495101452a44e494608e09b815be", + "a2b6902890b012b3b29ba29e3361325f88efd8d908f8b77167515f38b583bb7e", + "947d594617273e679458cccc554d0fc644f4f0e27818000d4a647877e210d764", + "ef2d039f4a10fea1fae019842efe40f559c3afd486898949357bc87b168e9890", + "f4aa103f81b50141e761d887b340dc2401c7625afc989478f5f722afa14341ce", + "39bf96522584662672e97ced3dfdb88d5ebe0c530ab5805aa3c4250a847b0dd2", + "1942ef2491dc05ff2f51528954b3c0c6551b768b47fdaec63540f6b6e0212ce1", + "e1132af43e595ae67091c0a2453fba89a219b6a5f251e3d897d4fb00f0a54656", + "af7805096f04f78039fa23bda8f3ed966c1543baca115a72a03ec5aec233d80e", + "9037043add6f24b50972795b56c27f0c3a4774e917ff137a7cfa6ed15e28165c", + "28580470c9dd4b9ffc7a1f930d7592100d4ecdbc1bf8a5e3f68639996f8319bb", + "71742b81845e9293f783947938306a7ee42653e6d931c6c8e4c1f0f65e59ae59", + "866ddcad02520a06a69d269ffa7cbebfdb8d26cad300917eb85ad52e78e9f79f", + "b5485687277c47dd7960bc3d9ec7486cb279a50498dc857b025897ef5ba75b69", + "7a243286699901d82987a16817cdd07f6c39cdbe9f840e16fe8b571309d74960", + "48ef1e6d705f737d07fbbf0921533e4146a1f4c8689cd339092e544990805d3c", + "2e716fd0fc2bc249f8f40c24d0b92501fe3d61d07d39a52845d5bba6145b71a8", + "32684158fb89cfebb5892d8fe6788fbf5ccc8cc27b502b8910e31a826848f67c", + "ff56db96baf76fa5e4258c24198fbde4212ed8cd3fde183daa3d3f43a7107948", + "c434a192f776ff53d854b1655e96f621596418d0f4ed5264fb495c32740b44a3", + "57411f42771d2cdebbcf5127a6b062d987b06ee5da2f9d03d7cc1e9f7b8728e8", + "79d4e5c0163ac93a53fc65069febc7f68da89781547ced498cff730dd4b7fb41", + "f27e5ddc23b8bd057a6774d0bdb7df52052efffb449b5aad89f8b445c314a9de", + "10a8e1c77fcaa55ae12f0af27e470298d50216317aba34d9bcc68b2ab486f1b9", + "abbd47565871f8e73c6a8ed3fc4e698533d9dd332b7f0a7a1a87e24dae262154", + "36486efaa1b8b34e4120a26d72bcfa30b544307904ab26d19bfbe028426c9092", + "394e5e62ffc9485b9a03c062f118553cbe987d73753a0733574c6e256443a66b", + "6b9b7c5f68ecacc4698aee67832436ad2fbf0933f5eefc16f3c1f58bd8f57d62", + "57a9fbcd275a825eb06c86ac8f1da874692fdc29893857a0bd7519e1bfd80ea4", + "3a19c43cc3504564662a28e8a9fa56cc8ff3e872405a636b046669a1cfd57727", + "004b3a1abd0d4cb0df215792512528233c204e31b5f1713350370df9260c2a3f", + "0cbae6fe68b4a450aecc524c14ff3d8b81765d98826752aefe07950731a8aaed", + "cc773a60db32fd4f0e7bc3896830bf0c4a9671a9565b2d143e8940cd210af9ec", + "e9973e136170c8e9f440caf5f4564d45400e30b4876f56ec667137e998983b43", + "1006cda7856019d34d4d2de586198c26cc544f8eac1abfe5f8d128d0f1808801", + "8faa885c67c4652bc808923f6b417a3e9c96b6dbb4c26ff8e14c35655d5fe923", + "7298df3e09cda63c352125499c50002a7a265f33f5ecbd20469544eccfb3c508", + "2565095bdd107fe67cf81ee78ffdbbc0da8a2b41cd2a398ae878f8f5d49d0234", + "b30bcf207c65996254b001e5fa96509898d24379d74a464516e0e4d4831a3063", + "2f12f01ec373e4e36b8efab54d0111cc6de2927976b8422967461fcc194fc73d", + "4f88176bb6b25b6e943cd49b340d90f0ffb4baa59f5f68d6a3183db6e5257100", + "78be0f78afb3c8516567f574bd7f9affeda5313c077d95373bbf7bbf125b8e5a", + "87185fe6d445f8f1e319f8f323ce022be6da50f6d2cc5609f9f1ad89886917e4", + "6834317f91f1af08752ceebdebc0b4c8b85cabec5ffd65627fc1a5762cddcead", + "a61c4710c3aa9aa8a2141e4aacfd8aecefd61d306c927babb2db7e7626a08a08", + "7faab4bd141c984ac8261b19031bd2e9f53947e7ca6c9ab0fa5cc668d1b0df4c", + "cd72590db2b12afbabe89a48b61fcf4cf1e0d2c0c300c1123927121ae14928f8", + "921ed85a41353f1ab8d1a4742c28dbecae3f465c12117ee0d687de9edaa8dd34", + "2571071c207de5279aed628f30c70aa4205eb17f4640916930c6814ec03491de", + "0eaa8a64c789bd2899e0d97b21697df66da9eb839b758a6e415430213bc7ab2b", + "ff0faf3a6a636df0f3bc8cc5ca6083e4e3c9b865f5590f8af06491cab3b26a93", + "28ae08bab0760cca6146019d2073ae0f1fcfc5673b79cecc494cab5ab04e7415", + "212a33b9ab1660c70ef01bda2b799701bb72eabf3fa25607d0e6dc423c2357b7", + "eb62a2f858a58a25689db989208d80322b78738b2eadf4d12bb9ac8b49065dba", + "0bd846ec567d9ebfbd5cee3632c698e9ab965143b48b436a0a9ba68701974e54", + "b4f77faa8f1ddaea781595cffb56ad9758b9fc468cf9304f0a3111a0860605da", + "e7536be50a3bc3b1550f27619d3f4d6f396910729755bb24318c3197f2a9f2ce", + "960bec33076c60abd7a0ed98f9b8fba2241455674d7e09691e0b1f1ed73cb551", + "1bbe1e85191480c8535e5c7806b360fab7d5f9f25d9b0865a190d2afd447a6c5", + "f52d0bef08a19736c83aba0bae8679867d860a5161de3e3b4684e1e8698f7920", + "4846be9705c305a8accac085c2c6851a8149f57ce56f314e480f1cd255e3cfce", + "0a38719d97b8168dcb539b9b6fad20a7163965272565cf429aad69e983fa7eaf", + "0f115d4e641b1ea7f90ebfcb220a178d49c79d0e55943bfab475bf24a5118411", + "b65e00a34c5a64519d94ee200d062ecc9637e12664b2cdb537b79c907d7843d0", + "b99c0e0e175913bc27d3fccb15010f4676f8b33d81e86e93e7f877d3a8a135dd", + "cca409ba122cf5447dfdfc035d2ea09b9433bd99df5f0b65726b1f13181da4e5", + "30f0ae3ef12a1b6da4ae8bf365da69d18855615e51c0802e35737e987dbc39af", + "908ad6c271ee7b68fb2a236f7b2722dee908c5819e3553894676bd2f729ba0da", + "b837b6fac7961b76e751100afaf8a73797797b9ac168c6373125ec959271c9cb", + "8f4be31de1d05af6f3184a1390f62112c7b17dea2c78ae9c8837b38cfbf70d72", + "cfb5edede2ebc049293d355666aad072be36227283b3984e01c6fc001a438431", + "1fba8731962da72c77e546241dacffa9b94fccd05c4135a231a192fb2f0f9ff3", + "5f0bd4c62cda5cb13762bff5461c635cddd8026a16e96ebd03b6adcc563a63db", + "00bc6b9edcfc5d2dd38228fdb59f2783df348e34679fb30cca5eb71228037a05", + "ce67786fe9f7b4d3bb05dd8e895cc39624563eea52813d23112fcc281d792de7", + "a95272a0500ea92f26315971d84975318a20514e81bc1f4e6ef165dace91cac7", + "b751b90f3123f1ea47ae0bf57d06ed1887f430dbb4df42afc34ab1ec35dfb6a6", + "74103cac68724b3c97b130e869c05adc328377382da98164d6e85efda67b62a3", + "de353e0471a63f737a6902a89a7cc7a02e204ac7e1dae513307ff35ce1496c42", + "ea3b0ebc134a340120f87eb0da8a3c2138d1194c0e64d3cd21dadafe97004332", + "71f99a8c50b64383080acc8231599415d4b9bf3052cff0076cda62d08fa2d4aa", + "810db81e616e41b5fe1f347e6535114d12f898cebf1fb3f32874360c26ec3076", + "8c6e9b916f022f76271eb5f5982456bb2c368dc7a6c3ba6d0ef335147526a5c5", + "22daab4b38c0d2186fc466a987abf0189c20bdcce7cfbb1920fe7576d0b33102", + "a0d54977c888ceb8ab5cef504058492ec761f1aad6eda0392f27bdaded0f4d3b", + "f1f2e6f43e05758a1bf2a1d4d3a904b0e2f0312a57eea2a49aa1deaab31869df", + "0c0bbb0bd7d4c6f49b6efe2babd37f8fd41d24e5c2d1e83a04333ecb4fb3309a", + "aa5407e099426b7bd7811c7218d9ff555e0ac62c3e0500d09e15cc137acfd255", + "24434303ff9fcfad02908d00d7a9283ce366ad35669e0c288c81da25fce0dfa5", + "c88fd0d4dcffc50d77a9b98b04149ff9f912d3fc1328ae8a464049f988d1f299", + "671fcadfda6fca9bffb53b7f85c679f5ec48a9fd9fc415471de871cb5c25a733", + "8222b71a07bd3d4e97205f9be84b5d596117c5de2bbd5d9f435cb8bf52e97af9", + "97d0c15b22281cc34c54f2d96b6854ea49560f04cdde6ad4f8773ee7714f97de", + "84b44d3a2d90bae12c3d9246bfaa4fad0f397b6b4f2350fb865137524c4dce6f", + "79c855777bb71cf420492dda76b6dff0374ec0f1981bc01f3651d41ac3f2b683", + "1a2bc7c53af0968f0cd1d4168e10fcb298f7c43add50ba679bbae4d10308bbd1", + "49efb83a3bb4e0b07f865136644f54c30c49f7d735e4e9bb57574d1eb903c0cf", + "a63e43bee92430a1b6df8a07958642940459158f1c44d539f303399c8eebadbb", + "b6ab0871570e30642df6dd9b395b2b21369980de75894c9b3521b4778efd271e", "81910b58f73bfed119d1ab8c03e22abb80be80d85f16214ba998f6b8593b2179", - "33249e435c044a6b3660f95dc40f4d4bf33c3fd49a2d2b76fb847f8a11c4a3c9", - "c4d151d0901716bf10ff9a4f2b5ab00aab01f6a2d97e586a08d2663257da8a61", - "90091185817e3a7dc46523e28be2175d7fa320fd617bf3bf65ab77f62897ecc7", - "ffaf7f9ba2956848629ab840e97e093993125d89ff6ebbb6428d7f195be6f9cf", - "6e7264f8903d3544b236de167a11e023eac3830c8cd72c51b6c6f3bc6f71a9fc", - "6b948e2d252c32dcf5640134dd7cfc3f59ced5c1b9cddabc47e1582a6dd15ec2", - "4acf7ed9d33375d9b0fec69ed65f1ad4d154cc93cbf36e60af96ab04dfce0a3b", - "160f1b31ce1cf10e5a711076853929159b4ecba1552d2e171c9ffa21a6d62099", - "56c87eb0821fce471d4a3c9839c0085da788bd3d8a8bfa1cc5540bbd3b11a565", - "f2abc6720d02175ba01d916d6d9d00749a2effd929ab028ffc35f778a49f060e", - "71d97c35b126aec3b2c3cdf181bf87ec3ba9321b5aa9d910d48e91b491cc5ce9", + "c03853f6d6847f853eabcaecb52f823533f05fe413d1b9b58d77700665b2c678", + "080d930064e212f23ac2ae07908a0654f0ccae6d5e76a9f26147fb3350035630", + "fc526d03d31721f35824162e067de36d8f3f16dd96e67302ce86c8496a3fd6a5", + "96a57080416dd18ed122de00eb7cc53f35af876a673ad117a325f1a2bea87924", + "b85607ad5321b4c87530776a65d60d5c5c97f2c74b1d793fdc4f7fdbcf8773dc", + "c74f0e4dc031536000fd072fef6cd6643401ad4f6d618ff6c4a7600f2a94967f", + "fd38024b35651f5ce8a05704c22504c0315d300e005e9dc9de4996297d68e33c", + "14a8a7b4fcda8fbee6ace9bf9052784a2c1c88c62f37c33d11c42c7003072f99", + "62e04de62f85b1c39dd465c115015d506ccef38f7dcf2cd738946a7d614811e4", + "5bfd6361aa55ec505118976d82651ec242b70d3e3d092f1d2f93267b51a8c5b6", + "03fb8bba0ae03f8032e08d9afd2b560a8f1f284c49fb67930e425e559c7872ae", + "ed1852c2974de8094880749b38632999430f696b4644494fa5bfd69c3d86fd74", + "2eab2090651df33c6c816d5cc6b1ec07b87f124e74cc78cb161db459d2d60df1", + "917907e2762239ce222e080da1cc41d171bfbe1c1f261d805a9ba0d46df0ee2b", + "e0fe0fb12b7a357ef42b7e12c2e3df933bf6d2092fa2c46b71cf217d0ce2bbc2", + "206527ad00085332f56c34e94e5559f61166fc8ba8c99e01dfa6bb382ba303f5", + "24e2fc5b1670f0fcd14041f0eec7ba07af03ac11b50ec0ca64d57b4f33cb6a31", + "2450038f030e137276cbe005c4df2b7dc91afa71d68d56e9563b450ef032125f", + "bb5c4fb583ac6a26fc2d6a5bc8bb0f663c04e8bfe25bf863265bfc3ae63ce22f", + "d9324342669674c22187d6100cd0c47f703b9f0c743ad16fe5b7d37fa8bbe6ac", + "ad039efa08f3e519c0be6b7cd86ec1758fe1cda598685b1e0152afb438ee81ce", + "50709f5079b85ddf22abce0aa5a938f37930d44ec6f6db6cc05437e0353d5fb8", + "ecc361f6d788119f05b7a09199b393a11f754f42c0304380326ec4e39cd5cb16", + "261a8471bccf590bb0b8e9c71c25b04b48f05aa59cc8bf81835cb619e7263f81", + "e12e7abf11e1b22719765d33153b8733aed335b2688d6947f9a733484b18592a", + "67e6f1d2ca51959abc4d86b6f62eb5fbfea16c93cf7fb70e3bf3d0428cf0d0bf", + "edafdbaf39496d2e7f18ee46f85dcc2716a7897e30f76831aae7c415bec8af85", + "be1eb33a2cbb67e62876bbf9ddb253d6236f2b67a00de27d25153362a0de9874", + "6dde66854c3d02bb5f0706db15ba658931db1b0b5eee91d5526df04fdaa8a270", + "da09bf265fdfce40160424d49689e2c4c316c2dc40b16a9edaab74ca18cf3dad", + "216cca51337baf5c8418f47b168b1761093eaa805eafe8a99bbd20a5c5b4d3be", + "7fcf476026fabd282fc961111261d374917583cba3c95ae07ff8a448a461c275", + "81cb3a40b780b5c75e12971ceea05810c43e81899ace861175d5f4945c097c44", + "89607091690523cdfb021e1201f287592573b4357ac12368a16a8116cd53aac9", + "3975f736b469899f7d9008453befcafc04c3b6349f6f149216e8aa4b6a74c118", + "3759d8d7e2832c294a29451bf585aa13b5c2af67cd91e387f500a697a447baa5", + "8f62c96296f59440bb64dcc810a4804468b79130cc550fd8e7247b0facf7ff44", + "21bc87f602923cb0144179bf78e93116bd7dddbbb0c2d5a3990175e8e8da3d5e", "c7157b1895cda9b5df49a74974d48fd3976f647f4fa8437bdc5040b85e2435be", - "b175ab532b64a089dd2600f4f8d8708b4836d0eebadf6cbb3aaffd47e7f7c85d", - "7087dff68e5ae8c835cd85712b7eba37b58e2777bbcaf8a5a4f5b40dd2a0b703", - "444b5601daa54ce66083790665a7779f9f825d6a071ad0edfa6e2b58618dcef1", - "1ec2933f3cd2ce51066754e10cfbb89112e8ba2f4910629debd865674f26a313", - "5500dc3599a5c26a7f0aaf4c84efd0b860b4da8b4ab14d339353897e00a8591f", - "7c612a8361950c14a07acd8ba243c31e42441995302125d3db787dad5e6ad338", - "41f46517535caa40a7eb080d16da21baa81af82b6cff1fa86a3a0ac2ad6cd8a4", - "7f9d8418bfb1c9da84f3891262a63be339d7daefde41e67bb75200c351a8a85c", - "9a0c6b9fd844c5ac3399174a0f4ce4a7a9f7275a695c22104de744e8d2fa5599", - "24942c219bf4a81a01ddb9c5b98c5e848fd619433906bbcdec06f08361d85cea", - "82f567b5a6e9652eb1b314d1d2fa31b07926986feb9a580e31f39e1e73413878", - "3ab451e867845c43ecbe25e41ce46512680ad39af828b9afa8d10048d8dbb448", - "4b3ec03348174f143acd32c724bfe0b470819a336dc7b17feaa5f9e7120f4c9d", - "aaa0a4fae5c64db8077d2b796f8143f18a1dbb10353ca3ba963253cb73437b3b", - "d637eb2ac5605211159cc9a492348aecfa69c5273f57e02f2d329b9b42bedb03", - "8774827fe8f2f9787f330cc540ec463744e4b473dca742650024a2def27d4c3d", - "88010b05c626a1f460f0ecd833a23152d0352c42405cb9d4dc8e480c11a2d53d", - "efaf6e1eca0f45a5387ef9ad4f5fa03be3964d591d19cdcd5d5a8186f9077f2f", - "f1a30391c75c83ecf7573040640231354f04c79c31be41955650323e7b6bccc2", - "fb4b785a94907294387f776a3d3bad78d746097fec8026028356ce181d8fcb76", - "d30983c394100ccfc16516624bde55f78c40b6168266d8b27ffc98f34e4495b1", + "ce67dc5889ea3f32dd5135d6af797eca24e8db6658a330635e4642a0e4824846", + "4b0facab98b030c856d39e4393eb7da8f43d8d7b40d5a49bf7edc611211ae317", + "5eefd4cb1b52920c73be9851c197770e6fe0289623920ca6ecfd005d4ae27785", + "e2ced08fad6ceaa6b6a5b90e051544ceeaac711f861797d88d6987a6f03fc13c", + "d7719047feea91cd981fe925fc01ddf943997881a5b28bc0bcd132303f70b61f", + "d02e0e599461ad5d1190a8d83c9bfc9e0d9467cd4c87cd8ba5782bf2f23af91b", + "ffe4c788144f17ef2f52b14ff0258dfb8f50ded884bbc7cb2e206600d174449d", + "d71f4775bd73a2a6a83ea98d94d9a77e5cd3edb4e2afe49e7118e452b1b18c43", + "298022f28ccd7e342356fc950aa38596a10499d01cde4ac99d7fdf05a90b4bce", + "a6ec3c139ba368f9488d8979936b57543d79e14adcebbbbc7b1dee48f13770ad", + "5276a4bfe2239aeac01cba02ce24851cbc7ed6c3ad92a979f11c53bedfd23a04", + "6d6bd432d91eaf444ee8684407c055eb7b61c53f1099f6f6860ca84e7bc5a81c", + "af4d1ef5466927d71d93cf8ddc0938117c92f1e835f33792cf4b2f7b8f84e55c", + "55d41eadfb72c76452d8d91cb2e34afa74568209edbe3a884c5a4f3c52416a90", + "b9a11cc0f706fa907e42e113cdd944b0f4dfc09ea8a6a084aad7a15ca8b1daf7", + "948e765ca6c6f95f06892e168cec5e12bc0b7c07022622e4ccda0328db8ea640", + "944099188aca93efd567320865440e6ad97cb621480485e3f1069af0323e8fe8", + "b58ba3ccb53b7db24736ab03452dea6f6a5e3e8e9edc0eb6e92cf39b59de84c8", + "d13b8b4c841597bf456f8519c9dd9cc19fcff9fcc3b7050025608d4e0cacb19e", + "b8eceae671c4db3ba336c51492fe2c63048a10579731dbbed787d4ecc7515aae", + "08dcbd4da483cd7803bc3d6553080855fa5ab850a3aba5d66f955ad6d97aa9d0", + "738ebc2b1d82b2e6a96ecddf4a22c70755a0adfd00826274141eb6bb63275007", + "f634b7d27199562ad09853fb43e26c4af06adc0ade7ba877d3a44f75772ca182", + "b0e674a1084815a3093d8e7ae19bf8ed4eb84719ab260a12ead536de96890288", + "468c5b461f8c4e87531a8512461af551516074800c76b636902c7163b3d00dd1", + "93cc3fe0655106c227c17b5b6ab301b1cf87a86a4a7d14418da27ddb79029340", + "c4cd1a6f8353d7f528148b9913d339e482e9bd852bb61cbbed6ab68a7f242e8d", + "3eabdedef1a727be41647a4708562a8a061d32a52ae89313f87772ebe3bc3191", + "d071bbc56f778de48c94a831debce62fbb6f7519e47ced58a19bd9918a1bacc7", + "ef53950285ce45af4de0bef5452ea1fb271102aab977a413acaa2117e15d570d", + "9912d00f3b27c20a6dbc6c7dc185ebc21f2896b4fcfaa1e7c9a6550f911a6a49", + "cd191cdb5d7e99ac06a822478baa4c50372950f637448f6fa7f09d3f23352555", + "d248eb941ce44abdd4f32480851bb6ee3fdcb698bd21677f10bf6e609a9e764a", + "2cf1bb4864a26f929b8c0425a60128293fadfb9d5be5185b54db9a9dc1755ffd", + "f712bd37162800b64fe157503d8f6bbc30e08dd6d192115a3bd65c0c205bf92c", + "042c1f9c29251e67c546726de59b0146eea48b1c63f4ec9a1375bb016cbdf4c9", + "1448f45f133897a6a980b4857a8c6fd7f86d01520a80f0079bc0bf2278384c1c", + "89b9f8043756e4d3f7e4fd46fb523e72b23f45b884c8fc532d9986ec4f616534", + "48b3678d47eccaabf36b426fc102d3c85ec87368e7f0b192d195bffc03e64704", + "424a2b8dc14389cf1db0078eaebd01735964c20029ff7a92e41c17c0e89ef4a7", + "af1a2c5cf9fe94e6e245d76d5c8982107cf47d7e8543c02c4cc7b874af3ab424", + "d25cb72e3c427ac19f08972d1d7b0a29633c772263e7b106c4ce1de9b4fe35d6", + "2775e9f1f4969f51a51f478ddcff8b55f5c199c2e4f20ffe276515108b8ca5d4", + "ab6ec4759ad0ee40ca177b5e631e51ba83c28c2ceaefc88d668b1fbb6fe686bb", + "e8d1b780e65c084b8fb9d1fa09d5dd021ff2a4b413eb07c96d2e56b58e5d479d", + "a0998b0fab775248d06a8b31ab577b1f61f4614ea4f0c42b3921b7ac1a276274", + "e17ef9a7212c5fa9dd8cd42b169d6fa99399de4cb7e871a19f8d7b0af8f9d25b", + "f7c28fd00705770e892a2e408c01c65713232aaececcac9ae48605cf0dff8a08", + "efde04bda7758aa79c4e5af330a62325d055aba737ea524309dcf31d024629f7", + "efb5b66591f7f870e952556676304181ecd886fadceaf02141cb63c443020798", + "a209c56861df0e27ec412e987f7fc2291c4636c9dd21562524bd42f723ea56f2", + "ab25cd5e8741ad231f6bbc30a238cb3ed78d88c307de23bb2e27debd2d0c9391", + "4dfebfe4b2fdcb96a02dfbb24ea028dff4055fc26ba062af9d99027703089747", + "55e70ecc6f5a637e51384724ddf3e6e581fc2e3c5463e73f1587cb0db27706d3", + "7c9991dde85ca552707c07b01d34e0b651f696de4903318884f466203dd17fbc", + "7e14154397c25a5a42ab74d264132407f71e0a79dfbfeea44456804c783866bd", + "979240a9568dd20af4b0b62f737a5c1069e01092e8589f150127387b6b766f9e", "c327d9a2e6d4d64da75202f9420800f4b4d79ec25d1f7003b64b8b3b66e87e16", - "b34375191aab9f94d554c9e7815720525a4f2df573d91af073ccc4ff379a9c90", - "e2b9493cca85007aac45e1d6e4c64c5b87448b1f2452ba0de884f5d6e316572c", - "098a067e3762286a19d8d7d0edecb304bf19cc36d36d4fc03066305ca3f9ae40", - "9ff6fe08185f5e2fd241277dd171690e8b1f8ab5c1f21dd8390fd98c2b78e254", - "ca8b6a1702c502e22ed0c755aebf631f3fc8f8f48f0d11f30b0662937d968926", - "655d14194899719f2ef9786f5a1a4c7b3e16a35a6f67c05312f7497e58d02dd6", - "477aaf021e13ca9e9edaa7bf69ca18ce35ccf29fd5db56f27d8570b490ebadd9", - "ab6217b1a683cc8d3dacdc7c519b3b2fdac8ecfd74937c2db964d160c1ff35ba", + "68a26bdc393b45bf7cc1f322f02c83e117dbb0f9821d032c7b65b62098c63c4d", + "bfdb6b451081473f6aa34cc108e7775e9eb4b25cb128b75775e997f7c10efcaa", + "4b8cbf28c8a8bf6a74d34a79e6d7c2a45a6bd50df8645153d301a4fd4c9a8522", + "40a50e339be237809c6a6f3513827a6c45c0f684879c0b215983d14557e295f2", + "83d0dc9a57d351c6de3ec487680f3ef1028bbf186ed4604d84fff5ce607d7ac3", + "8e7b3f4a139ece2f83e807c9ddca50d3450d8f38c6206849a78f3c6b2b2b31ff", "3fee80915efa815ba9d509f9f306cfeeddbd426900ea18fc255e22c0569b6288", - "6269aa1d4748cffe56823280803998856b26e2917c80a02530165836f7b9a23a", - "b2087795cecf51387443f3969346af88afef0665b5c7cee03287bf84d5e07c42", - "7f13ad908ca71af64f12d140be615abc767512416b3de1bcce1a50d67a32cdde", - "973c45cd7050cdb4848648d3158a27b3546bd63522ac35942c5006a1063b21b3", - "3b41fa645c2071dc1d7c917ff28b4a7561a2e6a34e521a96f8131a285794fef2", - "1c5725775fde71bf4b743e3cc93e60e90b76b5df9cec7c10c07c91e9222f237a", - "469b210dc5863c1ef100b24bbc40ec1306ac7ff828bb388fbed0b95cd7488fc7", + "47762f470814b0b47fe5efcf364b91c9466b79f5dbddb6dc59758dc97f698534", + "3e1fc96d823f56ef9d3aad980980431109efc3fafc49276bcc193c478249f4d2", + "06bd132bd3074330583a82e1f5abbd792576255a46e2eccc8580e31dff52d689", + "def600f7d4baa99f1c600df8863165a37ec8a31af302d6e85c69bfddac50e732", + "69fac92b4c3b9855f93c67cce4a691e291dbdeae3ad2d2b8e361bfbc10fb6fb4", + "1f12d1f8eb38bcb0f695a74b87140b9fb201adae1f2b4ee76b5a04b06902cdf6", + "84fe8b07f66ee639255f62ed6a53b454b480a503ec7a9f37c420451a2716dfe1", + "80ea28fa848434b51ca84964f0b55d6ae5e7992dcb35811bebe802e5b071724e", + "517ce57e60c605e6b659cd913f4d73d870bbdbdaebae6754170b86575fffa402", + "2cea53b03b73610a728e1d4fc198a9de7c0b89bffeb85e60969a3dd5bb1461af", + "f48cad285fa3891f85ecd223a9a89316b1ffbb52b68ef61a80bd816572e51466", + "85830189b4ff56ea82a2e20c7c9d4473ed6d4efc8fa40c68a8df0b762f9e237d", + "be957d8069af8b36d627e867fd6f1b0cf807dc735a1427f41f2410de03379e18", + "79970e872e3a053407f48fcbf104e98eb0851a7012c29106258f5d5d8ebfa8dc", + "9b484189743f9ba866a15afd075478a907801b65ab9a13348cee07ea1a90a057", + "d6f320624bd8579764c872d53ed3642d24745b19a431d697187c051bd930f214", + "a3c551cda78892f451872ac53bd2b3ed26446ef12b44946dba230c7091d9b439", + "9d2fd0ac6b129592b533528f92eb79e6c094b14e61dad66532604df9d20749f0", + "f9a2a0c9fcdf7b2f279866913fbb4a0f7edbb332165daaa4657115920d3f7051", + "ea93cd0bf69f323a82879a52fae7f201e3c3355ea4f749598bf29eb0d984be66", + "a49803d1510ffbe80b0cc8f477be2dfd51cdb0e45342a8745eb20b88f316de2b", + "ee411b089af995be3fddf7cb354370e0701fb114b9b5b3e5f95817ac9fc341b8", + "e73ef11e2f9ea1d2d7a11b1ef2eb8e823c1fff592afa7f0d48f63b8bd2d01819", + "ca8b6a1702c502e22ed0c755aebf631f3fc8f8f48f0d11f30b0662937d968926", + "819a86d606476241c6ade26d2c1a99a1c4666c6207752fd46c6de99675f88f0b", + "e5b331519b1b8feb9774d226aa768fb328343acdf3b36502062bdbd2e5aae183", + "91bf32897f50eff47b0d838ed3b93964fe1f9427a4a633b82781dcbc2c5501ae", + "68d0f30a48b953088c1ea2eabbc10014de9b03cda792457c3625cd3cf989bc94", + "557df3df32965ed6ed84e3c8d2b15eafb716ea2b8df0c4809320c7fc3c5dcdeb", + "3f69990a51b72459a3213c54e20c1fa02dcf29a60de64a254196b7c2d1c9a9b9", + "8fc3d81b834a08f08388a01e309c0269a5f50db2c05ba33ca8710e46cb1fc6bb", + "931a7bfec1fffeae0a4eea6558776d9f7a1780b620230253979244c2d7a34cc5", + "99f2555ff39bd86cc48f06a08482436c0f18905a897f2553788b50d76e5ef8fb", + "b370962d355f0072a98cbb2d45711656a821526a9eeb71e4ac8d12fcbca30178", + "a5b04c6343e283fd159bcaaa11fb7c94fadfc52a001c5acb615c8cd86ce7f511", + "d759d7e3db129f5a435bcdc19ed7e665a51810467af15aef9ba36f6bc7b47c90", + "1c65fb82661ea75011e020e8565bc87099591049c6e333ba454a568f591131c7", + "62b25c027ecb905da5417e017522d7f0ff398480cdaef2403e2b341785cf8d5b", + "8abbdd0a6317b7e5438864a01e522aea6042dbc8c0172037354a2f179123d47b", + "7c6741f517ca89879b2835cb61ea58da39c62738892c7c288b7b3ccb2142c051", + "a24de10495fbb58f3aa24d4904c41d53563a5035b6e8d487fec63c5f9eb76eea", + "d1832e228f9e0200fe2204b2997371f068d7046119cca2ede2d9acb807d61f99", + "8b8d36e46bcdc2b694646ba3b2d66fb8c24c7ff3179256acb141790b37f107f1", + "10857ee71ac7666f5a9411835849d66e756b012f7649bcff5b8b7bfb61455b5e", + "a574c91bd706dd75c9f06fc0f5c95bbe4102a233b95c4be49160bc71c9268b7c", + "4ec01ca602f4393d764290174242ebb27ce902271af81b5e8ebf0f49406b7cef", + "4377c67a005da24b3d9bfe55f2b5a424f4ae7308b3c0ed331a04148ba41e75e1", + "0d72bb35afd375061730a7a33a92ab89e5a2396903196becd420dbe861864295", + "254bdc0c051a40f6433ecd43f0cf9d01a5a1d61f0f79fff5dd91f0e35f329a42", + "90c915c3e681bdaf5a0bac03861cdea46c5e55bee716520cc85543a9c45172d7", + "3f9087dd241945afebef97430633924f98bdd5d84cc3f4a10585049d0f4e3e9e", + "40aecfcc85b5179ac7d390b5ba76ab01757ce9cbaf70639f60262441aaeed75c", + "c270d649a1e73d6ab8ef2e58f6390234325051ef031249c77b7ba98728046835", + "710492390d5d3397b8b88dbfc7aeb2287625e7fe27e7509238a9608c9d11c15d", + "368333d51fa41749aacbaaa12a8cc2bd4383da742cf6b334e9ca8f3d3dc4c367", + "ad1493e8b39b91938170f6fe5c2f1171ca8de0ec959b56bda7cc987dc08541c8", + "31d5eabc24e5f6958ca1c26d0c5cd5980b5ce786eb2466c688e656c00b1f5cc6", + "153dd7913b1782ff85b6ee21d31747f0648b60603a27ecc319566a7374262e79", + "f9f60a5bd3d455680bdfcd91326b4a027ff945cb03f70b74e342e286faa2ece1", + "1f65adf9c1f871423da95014a5f9cfbec9ab1c542c1ad22155c5e5bf665b89c5", + "16d4d0960b3dafae0a5b533a877f3f2c7328727801ba811154edf80c41eac4b9", + "736854b578698cc4ed8b5789b1b71468ef8406f0942dc2a8bd36437c99f20bab", + "5f24705ca5584420e69bb4772460ccd77537432d0c2e91780160bbcefbc83c1d", + "a5055090f5da1fab522eeb6959b4f5e34796455185a4294b61686f2b596faf0e", + "5a72920c1202ae977afdf97d5251eb59974016e0d089bfffe53a2c785b2a0ea8", + "99834322335b1e8b9ca4785008012af3eb475a3ff26e26d675f10e2beb3a559f", + "1c25fe3a4d0e136936cfbf5b368b5e0f196d4bcddea88c9d165af123606b17b2", + "cb298a5ce212906e29eeeb938e95202862f5575eb376163919fcb181c0a52d4c", + "6a23fbdd54df61c8b0c51b12639c8d74a90ee6675723cc99563bd705667d4d6b", + "27060c14318f20582723ca88f89b915074bc2e99d9ca11d03ac3f21542b155da", + "541a19a9df7f13fe7aea32675b4078c253fd29633f0f47220008d1817cf50f85", + "4717016327d53f639038544415e62c147e43e9cac1d1434c60af243967252a92", + "5613651890ad767bc402b9df8b27a945e266b1d91be67db180c7324f433493eb", + "af0774bacd97235d890cf032208a048346e7f4f3a7ed3a34bfda605240aa2896", + "01942d6bee88051cdfd39d3aaafae45c8636ee0b85b5859b78dccaf0ab605619", + "5f278f88ef0105762e69c21b3fd32ffba0eea51fee483035c6f2e2399cebaafe", + "24942c219bf4a81a01ddb9c5b98c5e848fd619433906bbcdec06f08361d85cea", + "aeaa22380585b10ab0c5c51d9ff2d7a3f3af04fcbf4322e1e01191d92d81f3c3", + "d5e7867914543068456e84c6149597f22dcba3350660ecd44d273945cd73b259", + "68140f091fd7fb20007039a116ccda85728ffafc49b0e87531d09b02d4c52f90", + "de1d3e444dc93d4db0efd56a94f8ca2e59772e12aaad19c764f370221b44b10b", + "822b6f55b5917a3260675f3e5a92fe86e2fb8dc6e6318fb279d0913e8ceabcda", + "6f0a702f71d77bb431244b7a67bfb24c822c7b34c3830259009baaa47f746429", + "042d9c3be9d9421a595847ec1f2dd488441f597cc8e4c10587861ff3f5bfee21", + "0b4a00ab9295278221e64642273515d11711fb7acebf49aff5b9233cd0725cf0", + "62984800bdf7a12222f390719937a933ff7041117cf83b28a8c2bf17ea586d0d", + "90028d3402197aeb66805fca84e8cd7f36caa3513b928279eebedce31158d32b", + "4e25edbb040b7623aa16f2635cb59ca42333590f46ff936eba615bc21c58cf22", + "1f29c5da553aab54a6c55937f7aca8fe052cf998f907acba63ee7dd5c597be34", + "d737dadf75556447f5c7ca878715cde1653a74f7600e23ee37017947ace3ea37", + "2d6e07df0d1d88dc4d379b6085fbfbaa10f447ad771b5c09038bd14e32ff9503", + "9c5a2b24e941f6570d473a20a1fd41175e0e8a4ec61c7db92b83c9c812060dd7", + "143d95f08acb7a211a728ad7b21df55df873585e7f90400320eaff7779d53d62", + "a643ba14817db65acbd6402131aeb11778434c841c719b2e88eb0afa089bd3c6", + "6cc733a91e5005a2645d80c4ccfb928de564264aeaa79c50e2f041b8a0de6b03", + "f7437160d7702c37b1d249acf31df87a5288e42b503f1d1dfa7d4bf8e6eb3c01", + "2ebe4de4b3f7f667b71a0e66d9e7067f09fa4fbbde313e07a7738500a2093404", + "75c5474c019716c90b29a1974908fde3171f9b6cdba0348ed15ec892ac2baaa9", + "31d7d101b7b2f782aeceacafc745be660d8586efd82d8522221a39461aa262fc", "8cc64df46191735fa6c9f08d6b805a8bacd071007825a76603fd798266e38552", - "ec3222e07f9cfcbe2b644afffaffdb9368febc5300fe53d2cfd9ee11da60efd4", - "7f646b5b4ebaf6628925acfb67bc0e33b805643cc5e6b2785f5b7299eb1b65fd", - "b8767c478ba40de6e0f24ddbb20a3aae9bf31b054adcefb7ecd629038cf378dd", - "65d0e14ebbd6e6b17112f9aeaf826dd7ce25386f6a3124e6d3ea7ce88a224117", - "acf1a25f2b4094e6137259607b76cee9e4dfcae8eccf0990a7d1cd371b3be6cf", - "5d7f8be430eeee9f0c88856372468761042cd4bc4f0d58b1499178438d87e11d", - "18dab881f510bb11776bec86f30538880e02e152e1f984b0743d3dca7d2a8df5", - "7dca47dda4a459169e508c19ca4ce7dc46a8760a80fa223543e7bb40178cd212", - "f73f6998468713ead95afe7d0e6d9d5578717c71695c6b0faa683efe1d940c11", - "35813db4d6f04e77bd6f7d8f665802f0f08ab3cbf27d3e6f7247405a3b09ac10", - "4eadee052430995cded3a4d31d75e89f32f7924ecdf2955b51255e2928bd550c", - "1e305cee9f2bc920601de4729682482f9c60bf99e141d372edd4d768082d14e5", - "bafcea35eaf66e410351b318ea14b491b9b19413cff66a9f23885934e2ba2fa1", - "9dd9ccd6e5d34ce9a55bb1c7bf1406ac3bd296c894cd73d66bac4294f7a313ff", - "3c21149959213c42c2f513b184c37a32bce305779242c33b6f276ada10723b9e", - "a894ea8e0df902833659d0d8252810393772f8cd12c7d10e2e3b5ac5a51019bc", - "9765e3f356181d2e9e885c5a0e0e21de30f51b7168f04739c2acfff89b4c7496", - "2920aa9ba6785635f88e1af0fd9f022100a610115cdf14e243e61b15a47eb5a8", - "1748ee775f38e1cf43a225f1b252d0760ec240323b7ec56b7c8811a97b01d2a9", - "8d41cfd5eb21a178bb4484d2efffe7f9e512f40e70759078c024ce35d3e6a9e0", - "5efa4ca18752d896210f378efc46e6990ffb49bae9a284805fe4558bc54cb486", - "8fc380bb9cde0dade1fb4a5a3915ddadd3ef2685e9dc57b876ab01a3ed3f65c1", - "5d04d837736fcabb546a2f1f1cb0040be8b426fb1c98bf0ea26bb6f77d7657bc", - "c909cd6bbac64c8e6e0b2754d8cf584b4487eea9d8c2b511a0240bf1298aa0f0", - "a4fe167f6caaa74842449fbcf59a7f58007280c80500e87597d80003a805cede", - "71742b81845e9293f783947938306a7ee42653e6d931c6c8e4c1f0f65e59ae59", - "193f3cb2085738ab9f6c7df378d91188efaa58fdd7b8e99ed532d55dfc8383e2", - "0c5279cac3a91211413eb96f51628174442204e615173ec7053fed0b78059fb2", - "6c4fcdc3674236bd4d9ebb3846f884fa22ed9e3fc925d6c557cd55cbc12dde85", - "89958987ac19c33534a9e6efea3b8784ff87a28d07d66b5c875757248c8b59ab", - "fee1f0e7a1b2b6c94aec5aa8c7861356d2233f872a94f0f4733ec9910683f428", - "61b9ae0b6101d21088c5aef0315e8e705eb520b8409b5e09cac8afdaf643dacd", - "24521dc72e67698c03e9e2592410c77a806d3508308e6704b59515b81a44a2b1", - "b2e04a9c81b6c3446eb215875138c1848203999e8fde9e2392bf8443f77ac9cb", - "171f268e10982f79fd02c3300b1edde19ef199b37f4448e714d0580c3390c1e4", - "c4353a7787e428954abfeaefafa4194e0986f643bc64973f89ce6ef0091412eb", - "acd13fbf5a3858ad28454b14c3d6f0d59e9f106c221ccdfc0f9e1fefcf0ec3bd", - "e330ceb9f47689a868f0cfe979bd6a3f484208b1355fbbb23947bb666cf5ed9d", - "424a2b8dc14389cf1db0078eaebd01735964c20029ff7a92e41c17c0e89ef4a7", - "6b39f04366cb47f226009d5302ddb464f759c32f6767de6e3857ffb51034885d", - "0084b0db24c0492ac9f312274c64c91119859c5799d93a73028f841d991074f8", - "16ddd9595926add7d9e61259f1a19f16345ad9c4349b5caca2af7be4deffc3b3", - "66bb81a8caed2fa2af138d47bc90265db8c3aed6b3073e56184085647c7b6eb5", - "fa6c89b26f2738aa12efb0b76f4423068a52cc3df1f63a97ebe8170978180e0e", - "c422068f23a91fe0163de122403ebe0de04ae106baa09be347694b23c306ca23", - "d4d79d246209b247dc56eeb293fbf7ee7d08e962e71e947cd6479197ae05f1f3", - "3faaadcff1a7065c94dede68f096e3cecf57be938f3eab2712c449d6a798fe77", - "aac611f34cdca1f75e946318d764083ae5b96b89b191ea581f8eee9c137fb2f2", - "1603f4b161dae20b4a56d5bb3022a4b7ed9f757c5168a82b7b4cff09c150ce69", + "f9b7abcbd7b67d44db23ab3ac2a1001463d1b44dd79db1e818b26d8fb4a97986", + "3cfc0e38a5d0299aff7ead800e319f7cf57d8944dbf67acbd3d82685f615752a", + "9d57d067780dc435b6f872b22dba4e0abf64de56173ee8e24a286a91d0b82cf3", + "b03f49d76c6bb85baba54f693b27ac78f7368c55b5e4ab0bb017f1fdf68a23f5", + "082f9d4b3e056030f2f42bebf3c1e499dcc28ce8c7f04656e63ff8c9fa037811", + "fb8b42bfab4b49b27bc59bc813d1f58777a43dd895e958d84f329a386d1bf55e", + "56c87eb0821fce471d4a3c9839c0085da788bd3d8a8bfa1cc5540bbd3b11a565", + "0479d19db4a161316a2f0480b66e9be82852f6aa9a610610fe9e6752fed21925", + "c591eb7c6530c6b96fc3ce10421310b9255165ca245a4cf954021e0dbfc110f9", + "b3cd3f065d1ddab27ee225412e61a4f49a897d8689a55fa452b4d93627529f4e", + "1739d7646412fa36612cb5d8559ba3134910194def287b612aa658a10563c480", + "895cedd3661ff9039636dd29d02bc5442f5c723519f2e7bfdb8fcaf2af281f13", + "29aa03fb4a073144d9cd2c2e686f559d8eb619ac2d84c32eddd44a0bda6bdef1", + "fddfdcad679a7e0cd769c182b813e92cd35514e87266c83ddd6e290a5c91a0b7", + "ce5eb50592100172e8a2788fc81c3a8bd41d88c99148a3df9ce0007857e7732e", + "9a1fc0b2d8934f7ad5bcb81e62830adbd6138580174154407cea2d3b2fd807f3", + "97ad37f38b047bc821e4735063a24e8fca4d4ba615c73be6f17dd3124255deff", + "f0c7d9cde40930e703128b7343d8afa84daf84ea826e6d59a7b851d5d3f7691a", + "c4a95ac3ac8644689a5b8ce1fd6b6ebe50b0e3590f0057371ac1b43002439025", + "423aaae38113acebbaf227a0420787b4c66999d0c9419e2cca3adb1947f84749", + "cdbc9f553cd3137b38c3c108578bad989961f110d976bb20fa07b09e41550c19", + "8f6fa4f6e16c58fbd54cef9880c1e2a35cd3ff393e1011aa036c9821884bad3b", "b4bdb58def80264103db3d86a7807bbfa0458c690710dc08b0f87fbcd75e10a8", - "e77111c9bd1e657d55c1007a03108f458ec128e8dfe465b43790317f7b9ef90a", - "ce67786fe9f7b4d3bb05dd8e895cc39624563eea52813d23112fcc281d792de7", - "097754afe79da6bdc469f8205a9757a563193dbf1a32744be2fe7ab080aa6e9b", - "ec8519a4abeaad9522903ee9c5de0eba64c12733a5be0b12c8a0c983161b19ff", - "3e9870cc80d578e863be4aba82b2201d0ab01952ca3e49c8d4f6941f9ff7e4e8", - "917907e2762239ce222e080da1cc41d171bfbe1c1f261d805a9ba0d46df0ee2b", - "bc6aef3126d9367ef75c50335aeb0be5d094c233424021090dd6f49129f374c2", - "b41b60969de2336f0eec830d8a8ee7585594d5cd65ab47266cd8cc713a835ada", - "1b063dd8ee6c0a10ec8b32e0df805a90fafa8cdc5627d9c5afa6676fde1176fd", + "a5914fdfffbe4c8b15cc1a20270f7edb596c39b19e01d00cb840afe73c9c4e01", + "f43027efb18a5c003ced16623bd096f351c6e4be5219cfeb27e5a1fba58b7093", + "95c2e0ef277c135b9340d3580eb101ebc19d7a8afb17f9a891f9ec7292dfe1c5", + "e6bf5d704c5410de0d9f647769ad1a2963969ea67bfbc15570d01eb56bc37ba9", + "2e5675e4282c9892993ca3e8339572a9f710ad38e83d5cec5083f8cbe762eafb", + "0487627d6cad59aac739398e2dad540aa74128a70b100fab7c58fd75938c04ec", + "45e1c79ab91ff128f838d8c8a11e6675d978993d2001326a481bfad03311bfbd", + "28e30b828f9f630833035c8a0e18e6bec763c63ea032d0dfe7af5827d8834d8f", + "dd6163cd44bbbba868777214e9645985b5582e2e854400bd626cfb17773f8cfd", + "16ef6227a027882ad8f813306d21a96accabbf052f80c7b79326797efafb1315", + "b5349c0e79a4d766fa50ea84ea51d73106550f0ab9cec7ea7a03376165a748ec", + "453c6adcd7477f31c4ae4597364b83cb7c602295446af7e4f1f3105eed4f7388", + "5ab83e9a77ad7322219bf2a8104d4696023006e957bc602e4ac5acb5483ab56f", + "4148196e0ec5686318a3f3ec53a81bd4999c1bf36276ec5a016a21bbdfab0ef0", + "53b4f04a75a392e305e8f0018c80030ae2e6f3c7704feb940aea5f4daad3534e", + "b4b3e7d1a538644cd9ad42a3287f51b4867e008405325fca5be1aa55bf0ebeb7", + "947dd30e53ed74f7247970746360495e7ddd8ff03fcd5e2654b53963816c8680", + "41f46517535caa40a7eb080d16da21baa81af82b6cff1fa86a3a0ac2ad6cd8a4", + "e1f4ec17a3d58c2cd674d1dfdb4321d94f0da2edfb447f0cfb8a204fcc7b48bb", + "c4d151d0901716bf10ff9a4f2b5ab00aab01f6a2d97e586a08d2663257da8a61", + "9e1e8b87df493b1741b182d2bb60eb244e1ddfc9685dbbedcb67e3d38c915329", + "7f0aaf93316bb01cb66a9b7eccb3fefcc7222e8984e953045b7ee89a75cfd49d", + "7a711a00c3bfa9f8aef7df1b498ea93e425b902cce497c3859af647ec78a9ce1", + "ea2ca122316c3136bae3b182de8d9196f300a663dba68416b9e20dd64c9fcbff", + "6c630c9422025c7b19beae2349dfbf7370dba610f74bd9295d88594d75f9f8db", + "91443c30f78a4d21a7ccf8b809264955d32a62a0b01f1145799418c754665628", + "4ba7094cf018d7e8d0528f5c2e9fff34077b20e1d5dad8254e89f61eb1b8a729", + "de4844f82b6828b987e13594e823454de22fb71f401328326f0415968def7906", + "49ff8f963fb1cef98955269c50fdfd239bc4a5b249dfafa885647c4632609ea6", + "14e99f023d14ce7715b0ea26c0a591fefe2f2b07f9926272af38e75705e249e7", + "b7130b4a9831ba31093965f620f09e25abf357c10e75f8a303b5c5f3aa433eb5", + "1dbd3ccfb0a10b4ab28d3b39778ed20e5bd48656cd991fc9751383c6605b8855", + "cea682cb4b8e29746dd2c83d53ffe16c5082ed512888323a1681d02da525f20b", + "8c6e3c684d4ba9d93fc08d2262182438d348a318e77a262a01d0ecad14e2faff", + "511208e994f79b15a177919fef038307d6b221d57ae685f3231d8fc127962fae", + "40b6367da9ab8c1ad45c3e6454fd880c738870b6d27c6b65b14f3ba5c301f4db", + "a015d40a8dce5919bdc43e87a13ea92863317702f9ad243f389811999f348f41", + "f73a4f43cb544f63bc16b05934f52899cb66e439ed87281272918859b123b4f4", + "edb7701568653130de8592456ece18e88621430ec0ef0779dcf137ea09236952", + "bd09d5cb757381daf4670ce7be30df3a9480ae9521609faefd85c1a2918b768a", + "ac8b9bdaa550f4e0deedf09a21e7d92b24a893364e46adc11871bfc2ab12d5f5", + "5620045c29c9ddf673eefeb37fe694dbf942fd1c07afbe112cda0c565ebb1264", + "946adb86a28c4e76aa3ce357e7bc49ee64379b799fea62da9e4a47a7c2077543", + "b6291339d5d900326731089269cbf05ee3cf13c2811edb7e21c1c45c24994968", + "03297d5abcdbf272f34d69f3585eef6fd40cf8c47a1f7197dc38f0b6ea83c4bf", + "834f98c2579bdb0d95a852de2fd4261a40e59bbd9621a6930e3a1d5fc36e4714", + "6005896c10a80f03a6f61e36c72ffd19902b4b208da347b9b9eecb99b21623b2", + "8dd5d061672cfa6a98046d1876c3f894421760ae145ab1d51093d88e0c2dfb35", + "fc404c72cc2f03d2b4524393b369f867e947a899d9885f67428593ffd9b044cc", + "1af371a8d7123a3979fa6d79a42204fa6403aca196dda141dd0abfd4b501b7b9", + "cd09fbcd0257042af8316d3095478e3322f31a49af48141ca8131f181f41c393", + "4b2f1e505099e18d96764bdd0a285d22a28261166f4c15244c8cc6f97c61c4e4", + "efbb91487a172650de6b9e1b8e44219cbdc57a98e33350449f0e10e9e0fac451", + "39b16df8b2fe6f29595e7f93416f7496146f21f4607b8ac847ae9c012692ac41", + "f3c317fb53ab7d8c4721c5ef695e8cfe67511a8e4a362b557963fa22295d7e0d", + "29cc9c3c593a3af1fba3a4b24f077a7814cb03517466bc3775c4591cad05d2a4", + "157cad9cbebb5ab9928d04a09a075dc98bab3389d7f5346b32a964aba240ae02", + "f96762804c6aeffea44ba41f91b743eeb2ffb533a289591cceb6e668d493a7b7", + "ff5a102f32f3622e1449ef53a4eb3b380e6beb6262d55252c86285453998a92e", + "fdd3a3c62b38f6e0c617946706cb5726827eaf6c9d95afe2b70bad0f730a1b91", + "4e256bc198df6c1cd23e9e44644e2a54b7a547aa1a3cc8e148c1eb55217412db", + "02065626f06389b3dd5368306fd493abcdda099ed46c434e1e24ef52f7b2c32e", + "eec049b5adcc14817fac6e795e90f4351c2601bb59d571fd249d22afab21d04e", + "5a76b265f7a453c57679878b4c44d82dcf6c36e69d8a1d273d0d1aecbd16e5f2", + "ad32a2a789a0ee578fad50139de9740de934b7198cb99f2de5f5379ff907dfc9", + "ed7ca086db90264edb36b0de5c503c638c2d9346d5c0f4b4e7f06a3fefbe2f5a", + "2d5ad4801d3914da1dc6bfaea91bb2197978810762803856fd16f2a94034b079", + "184f60434f9b89dd4f23b2b7172e67a26069053ffaf5763262f5f74d982e610d", + "792863f13cf1746d9a7914170e2ae0ded839a6c64d863129d61678f505a2f41d", + "2309609c1a9a969482166faab37e4ce4afe6a0a143091a8611924facf210a317", + "bc64bd1327d88767db8f1dd98d374baa95bbb1b12543ec8c14cffe0db7ab237f", + "7989b3646d4a2f0748655884548b9dd07ec932eb2b00e2a2cdeeb945a93e9487", + "9b3f189c5fd7b8523cd52d2f1742cd9de878a3ff032c96a386531aca41c87791", + "d7311b21d2f7d1772987fdc5f661e636828799a0e7e10b5da64bbec54fe41ae0", + "544410736c4027621f051a5ec1a61c8b70422b9e9bd17371f65423b7174c20a6", + "bff640c4c2b8c334b7b847e2f240510508cad4efa7b4a6ee27bd123618ea4c34", + "c3074d6aa9745793fc77843612d5bb5e2cb240af1482663816e17a2f3fb0dcb1", + "3c6e5958d9646c0153b87024e77a63e2d08f2048da38421f5c716709d9de22c1", + "674a0d4bd8a0134ab38f2769d6da22e2cc2bfd4084ab3b07947566f0ff5e44bd", + "0a8dbc8c4e19673e7335c6af61c5b03f83914e99b810a6568d44ac496f084305", + "46b7695e16d7e75fbca2807707ed26275f33171f9d9dc166c4cdfc27197bae0b", + "390041157d7d6a05f455a71ac5621faa21f2edad7c98f282facd895b0b940906", + "d4ec529a7c44978b800c9cc0ceb4e3b69c35285c7a3812aaedf0ceafd9d79e7e", + "a1e4f9daa5f124269240f48f55619c03317aa64472fad391ae6ca3f6df46d8a0", + "71d97c35b126aec3b2c3cdf181bf87ec3ba9321b5aa9d910d48e91b491cc5ce9", + "b833ec6afb29002f0c148b2e979400e653bc767662b33cc6251275c80b5a3261", + "fab0dc6e03a5da76a947bde3863e52deb7f0f081f1344058f284d4a6363af742", + "65e310661fd21b6ba15dce3e7f519375117ab9fb740032dfb7036c1d3cf1f59e", + "1855aee933b7f613a5981d0258eab97247dd8bd611540051b1c6becc355b337c", + "b48b50791ff146dd238956c76cf5e2b7aadb523556c8b6d1f54dfb8941de9c65", + "fc7bcc1c27f4307afe06f162af10fcdde7c2b2498c65d4198cbea6c8816169dd", + "33c4028e9d4aba6c3cc6bddf98722984a72a0bc939aecb51cc59c8ec68600ba9", + "37e592abb233bad3d20d72109c69bb4da16de6a70bb4ff7f666bd0725e6478f9", + "02a9bda78c94720bfc01e89a4663ed819acaf5bffdab19cfdb197d6d1f9e0059", + "dc11c121c4257099efdee55027c39350fe075a30d106367a91c4ba35389ad3c8", + "84a7dab6b547cd9ded3078112627e75f50f1570528b06a4eb19c4b8013d743a6", + "7103bd66d9797bbe9a989834165898c334430a226295084ab41371583491516e", + "849c13a2a2cb95094c43c6eee8f0fc8b4331e47759c96ea7b2696da86f046864", + "7bb565cefe89e6a6724a067b1f17f6c26e1b4b95da625b62384a1b0a7b61b942", + "67f9014caaaa6cec539e99db9f72b56308797105fbd7a3b6e65e73dd3ec94bf2", + "be479191ffad1354013d87fb918a8685dc9f55d5439ca5288745e4c5263b3566", + "66116dff74ef4335e09101d65a3e738e3ee87515287c4aa718f6edd260db34b9", + "b6d9597dfc4dcbb66d9e5943eeacf97d18313e08d02ee2e53ab8e39ba5d1b359", + "5a1383033b4308acb3a86bf2913a62c2d18fc481bb0c5ee5d3d06ae103860073", + "387bc308ea79ca4ce0ddbeb2448a7002c393308dc1ed648888c76f0bd1dcb343", + "f5beccd4afbab6e6b1feb93a2423d3d869745715b23079df40d63a7cf2791f46", + "1bb48c157255f4310d0171552ef423187df9748d46cee6f28c734e10088227ae", + "66525a545fc7a350d6be70293ffb2926c97ffb054547c3a21b0325df1186eebb", + "c9e693b999a8634ffe000cf8faf465f03d52607e09000eb9600412b75b5002f6", + "4244eb9d12e828b51dbcf50cadac2cde5221fad5fd86f77a1506e02b0b1a71ed", + "09f3b1262a75b7aaa4f77a0047ac5160446cd0e2820a061859753afd9d875717", + "0dea677023ba50e92aa3f1e55fc2a6fec4e3da9834d09b657e5d27ced0176b45", + "7a5cf83833020f61bbae596a00510f5520fc676b52ddc85418840355d56ce733", + "29aff1a5d38a3f506958b0b5f68f375bbaa45b326462fc6cf078eaab597d44be", + "4cd01eaf2779c4ffc9783956826482053c6d0bb3fb8b2f788c971fdc97374f27", + "ba6d56cbdd2766a0531c01f1ab6a8c443b4471fe1a8073f0657d4d47d6b68879", + "24c5e2455f75c13cd97ede49ef88a8adbb0f96845b9a4d884f3c1a33dcdb2a2a", + "b335f05b5bd01647205a618bd9b849bd7e8415916a9449981e42a508293d13f1", + "daceb92cdc9c96ed0d8c1e577f151aa563c42e8e0a79a85c4dd9e61e6e636b28", + "099e448b42d1a4ecc057e1b220b95e5d3457fad57e51b9b139f7ec90116aee3d", + "8c8611a2be6b830d5e9924deceaf6105ec63ee7ac7a510f59ea965b9c2241a13", + "b7b79fc322f099faa7e537889f7d2120f7283669fd23a80a1db8d878af302f1b", + "fcaf46967bcfab8db45b87fcca7ee39fe0d0dbd0b3fb2525432701973661bc6c", + "9d16cef96c6889cf5f3ff9d1688e72d3f3ec7c346d5828a7859119ac247faf3c", + "ba954e8508daf948e9f13d58e038c0f5d857fb6d94d55cae57b6ace29f46f062", + "b589ed7be672c6d90951b126bcef6f1a0dc8bd8c8ed2c1c038029f84085eec05", + "74312ded4ab3e60841818e5c20bff80a26acbb5749bcb7ad9a42c26e7a045821", + "c66097720c827e0fb694db8cc48198406908ec769020a713b743fd5ffa02ef52", + "84ba931e3c3f890bcbb185866add6b77629d0d609a5801b2ed3f4a1d7c666229", + "4b1191b7289971243df6fabf3e2c29f014ee28ce8042049c0ac723a2502f98f4", + "1dfe040a8603b435b7dd40f6a7f3bd909a201f69b6b824760596943a5a8ce552", + "4d667e0df92acc5503f70f58611e81f00ecd6b854b36b78e10f6595dd0fd695a", + "324bd470c85d4cbd58e7ca22066b8ade8cd385b55f69e632252622d2d048bb78", + "b175ab532b64a089dd2600f4f8d8708b4836d0eebadf6cbb3aaffd47e7f7c85d", + "fa00cad6318fcab5deac2dfc429b053a357dc1a9fcc63720946e0c232470e80a", + "b7505080ecff3551058cf9d990b8912a2c838f8eec137b86f6b3af5c8739f90b", + "9331aea4b6c1ae678b8d54b9fd9154d883196777d4c679d70df6121ca0fecaa9", + "44c1237a5857fa3fe39c7b6b1ea8b2a1665fde3260c854d703b0a5bc6dca7b46", + "8d62267b76ddc017aab7e2eb32f1d86e34596c8437d59048f344a4bf883c7873", + "c611bbbe97f0ce0ecec36f3e37cf8dd3f40bf5dff1954c2d40f52a6418381604", + "cb8d04c78eef0dbcf0ad59371e778d652b19d2fbbb46593ff6e21108b7710a00", + "3e748c871be69d4fd2fc6126610cb8f3ac91698fedddc225d6e04fe25ef5c81d", + "c661a074e2569f00b4df422ee87db627f1a270588656f031bf90712fda1acd56", + "1f77328762c808a05f51a9bf5d271083a3a17e31414930e86c60f0d130188fc3", + "9f7600277d0324fad0ee967e92699485c13c22334f8fe4131517f7f1282dfb3c", + "cba60ae0a3ddba66933d44e331ec902b9d1b21af2076c36029ebcc36af5f6e5b", + "02a0ae3625a77b4ec7b8400be9aebd56d4182ca7d6988d450d9bf7ad1c03a4e3", + "0bf371b173846476b2b1cb6402bc2eee57f600c789a2a521259d8dee589c02cd", + "03d4e8119de94b6f171ce9a28652a97f20bc66de49908b6f31ae22e0b5dc3eb4", + "232cf6e971d05ed8ffcc87e2a4f2b8392468b156fd22f75a5f4ccdddae97c51c", + "bbd8b2d98a791bc7856b6888d14a528d3d292cb70cce9b9f6ea59151643316a2", + "7d3767832f41dedabbad756876fd154d7993be065bfbfa2ef7e494303654096e", + "dae2f04d5c87fbe196eb167fe5255bfee01b0c3397b16a7c3c9ab1ed2d9fa5a7", + "975e464727abce072b2c46639fe135523dff1d27ab6d9d1d61ce355cc2186506", + "48f7747544993fb749cdcf2b4c37b3675f48b6c267cda7cae49cbd12f2523b1d", + "56db578a2a4f1dc279d9d57e07fdab8ae790cd4463ac9bc1411507ace8e7450e", + "2ed0d26d2507cb94cedca1ecc27ada33ee08acaea9f1ffc59b60864916c07e0d", + "3293e7353e731ede95b16eda37d5485dbe4b76b15b4594c7dad16ed2337f785c", + "38038d28d6414f90d1d51c574b47c1571c6f9119ed645852f455cc88a7f585e4", "f8330ed3cf34f4143f6684c12d62b40cd82356a4f4c2678521a81efd7f765c98", - "f484f27e94039c70312bf9232ec2958086a09ae9bb234a42060151e1367bddcd", - "043d50463fea7e22ce7795d47b4e13ef6ae65a7d334347ac8955f989bf80ce47", - "9338ef09cc1092b5098e072b37127840d4d6f1afe920d6b07ab48c88ace3cc0d", - "753cb35ee4847089ba1850bdd350dc892fbe6a5827bbe4bd63d13560f4353e5c", - "8947ef3879fe1ae4be375b211f8c1291a522ab5594a3347c85b91dddd4c41b90", - "80ac6948646738c7d905c376396e05801f22ec008a01e829b5e53ff5996052e0", - "3ae8dea85749ed673af5e3f8aa5dbdf989308e7a29f064fd25678ad00effbddf", - "45cf930852797dbea9d3f91224cf0f87a1b1cb6bb4d5ab91836d80c416893574", - "d3a6faba672815a2aa7f7385bdbb2464ff64dcc5110dc3ecaabd6ab97f91f70e", + "f59ecdf3df94bad6680b4e5fe1ab796ae662757da4505dc3f81d06bb31eb4881", + "8f69335e110c001e688d0a4e574dc3c3690f40492fa6544aa3feffcacd5975c5", + "d9e8e73a7d6d28a368d020464a9ea27adf61698602415016c2407accc31f1f2d", + "46106637d84337453151626a9488178661c3e7cdad537b4a2c0a2af0ea2c9b31", + "75821699ddcaf8ca96fdfb2004b93d9e1c5673d5f9489dd309d311aa8292aae6", + "53971779e072a356892da034fc69cae45febaa08b00d156cac73dad9a5c9e16d", + "cf8abb67c74697d90a3564a90231b5326fec8a302d1a3099825e68e7209baa77", + "ef6e96c4326231289c33c465c9deab43ef1eaa50d1aecce6a2d1ab3d0f4e12b7", + "3d0ab0e47cea53b9663e485d2b2847202047c7cefd1d125c22f1ff60e90bd48f", + "739f9d722a2967ace510c206327f7caee4851c64f9b1c969c82ac56ddf3587c4", + "599a4a4d3d976d6476ed7bc45e8006c92e02dee5a4133e541a0dbf3a6c80c6e0", + "51e658f27af66cf59b12dc9258be90286655575ed3fcb5a844943511947398cc", + "1bc9071ec297d6877b22518dcbe0d9003d2bd35c0568e178eadc52426a1f9ba5", "c06b7643bdbfc990d9d00fe4c041db5d527790193451600883db7bcc85a4fd8f", - "5d713a5ce59bfebbe1033ce61ba0d8427e1a1d284175f4bc16c49eabfba58917", - "b853a0e43ad5c6fbadff3f42fe6155f72972ea546b6948d079432f04d677cdc3", - "0e74f22d3088aca096d16cdccb25b1f7c78d5ae56a622d2e619a4f97cdb29f30", - "4cdbb6940681cb8955440185764ac041f6837e900ba7548acd988528a70b8c67", - "fab0dc6e03a5da76a947bde3863e52deb7f0f081f1344058f284d4a6363af742", - "141dbb05ebf8bba7476374676c8bf7afe903fd92158a3bb19563666f2c4693e1", + "958be5d6b373811e6c4e202f9d06c1688fa745606d3e734b313a6661707f4c42", + "e0bfcba5a98f2371a7693f8984f0a10db4f2ed3be1dffbc09117ee1076f43404", + "6cd5531c5877f98573508771c2ec16c8857a3a87dc2b955512b90f6a0435e38a", + "34703b4812a88caf3d4953eec29c1f7368165ecc4e34399a90a810827db33e5f", + "ab1eba49759df3b1ac1a13c307641d20aa92f349aba9e926c45de9d32b413b43", + "3017e2173fa8d6b5f703066a363c1b606b5d84060ea59772a11f6a155bd22e9c", + "f3ce8b1285fab08dd31a9a9263132d6259090b1bf01138596e99a892291fa087", + "e6e80fead119db118243287f30771c7b34c92e7d4347b75624beb6f1e32d406f", + "2d2747222c2d8d3bf87c61511a076594eacbb97139716556a29ee4c351a84830", + "4a4c6844e9e1af50ff3234037b1e3582ecea3173f5406a40b6bbda0b88f35f2d", + "ccb931cfc8c5167f98bfe5d180893cae3f0b0d5f1281ab3083e0874f15730fff", + "21062697ebafae6ca503c34d2324356dcc018709838b8a118e130f52d69441b4", + "2ece9f61861fb81bf63c442d8a96d3bec0580ec7bfdf294f13562063c404ce5d", + "bd77f2f38e02c21517d9ea3e66253871343f9d2b23d93cb6921bd84beeb5dbba", + "be3acb3349499300150b38aad3345a31e57fb656455711d4b7ab3979ea8e7740", + "38d39140ad23a2df2267be925a24177c30b1a2e904164f6e8e6943cc96900ce4", + "7c612a8361950c14a07acd8ba243c31e42441995302125d3db787dad5e6ad338", + "3a9069176630df2d75fd89f888d18744b98c68a1f31a199fce998939af146ab9", + "9765e3f356181d2e9e885c5a0e0e21de30f51b7168f04739c2acfff89b4c7496", + "3225deab4f62eca68dcfdc7619e9bed699e9ddb8d2969b1a6b63fc59f88056da", + "fc57f84f2123e9f521f937094f40898fe24189406e9bd4a8d44e834c64ec3ecd", + "aa1e3735c56c0214f56bbd1fe9fcfada30e04cac2c39c7cf772acb109c4e63ff", + "038bf256323497b5eb8a8b1a93301f4141c676b57ee3c6d5e7ef306818471e81", + "82c5a2a7a30d43c54983cde90e4898acf1b647b4d8b623b3ceee851bdb8ebee7", + "b41b60969de2336f0eec830d8a8ee7585594d5cd65ab47266cd8cc713a835ada", + "cf47296d77235ff3d6877dcaedc531bbfce39df317152161f3225b64adf97ad9", + "e76a0631c69ee5120fc7a886f6ca00268129b32ceaa1027bf80d7513c10a6637", + "1ec97cc8577bf0603daa585fd402bb4fb1d77b5d5bed6d0bfdbc6ffb8cc95157", + "4cd21d371c8193ffa06b50a979e1025c224a52c3213170b2a7db6f2e554f92d0", + "3ecd8fa1be87f215b0d7afa12c49742fb2e2ed907e2068caf7771853f8897c55", + "ad4b27d9e8cc25343539860f5340a944d8e14652d634a3a0113e625ff0d768fc", + "56f5950529dd6b96cdc18b4647b5d2f89e0f0b70624fef4adc2f8a6bdd91b7ee", + "0bacfa3d6cc5c8605e9196207191be56b77c23f497fb891b2bf1e2e11851b6ee", + "5891d3af8b90fa7de17f63cb39e5c79f34807355dc16fc821f6c4a5d7aabcaca", + "d4efc188f1625389997de9c51e685a4c693984565dda684f22b0a29ae9bf8c93", + "3f8a8ccb663b855fc664a780ac4d7dd966d8b4e550c6ac581799d3b2e1a16225", + "5f6f5d67b163998523fc5053b7ff6e87f6a8d1bef25623569b9f8b05e8fe73a2", + "d94a0b2f2737eef69d32a395974ac2010c1315f0ee001b146ce7028b09c8a3d0", + "8ce7d0eafad0dcc44caeaf5a69965a525d9291c1e0cb44db603e57fa767dacda", + "80701d9dc4b27d996b025b45083c04ef476a64d6720a0752a1dc0e75330b5369", + "ebba4f8c20215f999459499db1ab87f800ac19edb6501e797146dce7f89e18ca", + "7da8ec7654339933a14c1f368ec73be1e736eacd7ac0dde43f45a16ccd2b1712", + "175130c5dc414f388bae187df1d6ed73e75aef4e21a06bede0d6325202d6f116", + "2b9684ac67138333622c5fc28fac790b5bf153b9a0e2bfb8fe1ff56ccae10235", + "17dab4b13f9bf61cdfc84b3f2096b39d13952a9669ca78d8e74c14794b4d280b", + "8eb6905c8745867e38d145621aff4e8d3cc234b4747a58407374e5eae0ce9c42", + "e00fcf5fbced48fc1e3b3fd04e513091cbeb3f7621301eda6abb7c2bbec0561b", + "a4b93e84a16db5127adf03b941ecee90ec757c81e16a23242196c673a9619216", + "9c95547816e19d4b0b37a0acb8d37da605e32c01ccf7705c8bc39a5c1fed0f12", + "656720f943ac7e1d63d51b7ac666f61470c2d19fba83aa9b813a0600ef7f2e40", + "088d308908f7a568f64e849d32b520d9a76cd1e731fa2eea5ee29992c4fba0fd", + "1ff928c103c3af1c1d0ce613e9d4f10f4b34b9330abf84b2f1c0a6cb1873e766", + "adf877bb3f627379c6aaf8717f2baf3bc4d419685a9ad335a34f8dec53867979", + "e44b553f4551ff6848347c9fd29754c45c55e4d2898b723edba3076f14e6f0ca", + "b6cbc2fb6ed60ec9e07894084ffefd2939a80794f15635e63172ae062464bee9", + "33dff16f6649ac3d979ab4db6356f11dad52f6eb25fa8dcb49d48c25c407aee8", + "63666520fbd12c24cbeaa89abd9358d6569808eb3648450f9ee2ad6fb0681e63", + "80ff52b2d8e2a657085445a926316a5cfa29ccb2f32ad1901a6ec2fd61ec8417", + "5920843a393e1d6e33548935c1a7683f45c7008283228c05e7c999be6d7dbe50", "aa29bab534fc86443d4496cf50c371bd4729ba2034994705afcff2b55c8ad3e5", - "72b560fd3244dd0bac4fc2225c2df852c70d916b40ff29f339a27ffb9448e35c", - "f68d82ddc2728d4e35d2e493ba546d30424ca950680778852b16c6545350e5bf", - "d759d7e3db129f5a435bcdc19ed7e665a51810467af15aef9ba36f6bc7b47c90", - "d47b80ed8c67b430397e3838703b2f4108b524600fc0f9f48ec864cd4c0f0e01", - "e2b519a1792ed3e73f2375fa2502c6e3741ae3f2029319893551518eae9fa899", - "532cf49b0c7cf490c9fc9fa051de7d071ebcf0dfd4f05242e4d3d806233a71f5", - "028bfcbeac6c3825b9a277e83cce98e6bd2c0879fc361b0c9ed87bb77122ba7d", - "b0f42d23efb8a05087d8a8432d425b091dfc0b924e783e0c74a9d6e43ac4d41a", - "fb9cd9337b624ac6ee90662c1c95457b097bdaa35289f504643430cedd7be4a3", - "5b2662c1afbdc0c25b63dadc2a0feecf16b0891c106fa29a98ba95ead98e01bd", - "db2a8bfb9e1c3f65785804a3035a9f169d4eb70fc4224fac6c447cb63936a336", - "c97e20ded7eaf8374a86b96801546892272f280c6cf5eb79333759a620cd11af", - "368333d51fa41749aacbaaa12a8cc2bd4383da742cf6b334e9ca8f3d3dc4c367", - "1bc145c2d4653df22a7aa3b588897ff5e4ca4524be41d430ef4209af7fa364b4", - "9548aea75972eb2bd4fe7c80efa25917fb104659b0ae1c0889717792d6403712", - "a520e3e50e5604df0a00070623a4e0c851f67aea913aa51d381277675d708632", - "917677723937adca62cb907600c9e9d588d7f41af8395a477c52d9f337070fbc", - "7b5c006dcb5e5e406ac6dbaa09324590e2693f5c9152f3edc7c5fb90dfc3e83f", - "d737dadf75556447f5c7ca878715cde1653a74f7600e23ee37017947ace3ea37", + "2d79e0b0e3ca2e6e64f42a48d95657b2a965831831136f4d2d7d93d90b8e4f3a", + "08d35da7e61f04d700a70b05d20f1b78fdc39bc3ac4ff93978c0e07f2d5effef", + "85dfda2b834f5241592ace737c86bfef353ba8d7298a4d05ba677f5056aa592e", + "fa0ddc32674054e003337e8df2f106df53174f2a3b6673197b23588d1f1df2d6", + "b5126d508db599a49320e5de5046ae4b1c98f520a3b755a095d733ee8c56c969", + "d1877ebc1ce981dd2ff7922822203cdb7bc2a44f1ccca69016f06e4a6e6408a0", + "9519422422a309af27151deb38ea35646f950ac4c2aebdfd36be8f5fc2d731b3", + "5959fdd3c053ec58903b6d9ca9a80d0ece648fa398932bf8099e02c05318c9e3", + "934a7f4b342927add00892592eb7e34521a2e4a5871dd7a5c52a4502bb9d6d04", + "b17d5471d1594e75067298ff0986050aaefd21fe2f35043d3f99113457129eb8", + "23894a181eec6181aa5fb4bf91aa9eed3a838fb9460a64d6ce035e7e90ffe461", + "c22b47cc850b3895cc65e4771ef885b70ca6f0794e824eecb91e3c6213714bbf", + "1eebf4b74f232a59375f2ca949d4a7a30f98d9003c9554d05d0295e4ea1bc550", + "d637eb2ac5605211159cc9a492348aecfa69c5273f57e02f2d329b9b42bedb03", + "84569dea2e042a6e1b5c5fb3ee46cf3d234d38f169c2fa402d2eff78ec80ffd0", + "787af078f05ef25108ae779a0be0c05e69f8c05536aa9e0efd0eecdac8c67e45", + "2b5fbd38da9e0a8f6b61ddb8e2fba9f41b84a19351e981e751879d3e6d57793b", + "5dd9cc50c9c73bfffa22ad8b469f66d5b0ed84d400e1b9e8e89ae5af255f0250", + "5f55db44957979693dd5e3ae3bca79bee76d9c7d9f88bc61162087d3dcf34ff2", + "e374e788460b4f96cc686ed71e222fdd844e2e2cf32a58e946ec14b2beab1902", + "38f9494b4e2bb63ef2714b5f435414bd8a81fcb14a19aec342ea9bf1d5de78b9", + "eafe3c77fd4700a320b4b19c492095137cac11a6abbcb37645e35186acb2083b", + "57e809e0f5b5b19362f54e29736ef77a59b96f32fd72466c570ed002970b6fd8", + "66b32eef048ea0966de711a51e1d44b7567a8f49cc33b8afab6ec59b7dd44610", + "df831458ab8492b0c58746d8a50c5c687d562e86caa732f8dfd53f3fbe9f69a2", + "8b5142ea1b32aa1d0b21de2b0d3b8362a4c270863f5133e47e15c5eaf247c403", + "3e6bfbc079aad7e9c1c5a27157d67ac0158a73c0ab418df3103aab55a1894702", + "e11ef412abd160d665bfa7254827a7898c48f6be3f62e3c56ab607576ca62935", + "7446e455fff239375ce1f732e96f001e96e535289657207f346e78a77c72f5e0", + "61ca42f63205097a998c3621fd94a191f6946a471b5ab4afa602bc60f4cf8a43", + "55f67760c812ec1220fb341ae0395a4230718e5771bb93b30583994f5751c6ca", + "5fe3762a7d73fd2894465206c01e33205bcbe437aac7d6d15d3b1cd3c80e039e", + "1ecb5f0499be8ac12c04a088aee9a51210c03821330dcb37178c0922db9a00da", + "d3338935cd57f8070292d0fc3e8c6cc3f4047c1581c4b6dd761e9a1cf96e9790", + "4b3ec03348174f143acd32c724bfe0b470819a336dc7b17feaa5f9e7120f4c9d", + "634b5f1e31945329576e35d27431e1800d64b606e819eaf2c48be16baa0e7448", + "535f9c5fa5a6c9792541b3d51225068dd11a436379c98f7f6a2b583e371a130e", + "0caace4b108248ea9e851ddd7b8a5dc5f097d72f87040e8a8c597cf81c6aad56", + "23128742cd8f9e7ab1cbd85c4747869aff5de37f5b3d703ef348e622ee9a9a5c", + "ec79f47c81ed21ca8fa435d3d78b573fb12d66c64822b68afd80d4dfae98de9c", + "b6f69d998d028e254feee680e48ef6bf94885d3bdfc36bdbb34bfc8a8b1e4c66", + "41d9ad7703744f3c106f8826e7afa35a553032d45951c1fc1853862c5846d306", + "269c413dd3e17f29e6c008ff85775d98ab15f8ad7f2445f068ed9d028ddafc2b", + "e37a4ae6f6aa520cacb4c29b9f17c73f826d411a1380eaf2272e8aa3cc847c8c", + "2651d236bbe213ee55c4582c2cddfd64a2205828ad08ea75e9c9d300a672c7f4", + "9c5b892ce7621cfe287e75aaadbaca0c357ab9abb485b826f542da4d03c2d506", + "19c37d80ddf5eeb5e5c6321d039d95bd5490b80667a6815bbe86512733cab24a", "a048e024df6e874b66088d1539517916254e5bc6b7f74d095a129085cd55ceca", - "23f192da585414f59d7f88cb6364432b05daf19c207f89f996dc19b14bd20c1d", - "915c4f811b88a03aff22bf4d2e179bbcd51496e95f2411ea79b6cf5a2d4814bb", + "f0aac579629094dc3521a4022271373ce4724115420b7f57a0870b9ccadc4fa3", + "d5c613177a6dd81db689e6e85b4bddb53c212e11f1cfb0b2cc40ca28e56c21b1", + "ec50a3e6a729df31340b01006f9b857b9aa0b78ebe47c7897345e088542ed737", + "1c07610d40d4e2fa6801b38e17eac90298fd3792ec2caae732161c636d8a3de7", + "7f8ad7243876e0cf3b74fdb2f19e041b3b6eb245f8b61f7fe17f4a6153678ac1", + "3032c3e4b7491e233eb29ef960fb9d1c021b852659ac338be985c42a0ab44134", + "73e779fffdee19cbf60b0d60c408116aeae8ae74282d77a4e28be23889f48cb7", + "5475edee98501e04e40fa44d2d9e9cf2f8099aa77340591f7dcee3519c2632de", + "d0fd33630e4b0adac7c0637dbe999b62d5bfebc3790ba32f2c31fc9a5a2eb7ea", + "fdeee8c26eeb7db5e8561f0a9544b0c8b4835c138f1bb30c1f40180697f46930", + "e06860afab4651e6f7122e41abc639676e8b10c9b4c2590afcc2f437dcf1ee37", + "3362af7ade8090063e398c57ea8350ea1fdf97751e2b73532a81a3dec06b73ba", + "6b03e96995b645d7501d4bba8796ae09cd9e4f806637fb767e11124af149769e", + "a737a908f36830672600019a23ce01a758c324aab701c4ba10c118570a939320", + "d3f525f79c1c274e0cba6cb9a1bc96ff1f4d8d1d3c488d8f02d7f460b1d63e91", + "36494d09eda9fb276c571ed74c4f6b84d416413fc4cce40391a98e608a637359", + "1e305cee9f2bc920601de4729682482f9c60bf99e141d372edd4d768082d14e5", + "e9632fa4db1dbcea724c92f97a2cc1f6eb3f8675e766a3abe23aa2758d5063d8", + "48a2c73f28bd2c39be7c629790936afe29f78a1c2855463835f3c0d6e4c7b2b6", + "1e46e6e0886e4f03af17790e5e87b0e902b94dee27e1190b6cede0e7cd670f4e", + "56feae60ab8673fc73973f652dea9fac1e0b46672db4fcfc5a6fd23a670306ac", + "e7fa76e7747a041a702d7ba25d7bbd6d1a83ba5c35879c90c793193f223f36e7", + "ca1b9111c8c4f43ef275cf59e89e69adcafc5b7cd9960b031d41509f925e6a31", + "451fc323ad59f81b8278d059a07d873d2002fc79a8bf49b5b94eb0c356c0d2dc", + "33db90b2947590c9bfe29b276715249ec08b2e8f8280c5a1c980040c0f6b7abf", + "1e4dbeb81c69af17bdfb418fab343b2d18ba0cac15175d31d7d9c6312412a3c5", + "5667ab68b9874896ac5c6c366b66809b271d8615d78f641a90464b0103f20c9e", + "03d15587b1f0e103b673839201714b8d5f5c8b945003fa56db766c9fe5663e35", + "e58c4a2b4af9f4d44d689fd4fe0924b5d1df60a17831d2bf3d5f7ce5a69eb0bd", + "1d894b6a83b72e490a1a7ec5aa7d2654c20ada1b7c58bed4deb8d52897747c2c", + "3c4ad87c6e846a3a8bd2f7a6f783e50c58d12af954740d878e06ad7400312935", + "181ade391d0cc3decc29f125e6ebbe9e1dc25856f137745e1a0ab480915fec07", + "1fcdd3ed13f175b4886bd687a5ba519dc222122d1a3d39ae463202e71570cfeb", + "68f58869a12b8ad35349a22795f2ffb9066e99571437134aa14c97ebbf47661d", + "81bc30640f4a434e0dd72ebd38c9ec9acd0fb36da1ade4416545d67d1b827600", + "5d2bd8c0ace54b413947bb42e7f185e661deca2495c088f5c13ee2d8f8f9782d", + "9baf527ed085df06f1a2840a9a4e1f4f98cc66a351aff6955289398242da53ff", + "1c5725775fde71bf4b743e3cc93e60e90b76b5df9cec7c10c07c91e9222f237a", + "4e4eb49c926a7e7030ec3cb7f15f6eb5dbd00193568e9749b2d6459301dab0a7", + "2558ebd88dcda2a9c7399761489b731ea62c98782455ae0bc6a0550edb1f97b5", + "268a3006241c4d365bb5b3655ee1476a046b8177c586e8cbcda8fd0ff9264b7f", + "350ceec4c87bc78d8d44881f385b27bb09075d84e4e8ba17c5367de952e1483e", + "deded155b9a3e41b1b5bc55cbba4bc044d8d537551078d29508e356dccfc5446", + "1286ca49bc6ad665e96251c91b82c6719dfb2410e2c9887ec4e4d6ef1ca9875c", + "4245d700f81a1b406c0d702817f36b18ac7ed2a9b6a0715fd7320f51f8599531", + "42ed1265a481bf35d2d063aa41fbabd7cab2e7acf33d3082aaa6d2d413a59a60", + "8fcc424356bf9a4a6d8f7f9eec68281e3011d374950b55d4313b182ec06a8989", + "2eaf350c4c491cbe94e530565b8a2ea8ed0f9e9635db1245fc12c2ea08830124", + "c30f987a0484566464a8e615faee37cc5a6a5859623a19dda89929373dcc4865", + "4f75a7680c0bc5a913d3e78766c0535c89877e9aa04980c28d1d8717db2be3de", + "0ada516c29fb2598b97a436c49766e25e1e27ee1801f272a630af3e0dac15700", + "4346c47e8f6b57b03d62c29710d01b59589da313d0ece0e880db11ebfbc7c6f5", + "d8e515b5ac67f5c4a6a0c9738c8ecf34e32faa2eff6ed584b0e179fd10d093c0", + "7403cf81b2b5a33a939487d956faa91006c4378f67998c87cdd179763a57f5a4", + "8174f89184060d9fde843d0c2c9a3a44659630c843f71e481396edc2406d6b8d", + "53d52feb7adebe4052f8cb5f3290f3e669b4f004bd4b8cf3452070dd3524af81", + "c6c41bc10875485e273153a9bee98b02c849f0ba2e1ffeb98895173811d33711", + "09bdb72cc282d8e8642d2ece4492434d7d73850b05b7a33cc0f976460a104c81", + "f1a30391c75c83ecf7573040640231354f04c79c31be41955650323e7b6bccc2", + "57a3f5cbf8bc23888bed1fb864faa6e6e4591cc9e865af5dd7d0db10fec44157", + "d2ab87088ac93793b59567c9387965251e5d7736f873499a609841e756b2574f", + "721949870a6496848b29ce4573c958f89d0087f65deaef9aada8ee601bf0edde", + "ac2645e8b8cc6e98a2ebba9f625ec90e4819bfb54f44187acbd9b783d9653c11", + "2745607bdf4dbc5653cc56c49949819babdc35a8b796f730a125f449d0e5e087", + "9c3aa0a51e411fccb3755078cb24f7c612eb16594bcfe21df605f6424c40f767", + "c03599aac0bd448e042cb4db8114753936d49ae79217877ed84f189710c5d69d", + "64befff9b59f1e1d85cd0427abdc2929c4fa4d0f4f2b620c203491f763159c69", + "018a7e18a20007d9dfb83a45c4f06f26446b4d272b9f212aca1d324e429fefc3", + "934588fb786aa18b07c8eea9343592577f10b6e86e4c42752db6404db6b4e601", + "957d525e8aa523b7fd6c4236f66c554fa8add51089beaacb7f27dbbd3f725fda", + "9176e62d1246298ffd8702e97f8b9a5992288868e526a367eba0d67817d2683e", + "22e552d59f5d6e86d66997ea5fa0f0fefd484fad02f80d9f8a9be7d000ccfc6b", + "14a0e2dd045c94c9bb5064d139b62d8171a56214a65cdbda10dac3e5c24169ce", + "7bfc73b3012f2e4c1e0ba3146a5f9afc2a1ad9bb026715de2f00f943bce18e43", + "fc8a74259aa7c317c42141f18dd50bfe09675f09309c90595d56e186ed1ba8e7", + "22494816e3fac3312aaf8688671efea2eb4b1364e3e80cb6c2ef636eaaae89b4", + "fb58933c95416880b99bcf5013c466afcf693cea54b734a116ab9d2324ee9242", + "72980342cf07d68e2400ecd4e245d87915fc658718871242b8cae25e7dcb34ac", + "2efcde941bacf3b4caf68e30ff9ece667a4ee9d990c737f25633369ca0499c82", + "a4dab94534075fc96bdf2c1847a05243f0026014c26641ae19d4e97b8b41f6e9", + "fc81ff4981c88cfa2aa4a525ab634d0496414257e39d4df28f70ab6eaf2c7741", + "dc9d0eeaa77c80d024ba850d32fda4bdf46ffb509a40886eaf5564d1b2a77d86", + "6b4ab9418041f5a8c67c4b179fd7f80609fdb410a074f8e71d8d52848b7a76d6", + "c1bba0da524c3926cf1d28254763163cc3499b27673c255bf9e19bb6a9b2a663", + "2b461d909824c242e3ebe1debe0852a4c892fe792590849f8e6966c4ee781367", + "89916258c879edc79db0bd39f7a53590a1372033931e07ec281f1b20ac1af485", + "4f83a47f7ad2954172e15267a02fd6d45a1c4a25ac41bac1870145e82b36a896", + "8f3ec88baa40f6b987ea46366b2544a0f0365896f03eafa9c310891a87fcc6f5", + "583eff3107d7267c581c2a03a9b7798c14db3952996ec21f404a9c1f7589dafd", + "4bd849e2880a12764cdc1adadde14e5a8f1972baf17b7a677fc9e916695d6c95", + "f4629eee8d5a88f87438d097060891c9e54ea92bcd995b97fc4702342db53b56", + "5943a54c5d1d8b09e5fddb05b77a5c6916194bb2f49ce9a9bb19f3bbc52b509d", + "2573125cd429cf558be813f24633352fff1a93da905501321bc2a2cd951d71df", + "31171a6005bd5c6c3c0fcb9d98dc7c4ed111292738453d9d2ed40787e4427f95", + "cfe1812b981ad0958b49ed47e747fa506c117e10d245dff6942aa8e309d8f028", + "7b5c006dcb5e5e406ac6dbaa09324590e2693f5c9152f3edc7c5fb90dfc3e83f", + "1dac5cc749a1a967e6dc1a7f7a1489b30e18aac943b7c494b32d9fa4b64b4fde", + "7d41311c2845d30c20a2f1776fd87414757df2f88f2661b7092435a42e783af1", + "e64990018b44469bd55892207ca7b603675d5a661cc68fb863bbc6d00ada16d7", + "038d518c60b6a97db15cf3db5c44284dcfd244b87c4c31b7a4d4142d928627d4", + "70642f7aa2e3ddb2938cc2b1afe07e306840a45ff7e537c2466d3483bc8c898f", "124e5829ba8806e9f78f116e92d81a4aa834fc64392ebc0252543eaef48c46ba", - "e0a9aedc4760d341d8adf58807617094260c513ccce4bbc7fc54af2e2bbe426b", - "946adb86a28c4e76aa3ce357e7bc49ee64379b799fea62da9e4a47a7c2077543", - "53b983f4f686f27a3ebac6ef4c62c2ca2dff241b9f4abf9b831bdaae7cbb82ea", - "a41290b70ee6d5ea9b50e7b94a3b227dcf7dc24a6b309e7890c0129f7e7ef270", - "09d12977620b274d00185548e6c2a9194375fc0a2bf8feba4be6d22d78bbdadb", - "bc939407a6147f3fa7e8edd1c571bbc68dcf1bf20916f6122482df51c28f44a7", - "fa40c5c696398716c5e736647a8f1c943c156b7fd131547a487f0a5c7ad610e1", - "e374e788460b4f96cc686ed71e222fdd844e2e2cf32a58e946ec14b2beab1902", - "f7da3d2573a3735642071a909d18fd04de7838ed4f0171b18ee9985efaf24fbf", - "10a8e1c77fcaa55ae12f0af27e470298d50216317aba34d9bcc68b2ab486f1b9", - "605d19a96c5e06ac5f83edc01259636d3a2c052774f652b780012dc8e19c9c47", - "56a5d304df0abe6be4e67f25dd700204f868fa9bf6e6e50704c4cbacd4b51018", - "6eea41de0c8909f225e83dbb289c4176682916f6d0c5edae979e5fac7ccd544c", - "084b9a8077db8f47311ae8c6b14a049f6d9701726460e690117f3833821ee501", - "af7805096f04f78039fa23bda8f3ed966c1543baca115a72a03ec5aec233d80e", - "7d42ad9cb15ab7fa75c96b4591e6164ef5314dd8b65e61bd921f306fa4fa27f8", - "78faaaecdc3439ce2a0b6f46f911373fb73fe833fffa64d321c1ff05de6cddee", - "559a78ee1b672f56627d376ab726294aaccbe218dbc435c2201a813493bb1e83", - "97456379a258283c71a75af0365407cc3d2f770ef0ea912e57529e06c0775f59", - "67944e37e8de925bae505c6ce67ade18794efd81f53aaab485b7e2de4cf6968b", - "d71f4775bd73a2a6a83ea98d94d9a77e5cd3edb4e2afe49e7118e452b1b18c43", - "74103cac68724b3c97b130e869c05adc328377382da98164d6e85efda67b62a3", - "4f82c906b9b0fda184e3137116e52c746b9014f732217d9a67096922e8d4ebb5", - "8670d9c9b9bf77820c9df3ea78c9ab0999104316a078172fe9b94ca6461610db", - "eaf7cdbc86897784041e2c667bcce0634418d3f8c1cd34d18840331044772477", - "fcaf46967bcfab8db45b87fcca7ee39fe0d0dbd0b3fb2525432701973661bc6c", - "86eb7d97a4c48df4d6bcc4832536f1b8ae6049815bf54934474d41b2f8294209", - "3d1bbc2c84154096b20c468ccf332697c17891cda610ddd64035ea3d5256dc61", - "55f9987ea0b7c7b9ba84658b25c4bcca4487199955231c2c05e58b9e95245f48", - "3e87c2aba10caac3443f27ab36ef88c386ef848b3e23018b99c8fa9b57d8231f", - "f27e3a4f575cd0d5cc9fe7a58271630750bb84ea647d979bf934e1b5d70b3a34", - "6d7b009f1ac86a0b3a83ce09d7ccd0c0a83ce42bc9c4132d09a9f13ef8579e3c", - "9d2aec493640c33ab01830e698c10dd8e7ecedcebaa0c46b0fd67dfe9f0406b9", - "6b9b7c5f68ecacc4698aee67832436ad2fbf0933f5eefc16f3c1f58bd8f57d62", - "a70693501e2c6a7f4f78d0c8cb3274c3e02e90d3b651263aeb813fd8ec4368c2", - "e6773493b9f104a339aa88ab390aaf5c799dfadf43a8133acd6b64dea72ff942", - "367306f71ac377334b36e9a69e8146aa5a0ab2b54856f1ca5447b6bcf799e940", - "266311def1e1dd5fa1f58652adcfd57a3e891b05f2f234f02af09d2ee9de0ebc", - "fb4ad272e5b215f8aa316babdfdd059ca87059205c1c06d64c58988bafdeac62", - "ccbe7ef95bc1fff376f53f9eebba31fad8e5ddebebe4e583398e2f38c4d38d56", - "ff1736259af3d69fbcd752095bdefa96c13bbe5ca88a52b724f791d67e3b02f2", - "bf2e3af74c876546d21c3f14042a8e0e65631d6268dd34d23934c138d8bba403", - "2e5675e4282c9892993ca3e8339572a9f710ad38e83d5cec5083f8cbe762eafb", + "c463a52bb7e61135bef48b5f6d3ef2536f6ac729b733ca57536cc07bcd32675e", + "50cfcbf6536bef8310a837a8232809604ae53f72f5623ef94cf2a90a003a38d4", + "ad0eaa3e9ba7757ebbb6beedbdcb7958599eb06e9ffb49580d020259d69b9531", + "6290bdf556ce7a5aaba9053b0c88a9eeb221e9144d0abf24c1bab54514301d7f", + "9e7e10912e27773019e95431bf29ddf3bc90b118fc093d2a3096286552a0857b", + "0d7d61cc6ab6b5f76c7b719e8ca16bf585bcfa02c8a58412fa5fd7db5d0142c8", + "a894ea8e0df902833659d0d8252810393772f8cd12c7d10e2e3b5ac5a51019bc", + "fe475b13bfb6aad5e0527a23e3c16f89240418ae585458d105a257848eaa5a71", + "4b6acf9aef315267f00157645152d2d3197a845af60e9d9479e364c3a2de9adc", + "8b88898315b58e1dc5de516f151fd440a12ad81504293012d44483b20437ae82", + "ff1d80849e9ce6105a54bb043b6604c497dd2e2ffd5b6c05aa32f42e27a12fa4", + "ebbd11f4c892363c032812116eee801a88d1b3770cfcfb591a3eb8eb28ceb9a7", + "c4735514654f26154befe5459f638088eaa20ee3650fa6422229a7ce84250a54", + "d706b3a3df747159676fb9cf785fb8ea0acc287d56aa22b585a333ad00d24d98", + "580ad7045357635ff56205e71e0b46e90add2e28229dc7c4dc441b1a8d4fdeca", + "659cbb3b90b3824e15b5f21dbcaf6a53e04d629cad62ce1633ccec87bd2fc91e", + "ebdee3a0f32c343a2b80f3be5cf5b513300606051c15b2b4ad6d9093cceddcd5", + "66386d8fc3509c38d287b079035d7ff9f8477755e1aabe1d83b8eb3c1c1d6485", "ea35762e9b66281b05a4bfda250fe8c0b6e4f29f80eca5c38da3ccfc42f274f6", - "25fe769bc1472f65f4f2fb40dad84f4d18f6221546adbb32d22e38b745de1703", - "834762685e8ac1d43bf98b2a9f6aa1fa05d4c1d61e3e7394f27c0d01fd485d85", - "d3c22df7a989af10820f173b70052d31990ccca8d095c161f453ae979dd8fd60", - "74312ded4ab3e60841818e5c20bff80a26acbb5749bcb7ad9a42c26e7a045821", - "3d73c2e9fc04fd4a5c39d4106adf341a9f9af729a298220b33012631dcc7fb61", - "1ec97cc8577bf0603daa585fd402bb4fb1d77b5d5bed6d0bfdbc6ffb8cc95157", - "8f0be68694d4b7f7446ff4d38330a4da01fe6c03d641dafa8d0ec1dc7f9cec8f", - "037f739c101b1ccbb99edf283dde8336a1ab50926f9043bff5dd4cc2f1c4cf30", - "4eebb6d0201d75730d1c55e244811d65b1ce006d2ed4cf5b44032ece475ae083", - "6b62a53a9ad9a289d11567b3fb525e9a55cb0fb2f6a7bec08948f3daa324cc37", - "9141843d407abcba6ca8987d9d3f3e84837fd91f82ae18f4eb7bd6949eb8553c", - "58b32639f3b1095d417d64b879be7873672c60137451e26412314d39fbd622aa", - "b751b90f3123f1ea47ae0bf57d06ed1887f430dbb4df42afc34ab1ec35dfb6a6", - "c66097720c827e0fb694db8cc48198406908ec769020a713b743fd5ffa02ef52", - "995a2a16cc7b91c92bcde962e8f94d7cf20ebd7ab149d9ebd3505e6c80ef6a18", - "63883408082fd82e95a417a47a5418063da555b99f60de65177fdf1fbc7930f9", + "e8d95488ca09e31fa1b4de0f0be2a367384583bee0d1a826c621a5dcbe9a1f26", + "9b43decbbc78f64db434f681c206efbd5d0f0d6f14a0279c78d917b43abf0119", + "0a5a58219ad088efd963f4e4a40ebed3d526f4f434dd4550b1d36c8085de2b14", + "a64208c4893bf07591a3709d91c9e4055fa27f368ee6d2ba9d9032b65e06a20f", + "febe4362f08bf4a68d2d8d38cfcf74cc0ed2c16c72845d824bd2a52ed270d89d", + "5d54fe131050c9078080a041b359e0ddbc9ac30f1fed527b4ab99b2e048e5d2c", + "cb21c373e5f8d26c80ad199032f5a5c9b1520c44fb2fe335ef3c5926cda9ed06", + "16ccb2d8cc4194e003230ed469e7e3105a84eff2fab81f22a381db6f6c5d565c", + "3f5874b4ce0ec882c2c75e8c65a8fe1d8716171e8d0e15a50e42e89b7e933d11", + "0d3fe65d690694f2ee40b1a11062aa2cc979a3a17ea3e200626e72444429ab0c", + "337dad054e31d487e2549dab6ee7673b75be14d481190eb22017fb02a529292a", + "16f661b183757b34f4a7370a442f14cbcc7afd1d63f02478b766b2f80f32b581", + "5f6c33e6ef7da331066a2fb896999c37265fe04ff6db14a06456d3c3815285b8", + "9fc74a12a953aa177a63a97d31e836d0c61fe2a1f5947ddd8e1403956f0515eb", + "b20b107ce58db5e4a43a1d68ac1a427fce50d5d22f0436154afd69c82e077278", + "86eb7d97a4c48df4d6bcc4832536f1b8ae6049815bf54934474d41b2f8294209", + "4ae3cd65309776a8964ae692fc5f737bfa4ad567929c65193f38007e00c70cb5", + "4ab83b6bd200cb62c6232ecd28f3696e0863e4cdc0b880b643758eee8d718549", + "a8a8733d03367f04f4d1cfbb528425a22c22e32dc2b093b15df33282d810622a", + "5c92d31aacc40ea6b0ce4a1f1f5c1c894322ff39538d5cf9592586ffe63b2622", + "efaf6e1eca0f45a5387ef9ad4f5fa03be3964d591d19cdcd5d5a8186f9077f2f", + "2983b77265ed3864afe660935e2fe55c6165848ea0fa98b60ddd1721f89544f4", + "5fdc802c185ac236e80c43daee9a3db503d0a03f914560ba6fa174404708093c", + "18ad14fa170ee31924c97c0ded842125e4989c082f9c18158577f7346244a582", + "ead0c05b841ac3c5d1abdeb06811cbadbe57d5b893b933cfe72a435c12836618", + "41570a508f4f9465d6bdd5a0c66399ce377fd25757f3c3ca3359d2d0ab2f13bb", + "3909da28ee35246064478980e92d3681a2871775984aa3e363ddab1357973f97", + "276e8f7b5a81d2ce52ff0f92de4c25345e934b5bd2eaf676f38a8d688f4420eb", + "69fdd9de68f0e5d4e05c9584e8945fe77018a435a60ed94239aa177fcbde4808", + "9e8a36bc289025a49380835c95155d57c5968d55543b36f380f6ab2f37ca5f81", + "d062089d90f337968fdace7e0a374218468824b5110ed379785181696a09b4df", + "e2e13e237f734cb7bb2e46c1cd4965e4538ba116e611c45e9552ea65da06913c", + "063e06e1baec4bcb40787a30f995932cabb8cf145d675f96743d061ea7dc9b1c", + "3e0c28c09c63a512e521ea290a980c1eb39e00f97248a823bfb19f09e942120b", + "69214e62da42fa202f7c5cb984da65adecfbec19a5b201b2fec08a764483d1a9", + "9220b0f709d0a4462245500d9da90d40303cad69f0a559ad5d4118a407fef11c", + "5bcfcba38ad6d0dcfe114331dfdffbf5686dbc6d10e260473ced52486540fbdb", + "61392903a432317378d3622b6bcdc2f1acf2bf264d09ccc5a04b38e762a19c1f", + "7434c01ac25cfc7253dd1b85fd5d8aaf53de73b997ab5024fa1d6eb07f1e5cf4", + "9ca7ff5495ca2d1a8dad1fd1992d8f883598c10631d4d305c3db29c5cfb63194", + "4d6e362d91e22d18cb5c5ab2c5b6f7f3068fb7efbf17c45cce38892bac597bec", + "1e67e17d8f08ca81624101cb2a0077793d5a8645187562c28cb59bd5047fbc5c", + "2eea68adbbb843ff14cb12cc7318cdd2c42e0a9b107b12c7cab380e5190aa77c", + "0c54db8d68bcfef16b4af16b7f73e7f0caadb279f88bb1a39dbd54f2a0599e73", + "e5b358fa4823e644b12a994925bcfe366f08e35e4a3426fbbd877f3063bf405b", + "63f5c7e8679b923332afcfc5b9c681397214a67fc8d5effdcd35c49ec5d433e9", + "9306ea1ba4089ad317a4004356af8ce788c285aaf838d8f280744abba711c173", + "17527a847ebff9fa080ca0395cdcdce7c30c5bd4e9469876fb96a4b3134a63a6", + "fee1f0e7a1b2b6c94aec5aa8c7861356d2233f872a94f0f4733ec9910683f428", + "86ff92e8c8023b2ab3def4ffdb2b32203241c155729c807d08b2a6fca8d25f56", + "b6058ca9eea7c3b492bd9fb5c14588721dba3d3b481ae15de643f737dbf325f2", + "26f0e251ad40ee01ce4125fa66d84e04c48837f3c3999d57d9b356bfc4f12d01", + "f9beafbcd03802675eccbbed4409fb90733842150bacd8595811931f73a0c10b", + "af2876c6c618aa981b7aed761c5fdb6aefee9edb5ccbfb1b80d6f964804d2595", + "cbc3aec249f6b1317bea0dbebadddc0882827350c4027727763b32bf179d7340", + "97c5b7d8e25d39be24e77cae0ec3dc7fcfff185215c8a3ab61db889c851580a1", + "a538dd3a74396bc0a342eff590b52def891647dca6c93ed39062b715fd8eefb4", + "c1bb8fcb195f017bd96e376ff450ba4212fca15c87b638f9f8fba4bd16f96d1f", + "dfd33f8a0325612caed307b3a78db6a2207624604774dc8e5040a40c2501e541", + "3afb778e3278269af010255673e5481821d61e312a1f2de6e67ab1dcf20ce45c", + "f48a1b8dbc1418a0d8e3e45f83cb53716a5ab4a2beabdac843aadf89b1ef98fb", + "eaf7cdbc86897784041e2c667bcce0634418d3f8c1cd34d18840331044772477", + "45b4e658f32ed2d317853888bacb85450b6489205597ca61b2cb4bc7d00ede10", + "25099c0a825ad15e3655c2450facbfbd413736e395dd5460d09c746ff7d41933", + "510b1312005b62b4fbda15164179dac6df8375aa0b5f9f472998c2a53eb427b3", + "3e96f9605b634e622d2a34ef861dde7f1799de7cf03ca2c16ae42eb52c818a23", + "20a342752428535522675ba57ccad285e29073dd2164e6ddc923dd51f138c62c", + "72fa2d18acc566b6c7ffb287a0a8fbb0a29af71c7fde5cacd33bd2028cf23d3a", + "bef0df48471fb9909563ede8c30f5b9150136d7662e2d8156d608b45b28e8d4c", + "bfe759dd43de35f3c3c068528541e531e32307d80d5781c661c2e04f1e2159d2", + "6399df8e4b616e216b85d163ca54813242d2b00ea8d5e97380e16c556412bd4d", + "322ae8fcd1f52c7ef116a5d5437dee0d66dd650f570f1d6b5d0a91a3a334ad7c", + "32af066aed929c0df85376f9bc2ac54164636b60f30a8919f9f3b4c0f6f202f6", + "6fbf7221445ca63693e5107fb88054e931931986d744f61184ba4718a817eec2", + "4c876808ccd12c4254655beff6a14b2cdc7930821bd9f4a83bdade90ba5c2961", + "e9aa5948a4c223ba9eb3d8a60a98707fc5c80758fe0325e09e18349e368ab148", + "5f0df6e17aac09d31352e5df885fc34f8fb0fb6799006bfee649c1097aa75cda", + "c0556c414284e88b0e660571c74c02cdbf6e1a9e33a24a240a40da35b0768a4b", + "f1fc6e988730b11a9cb5c010dbb49154cf04ad0036fc0642426b423fcb7cc4da", + "24fe7353bd96e363a3e94b09e067b7cb8f0141ba58acc84ee4132f0e3e8b7db5", + "2d88bc1581333fd8183afa8d40e11b8228a55459769756672f585f8fdabfa175", + "36b1dafa9036ab548944b382787bbe8b9dfcd873840de58f738571df2385dc47", + "9f61f89454c88e32dec5212bf9ad909a7ae10edbbe3388341195beccd3a01ef9", + "b076d383a3a1da8779de4933bcb026b63a747bc4b7331f1822139dab3f9df945", + "8648e2eb953ac0e783419dee072103b06013745d11e836af9122c5bdeea7ccef", + "d29d9cd8499d02f0ed6594b759c37fa9312bbfedf8e5f2a103e88874f2d4a665", + "206a8319902e9de9654852adfb8e6c4805c933aac1240c60600a7ced3a73c17c", + "68777d77c67c37ae1cd5ec8a77e154f2c01d6897b17e0d030b5ff695062c716a", + "3aa25397502994a4c54f1796c3671c6a3eb926f82ccf45c3a21a8a09f1730e55", "faaf73d9cd0d0ab77f5361a84655da0e5bc35abd1f969b8b4204b36e10013044", - "0c3f6ca1b275fe12eee752aa6370da0c2ff58c9433268b5dd4c6c58c9a0c71ed", - "a02277e9a820346f3a3cb34baf0acdd0e339d874ffc694e4b4f5eb6fd1fbdca4", - "3f2c66974c726209c62de5278b500637e169b547d395bbde68149513149786b1", - "468c5b461f8c4e87531a8512461af551516074800c76b636902c7163b3d00dd1", - "6bb524f2b9ec9fb3bd53abfc0693baf55ab1a9c3c60b6d316cdbf532a3df60bd", - "a74676db4c63d73df97e998f85119feea1a328e5ee3bdf996e29ff99e109e698", - "405d7b084e604273e8a724499a84c6d804aeace820b9fec313a2702a301b519b", - "1011245303ea0fc2caf6630d14255d950063e7459610df5eb21eb912a8c3ea8f", - "ad32a2a789a0ee578fad50139de9740de934b7198cb99f2de5f5379ff907dfc9", - "038bf256323497b5eb8a8b1a93301f4141c676b57ee3c6d5e7ef306818471e81", - "e36e733e044f411c6f7d1b53b1723c70f8ba715cd85cfa2d7988de603e884ea3", - "30cc8d21303caff76188bdb08c6220c027378fa4820b1339ed11e1634df39a5e", - "f624a4ace0bba7636ce0946cc90fce1ef4a92a12cae6ec3684a81ca6de3137fd", - "9b0c9838f646ce09a1cfff7c465d7f96d9ce36dd296a20d90d0ba2e99cd81a20", - "a20fdc8e1cd70606e3743111a2bfdc4220e239f53ee7ef0322b1a0c8669cd541", - "0518223388c25add09bd4c4c872ac5c133db4428652fb2f06d3eef0135cee845", - "5c6fd283afe0811d419ee0f18df577a4e9a178ac6aa745ad88d69c236e334da0", - "1bae704a7481e6a8742fce18f16df78dadb1377c7162836b8d29eb213d2a9a52", - "6c42c9b166e40ddac5a0709f448917cc83194982fcd59f7fa8aa67c9f49bfc00", - "4a793e37e45e9a77854762e08a8e8b8ab4d5ebfa5b5aca8a7b39aa88b8432f5b", - "7eaf04ee49d9ba5d69354cf11bc0c1e99b7c598042d7a1c84bc746e3e369e5a1", - "da8734dadd1dec2229cdc5c08a4b0834b78aac4d42f3b9071d1f543bfe15f329", - "ff1ee3533dafaebe36b2ebb0bf761ceead930f2d0724c596707b75d52ae4bb97", - "08b7b635a8a8cf66cf36e2dd9d9f65998681b619ea58ea1862d5141208c8c575", - "0d5eee783396740ba21f259bab0f9f0b45412cf5342ea1991bfa7c9d6fa0bca2", - "3c406be14b5f0fb3ee090a9a9c86ec98ba42f66ce1dad636ced0608fca68b323", - "b8bfeaa751993a74d971d2e11e84587100f0e12606f1924b3732cb9a627e3415", + "6b9c3a77dfe9f64f0996bd8c279881cc6f78090394d8406896fda77492f8e3ee", + "6333fd5a77b72c623df665bf999f3a26315df8d30b24b153465d8d6612c923fb", + "6db68b3719a1bb2a4e1b3ba7f774ca9f94fb7b40b5a8e31fdc38bbe27c44f381", + "5d713a5ce59bfebbe1033ce61ba0d8427e1a1d284175f4bc16c49eabfba58917", + "07ca8d6ca361f16e47c97a92761805b10e55dc3a0b186bcdf88ff009ec9198aa", + "a308a15ad1eab2581ef5ecbc78fccdd06b4e6a026fc9de22370d9e1873c5f56b", + "418646502db7e22595d517e9152e6aab2a7275a6086553a0039277f531b90b43", + "bc6d2467200af8d3c98e88fc2d42efbbf4ead3cc1fcbf973ef28b111cc9591a6", + "415ff58a8e82e96c9b0f1970be36b6df317afa124a4fd5ec733cb27fefd98c50", + "ee082b9a4a5bb337118c39e4c1f7a747e1067f347e997a7cf4a42d8d60eed0b9", + "c58acd86d2d1e0df779b60b27bd1feb454f0cffd6d1ba7a4ef0563f3a8228221", + "5f171936519a31898c0b39af8a2d59a8608c1fc19ad840ea7d1400cf57705250", + "d35e42224b7ec69f20fa429c814a7b82e1d84d591bdc2386953275c4094e3218", + "41676d615e501921029e3c8190e52bf93eb6461d8b00ba37911b337b436fdc7a", + "99fcefa02bb2586fb0b279f340cfb2ab35447872dcfee961972e7985299922b2", + "a450e375da0f5bf91cbbd3bbd64b225fb0ab2024abc1706e753dfff0f75a0ec7", + "d8f476f32d8fae21f7a2e35db42e12dd00f22c0aeee4401e02f4f956a66dba86", + "320e22a5079bf644d6835d6986f7b1ae805b1e05015d877d43a83a68e6834944", + "5fa84b847c49398b6979d4eef5e639b56e956c227fc766146d4bff37fbd33227", + "114fb040cd0ac5e719df97cf928007b2fa1a890448025b6144ba6e7931ed4111", + "3e4ffcc26c781e534df06cf5d498ba6ecad7b1b3d96d5eb0c9fae88bf3ef60e3", + "9110c495ae41b1d3cb7b2e5099e5dc5a7c01efbb47c40e2cb82365f810c83383", + "19226632d93e5ca22276cde73e7d90fad4b7b9e59016112fcec1eb60173bfab0", + "8cfab564572292bd524fe0b68519ec89be8b2fb658397596cab875ca04e9a3f4", + "948283d9629ca8746bc2b94be5878911a39bfb19a1ae1d0d6c5018c06f31654e", + "b5b9514ef445df71b7823b1da948f37e59e8c713de099f2956d58c7e8a684d7e", + "c1b4c1022ba9da764d4c448dff8a47ee61b658427a20b612f771adf2614e4309", + "2a92aaa36e025a39cd9913d992da9959a012843160424acb2a44a0231e64c148", + "03cea2556ad0058a660b06d9cfc191757f3c2d752fa96e7fb3d99f6d2b973092", + "d7a84cc931f1fd93f03878533c8ecbc861bf53a9c12e673829a7d806e3e85679", + "42aa3ab02dfb5ceb0819252a405b41f53c8e8b64fc9eee87bdad13cf7c16ce58", + "0c3250738d14c98e740de2afb1751a291f38ecc29a22887c7cf73889a948172a", + "289b00bf6cf961ad59b1ecfabea805a0b4de763ecf8cfe9c364d4ded4d920286", + "e7b22f2e9967f59da82618f6919c3dd8b4899e625433bc3517450028848ad456", + "3f8f1813a58cd84aa6e6609178a35bd4c78995bc6cda768d0ca8c47e94ad48b7", + "6c52084c3f84e576307f76654da7a2ab241f3a22f04d4005bd36311a3e8ce68c", + "ed78a3ab6e4659ce869a2c1b25a5f6402e4ab37aad3ee12a745fea82e9ca8d8d", "dca213c5ce75c4a990458f0544534259477a08a23fc2b268a91213c4cfeeb85a", - "e33517ec69a6d577f36fd6352328534ab5a50c22d731c6e904928c9f02e971da", - "55f67760c812ec1220fb341ae0395a4230718e5771bb93b30583994f5751c6ca", - "4cd01eaf2779c4ffc9783956826482053c6d0bb3fb8b2f788c971fdc97374f27", - "01fdc6c5fa3be12030be97672a7f9c98c9b42162f0e4a7355d0b8831803c714d", - "ab5a62660ee3da52b03cb49b2183058e965b6e386cbb3ba173347cd7e06e78db", + "ce2a28281a34bb1da7f25356fff1b3c1b524759900cc8c77a7d296be4b6b4ef7", + "ffaf7f9ba2956848629ab840e97e093993125d89ff6ebbb6428d7f195be6f9cf", + "6fe818a99644f329aa5ae1eba2cff8d617fd6ccd09636159f4efabd23b057b34", + "8de6816aaed632d622a7e37e0e1f6a873e2e121572e1e761e38c001384822095", + "c26b3e98980c62b4604e59b0e875cc118d0c95f729531e1a9acec0ecbf4282a0", + "2f9645a4932d0076fa554e33bbbd380c591a0243ebce02f03509077fac2ebe76", + "924d54acc8fd4da8b036f5324ad3f4acb4859ed0a265eb98fd5d9723c495f09a", + "24035e038e7a6a8ad12c9928880e7996d3488a311f83e69675566a4af8f8aeab", + "8f99ff6a214fbe73dab68aa6fc0512a7c97707ff4d186036bd95f0bd92f8c8ce", + "ccbe7ef95bc1fff376f53f9eebba31fad8e5ddebebe4e583398e2f38c4d38d56", + "2e4a7d5047ee2e964d32cfebfc588f035e060d9bb4daa31316b9e6691c22f9a1", + "bad82874d2ad4e780205b55d63d17aabebca991337728d0d4ac5af7f77239fc7", + "ef41cade46cde77fef820cf499bfaf4480e5fc60bb0c6d1ffbae62a14ef5b6c6", + "920b4fa1c76b90ecbdea193250d301250c9f0d46137625b1d1fe52aa409271ac", + "9113edabb02573c3a490ac7da86d1107cd6861d824385faf7c28204ac439b125", + "4b9f46ce3f7de0bc38bd8647fa76f1ea60a312ee0a8784a2fb680873cea52e98", + "db9cb7e88dcc03d6f9ced278b7a68d3fd9496d28268d86754a9e660d7c7a4f4d", + "0aabb3d71736f2f31a26dfc9185c9f6bbe6183235a94654cb4971baa16469933", + "fc5b6a1d71601ece1c1a1bac2ffde7a22d26d0836bc713fedbe1e417f2ea587c", + "9433645f919258600f84f5061d6d76d5e4271df70c9bc84c934dcf5b0d448556", + "71e0cd324f86b739511442486b8f5a24b7c16b5878040a23f394c80d6db8aba9", + "e2b519a1792ed3e73f2375fa2502c6e3741ae3f2029319893551518eae9fa899", + "fc7593355d12e83716f11e5197a236ffb74c49bbe17996c49434993c40fac947", + "4fce5cd4bc41ae22c1949bd50cb613e18cdc25a242836f92ffc4c71594233d7e", + "939086acd9da2286401ce78b7f6905d5b2513357e55ec3cc4f1a79fdd47b6297", + "a38de5ae4ca4f11019ae7986b022f33d0af45c923a3bd6fd4d2ecd537bd289d5", + "8917db1239064b966706abd82d308f2d35177c007a5c034b2be52cdf0b4431f0", + "a43a1ef6de7c3e17c82881b6b0182e35b4e10e1f353b725fc749defe9799cf9f", + "53b983f4f686f27a3ebac6ef4c62c2ca2dff241b9f4abf9b831bdaae7cbb82ea", + "0539d7019e4d47822e92121b1c1de118890d844ca89064938df8b8d86ab18b2b", + "7f39cfda7c7d587c8ab6fbb83604dde471951d283b86e37accb11d8734bfd490", + "52fae125a52d3ae21aa2b663f28946b52004e86a00aca2f44a567de1bc9db5d4", + "ae6b324b4640ff1d5f711adadd03897ef3136eb27f1a3b5c338432084bcc8ef0", + "678181cf803e41a8440a003478c5595dac86495234877d0960ea71d39558b881", + "25bae733a439b4295c42c60008dc73f0a239d11fdaa7a6fb2d9600ae1bad0c3e", + "62283c3ec45066063769087ed57a39a52ba9978108c11adbab8a59eefb241dc2", + "d6340c0c702e056ba690c0bd79e54887f6f88b3f211919f8780a3110252375b1", + "3a8484f2abce6cf1df9d1b6a2a3392701c0c355302e6f23e1f589cddf5dbf597", + "8ef8774d49d93a578563625af4bd02d3616da5a7c204af0bf91d6d3c040526d0", + "71ce9a4b1d3a604c4610dc557ac9ec0e00f23298a10bad649d8c736b918970ec", + "1c4f82be36da1aee2e218e904dd79ce1766afdfce9abbcb3f31b74b728a9e97a", + "6fc7ffeed5a3b6affd0e510098f302ba2714a7ab5a8214d6c73016b14bc2874c", + "6ef546a4c55e98a9c5c0109fd838b1c498981b64be4a975e7c36732c8efc5965", + "5191764717ffbfffac9ab4bc3f302728f8605bca3110ff83912437d1aea227a4", + "d8d5a00968d17e92d52ffca34b58b2eedd5fc76dac027d293d8e83d72b901cce", + "4d13625a8629ce9976d832c4ac8f4db9413b750c862fcf88cde29f76b8c5985c", + "bc878e1175f47b2fc23d79d6243ee005b3f7714b65674e8c27c98f71b63c84e7", + "34477dcf2ee1613630def5924cc443f14d4feff5bb52657298e162cfdf23076f", + "509784fc075aa8aca857a3a527a6c80beec2ec3af53a329fd96814e315e4b966", + "6a130f11dc1ed98420041a1dfa0fd3609ba6d5ea843a97c45c1dc9789ec4af94", + "541a69bf11776b9b6976c687986492dac4063bcc9bf46d4b83e03c9bde1dbf3e", + "0a4a76d4bebcfee1d9f08413efc186948031e3ec7d8c2707c16ad5377d4bf156", + "4c97a86ad229dc3da66510f6bd046669dc19dec1256986940259353e2dc281aa", + "7011b7a1643bbf6b4d1f368bf20108ccf92e2f5cad600e287d5377baed991e67", + "085ae60501219b9bb085ed503810a76227a96435ea5a9f48f67dd6490cac0c4d", + "1d9cce670f8e88484e261206a8b6cdd0b67e10df3afa83872ad2d1a7cde70664", + "5a1158584f1cbc3e2990f35327c80ef0ee0d95e10824f02d5ea250bd3e4e5649", + "53a5661a2e9fa6f60b9e3d2299e6cf4fe0196d4a850be93fea8b007ea6142dae", + "b411765d7248297515aa36569b765fa11818f1c077535682e289c07b4c6d0eab", + "64d8a8d25b623ad324bbbfe34ef88a9cbda105eecb6390f30a9a437c4afdcfbb", + "0f25dc623a3784ecf6e0e4f2846cba9f817ca22b393379b3bc7b9bc2d7f18c5b", + "5c4e481352ca86b49924eba14cc1a4ea3b61bd2f4c7ed0ca7a9de07573bc25ed", + "7911dabef72f1a49eb416a71f6639d94fd5f1e1842d70be8628743ac7680d595", + "4b1356891102bca011c7e10f3809ba6bab91b9f41de8aa3a5b3552ada5b6afdf", + "87d6252634e2388d67587f1388773e34f8e5a97e37402ea21cfd69af845d0d23", + "adbf9ea4a068cb22500c34c94ec88c0b57a48b53c2cf9603a9c8e68bcc8f2b0e", + "07d4ccc9dfdef28f472a454c224c51cdbb5b41b4c9f18ada88eff14662e65071", "27c12a0ddbf31a2c619cb91c42cc5e5ebae0608dbf46e0bf53f986e2d9d05770", - "9caa70b23c8d8d2a36992d53dddf8d04ff8db21fc8e2f3d90dbb8f1e49b11c73", - "82bf7442462018b98a362da6df19982831dc7ef98a1cb710bc6d1b3c5a9fb9d7", - "f23e6d1515622b6acc45d60b922429578850f2135b35f06e6650995bb52a1047", - "63666520fbd12c24cbeaa89abd9358d6569808eb3648450f9ee2ad6fb0681e63", - "f6b0c8e156b00d7d2645325a7288b12de04e93b9e7978b64fe28d03e359d7a0e", - "96d836a9dae9f603178e26ddb21e95e3e826adf7a0e3fafd482182fe8f7d077b", - "7506be0bb255471fcbe31ac6fecb7dd05b4160138ba792b7229ed864255aa31f", - "090b31052c10e7d81ece797794e34ae952a7f483c5d1015f8cceaaf18472fe8e", - "4933f477ee38ccb597bbc512eec2751ef24d190df6d1415f166df39c761bef4c", - "0143be881ffec936a22957c75d1b38317ca3bfaf4ed7a7623992b6dbb68e5170", - "42ebdc3dc91b842d1a43e9374d8fd6aa3f67653f9ddc67a54ee487d896468eea", - "237770240ca0682db1a58625b670105a9667440400548731f802daad2e429796", - "4b2ebab52ba5ddc868bfde99d026be500ee011824477b173c0a52e43b9360f29", - "a64208c4893bf07591a3709d91c9e4055fa27f368ee6d2ba9d9032b65e06a20f", - "c172354584d21d8df4c566e3f3209aeab91b8eaf177ebfa60c7868dbb0259209", + "53b4f612bea7098cc6c17532693238c2c4021744e3529fddecf79f57aa041016", + "7f39ea0a20e5ba978e7d2804ffe7a9bfaf7195dcf95028e96ec6dbbd8939c7bc", + "34e76d0751029aac5559860d476b857dec12f18820584afb492dd7bf13ef8c73", + "aa90337da124bc8d5a6c663613b9591e3854235cd88735967ce8b10f6aba71a1", + "baa56e3ebeecb601e1807af0f7f78703194f19f6e4faee964126d99675025b63", + "390e4c67e1d8055777aba2e0383a2ffd08db48095e83c547689e528c9e9291ac", + "0d7c96099c6b36734b7e1e39a04605bd4bcb63011af1dae118a6f5f6874554b7", + "33758e568e0b246849d57bbc194fd72ff1dce72c24d021b68cb9bb9d4bac9281", + "d96ff2858ae4ca52863fbdcc56df6ce6fcb626bad18138c2c0cffb66873e157d", + "3f37d264e4777e2a3f6ecbfaaad2179dae1d12b619aa3ef4eddc11c5dc5b9678", + "6b867854b75fb66d75a40d5d7de7787fdd9627a37447bc4e71c951e2a5f7dbaa", + "09db6e42619c27988c97f877ad989096d1f0faafa5aa2a663a0e41d60b16b4d4", + "27c3b632fb5a97d1ad19e820c72f574e39c9302a2576b23965c8b13ddec212e8", + "43669141d0f139e9176152bf54e73f947c553423a234c272ac4a98f0c63113d5", + "a39ed2c29ea96f52db8c86eec7ecd63b9098f0d9419d40586921c8311bbf64cc", + "eae94e4c3473644cff7d80a5ef4c90aafe0a2aa58ddfab5befe181d80aae5637", + "23f90f519142b1a80dda87183ffc4fd6a6a800582f9c721f8df4c4b24f58fcef", + "d3c22df7a989af10820f173b70052d31990ccca8d095c161f453ae979dd8fd60", + "bd8ddbd7cfae217aefbd164ecbc77fff2f7151991fad38ec3a82ce03040d520a", + "bba5d0dc5583a1a1394687fe610cbfb8d4e136e2c5d67af7503ce3e9c59438e6", + "05c01ca8a018abe9c9c51245a5cfe575aef1cb1a1e66c265df530cd532855d63", + "82bf738330f84751621ffaae4375541996b615d33a4bda8ec8cd4b008d26e931", + "864377db1bb29fcc1d8e9b22e27060c3c25de36b8060dee98cb09eb213a97354", + "de369cb9a0860e85e0a3498d3e98696e8c29208e953a31a3f0df3e9c10153ef3", + "bec34421966fe97582849d79076230ad9360d2a3a17af21c84a0d3092071f56d", + "705836a3d2dbab4981d8f156671b1e54c706e5a3a1fa947d79a5ec26431ab2b6", + "95e8aafd822013a705471a91f1fd2e6b9ccfb681448ee2079b6098955e189876", + "debe49d8e682fa253ed94ac89aa38ba32dcee9808d53a9121100e1af312a48f7", + "92b14d1deb272f0e144f30ff5db18a02ae169dfaf3b39f908041544ab5d117ce", + "43a89e569b53fdb671dea115b90594ff39cee38a621159516dbb2d54b320edc9", + "d69647bd23b1402d026e3ba5489c741e081d9e3f74528531b3baea666613f771", + "de0ef822f5b94a70a16d73437f4e9643c92697e102b49f0392329be74bb0ced1", + "3fefe304bfdae2fc7b549c23df734a93b6bdc120424487e821358e60f7109bfb", + "fcc119d507f8363eb8cc2d6b5282713fe6714774e94f582ae13767521945fa5e", + "e32c35d279c31c2441c752b158aa353f46ecc56b78fc82f0fa85d952e3965b31", + "8d5ecb2edcb6823f36a9a904eb3c660168b53d6860aa23aaee0f4d6a406e6bf1", + "db29f286e12550ee682af6a9660e58a0aed160a12927dc1994ba43eb05c87e33", + "75ff8f22bc5934a3f6c27cc6e5b76ef55e8bfbd3ce775a8c3de9c6f4d84cb5ef", + "b853a0e43ad5c6fbadff3f42fe6155f72972ea546b6948d079432f04d677cdc3", + "57decd8b8aff12b0a919fc01bfb7938112b946b1145227b4d51b0aaf378d84b9", + "6fb8ac6effeaecc6a058e23c832ecefefb1be23fd8c6f3984eb11939e1036bd1", + "c518e16cf731bbfb373e13b8da72658a78cb03959ceee7795f2f9eea4f9b3861", + "46b82744348fa9c30267bf87a38f7ef3ced3edeaff689c52e71a43f6d90bba35", + "92fcfe95909152d08396dd7fc40cac2882f24ec71d299c1de56d3226285c90ef", + "1d30402ae739ae34608ae76aa738ebd7bcda885c88e6b33e6e2c307d91e66cd5", + "4aa7835a7ab0fb3f479ebd8e3376162f64ab324587191d49eb3d2ce1e0ec1274", + "e515e25b32cd681e302172ac21ec68afd4112f290a51decd65a57be01bf877c1", + "2dca1b29e5bbcb2f30701b82d176716e480d389959d9000bc90f723e6ff23e61", + "933fc0d1bdb541591877bce18a1e0e3a76fce5c5b3636b18b453fd9e765f891e", + "67142be06ede0bea368bdc91e9ee55c8edee39b68a3a485dda0653b148e1b4a0", + "9e057982e001541ab7238a8a606a2b759e0f13297ea34e312cfa5dab62b55517", + "2cefdceb5f3d6eae90108ef18e155e960bb5d4994c53e93e6d17c09792187069", + "a3eecee2a3d4c1da45421e52253e63965ee35d7aad856807511e481699783419", + "09d2125dd7054df44c5e07c0326bd88da4584b7c4d52ceff8b895fed41906408", + "58e04c83937631522951ad3f25d94ce83f4606d00fbfb166741da921e9e2797a", + "f4625440c0f6e654425113e4033b2a215d2d613194ade44add9feda73b3c04b3", + "e21d85847ef158e35d02243914e845c9a1eacf77e872c2ad404661678df85567", + "12797b2ff5b308354003e525806f33107d77acc9b31dba14f94eeb062c4ff974", + "6e082896bf5a31ceb2b31f2c1529f58ce7ba4d19d3f1fea7ea12706ddcbaf2e0", + "b212d3e8737bb817946fc6218b96f0ed91d630eb0d619bb2c1a70018af8e1e6c", + "a62ac60f0de1fee047d66a9541cf2563ab0d517b2969705d2465334fa7ef0a3f", + "491560e40bca9284eac2a115c7bd0254525c9d98583bbb5948cea74f2d0a392d", + "a20fdc8e1cd70606e3743111a2bfdc4220e239f53ee7ef0322b1a0c8669cd541", + "583f6e87a3dd55ca18736f0da544f28d33fe36cd9de79ad0018cd07ba94dd480", + "38aac22559142454b613bb0d05f8ac0b370dcb384323b797ddf687e39c5cae89", + "8bc73ac1d565247dbb03ca05229a0828a596e92c1796f05491973ed1c1ae0b0f", + "6af2ad3001cba5ca9d74cd87911b7288f2a5871f99fcc1c513874c6028b069e9", + "5b8f74b233a8623a05c12fcd1a29d067d48e230cf519f322ffac667a7cc22717", + "e2b9493cca85007aac45e1d6e4c64c5b87448b1f2452ba0de884f5d6e316572c", + "4f82c906b9b0fda184e3137116e52c746b9014f732217d9a67096922e8d4ebb5", + "8b1b96adbc362f36135b93b1b4ec5517fa525974256d702dddf2fc5c5d46c009", + "e8ba17b0ee878c8a6cee986a5201dbed34b6a8f1add62573cfb7f6c5f085b836", + "06797182e119f2ed10a561dd4539b7e0af3fe1e4be05328f2119adda98387e8d", + "fac021fbc7584fc5ebda8564bfec65f009dd6720fcf12d62da5d94ae9fc327c6", + "c997ed8359257c200e52e8fe42f190e825855cd9dc9a84786a22f83a77f7c80c", + "4549573b755cf4e78033bf8f5c12fb3bb1dae5a5ad788520a45ed96baa63e55b", + "87c815b695f5fd65e6a52662e61390a521fe493e519d8addfe8f1635dd706688", + "b655ad3befef0fd39b426efda172173dd30d45a7586f575a8905354427706586", + "fd50afee291bad6f90ea5528fc0423780d0beb834ceae6f05de3e91bcf201ff6", + "148e0b15a53f9fecab764c237e4526566aa0fb35ad98650c607989536fac3065", + "324b418f179fd7969e13f4ca7e67255abcccd80bf3953a50055105a74a389d33", + "f64346bc91994afe38bd18aa4571f0d34e794301c640e6496daba03d7a64c584", + "4fc576f965c1a3e5b04737d9d92dd5977a73b33ce6bedda6d10e451c1e7e99b5", + "9ff6fe08185f5e2fd241277dd171690e8b1f8ab5c1f21dd8390fd98c2b78e254", + "c7ed8b41552edb8164902344c83573b9012ca73e71ef9e49e507d707e283b5bc", + "834e6a76d713fed9a6fde1d1ff1ff462184836f092f0e26a65af77d97923f94e", + "ff3e6ded0f1bc5ba57cf7713ed276be2b6250c7d78d004610c0724f89797ac4d", + "9141843d407abcba6ca8987d9d3f3e84837fd91f82ae18f4eb7bd6949eb8553c", + "7866bff2427231a9b52b25c9d4fce5a973350bbf2c99c4732a8e0705da23ce12", + "54c7775caa279981fc141fe60fdf2da2dab6e0d71f82e1e264fac9da9567294c", + "de28a45e84290d5428273ed06297e9a765f5fca4ab3496871a19fab6792b9c15", + "3f4bfc805e6ceda302cd58a02059daeb8ec094f5b542c0877634c397dbf387d3", + "0c84af9aa1a8b353b6dd7821e45d588085898a878ebf6585dfbc5a5a136c59b6", + "a1c26c672d1387f0b6d6dd0aae23c5e6ec491a697b30656c88763478001b9db2", + "ebc4a9fcdea66db828e4eddb78309aaefb4add101510f4fbcc6c4d2adcc6e71e", + "184a705325944bae67df0920b5816ac6ee526032037dce2a10950ed53a3f46bc", + "fe618357195a2e990ffdab24c862fa9c75185d9afe0164ec95d8c60b1636ea0e", + "f04d15788ff2fb2695ba8975d4f0ef2f2ffd3fa3db5e80a2a05d324a61c38270", + "aa58e428c357f755357c75bcc2303d838dc99d34f726e3b9639f9a171e685caf", + "9935fa9c50c32b67717201dc9b883417e686754544eee24b554015fadfa22e9d", + "42e0938be484a074742f067d544bf4cd4ba2037d2ef9a969983c1d456a55d76c", + "bbc2bd1c600d9a276e2a50e4269ebcbe9be707a326f83d673c1b18a4e38c6caa", + "84f1ec228dc7c3fe6bf43ea3923082223f135ce064b05613711bc0d6662233d3", + "3396ff813319c6524486c0e2f17ca8440e2da7e1acae7d03f8a006e73d9ea61c", + "1431682efd24fc82af3c71c162d82b1fa220c70dd346196da66b686df7897cde", + "917677723937adca62cb907600c9e9d588d7f41af8395a477c52d9f337070fbc", + "5b3f446ee4e4f861c04f688913e79ff0a9081dbe97a1b2d4c294deaf48451fde", + "962bb0da7a2e86e1e2046f1a73b63321ea268d54e65a1b32479cd71519d4a63d", + "32ffe24132ec802ac7da12ace5383c12c4bcec2845f9375a07c55cc778910ac2", + "063234a96d8fb3bdd0472e4e5dc69bc26df3e2041a403139f003d1a78398887a", + "4034d131dbe5c47ac0cbb81a021cc1a61fedc0a4caba06036e71ab442d379d4e", + "33ca8892f263f1babd3926397f22a2cdfa57175a7a60d1894978a7444142aaaa", + "cc7f12fa75bae9b95f5dc32d6d6adfc5ba56a59eef5fbdd61551e208e7325634", + "c40aa61773640b62ae7e007c54fd2e7518a6fb07fac79a48c659549e97ccf164", + "3d5093a661a6b7435182aa766dc1d488b5ef6b7756c7bb566b168361f810fbb6", "efd53608e97831a210027707ba4dce5c16c696a51a721f4bb1e9b8b297dc33a9", - "29e766938312351dc36df07c44749243c7461c3e654caa800d62d963492e9268", - "90e2ad59a736d6d31968a987fdb90683f11bc14b350ec6149a3708fef0b32975", - "2867b050b4f3f89d0b9f75abc39bdc5ea6e3083fe82184b9c37ad5492cc71ce9", - "fc2dfdf7b4b4cfc8944aaaced50bfee45e8375dbeccbe56590aa41134a1941c5", - "a8a44dce0be536f6705c737906895f0f58031e1f2c1c94f07854da26997d35ce", - "b335f05b5bd01647205a618bd9b849bd7e8415916a9449981e42a508293d13f1", - "cb5bc1062c17c9a2433cc6656d37d756e354690b3c0d1921d506321b4af0c3ba", - "e34239bc64f360077df32003dd9fe15934ed07af3452d8f2af1a2bd5c2cf8baf", - "110bfdf3edf15e2cfb0f763e431d0d3eb03e637ced6d6deb53ea32ab44871176", - "f61e687275ff718e08a7c095e02cf3a7f6fddd3c93d88c6b6acb992d42df8066", + "2b4e04a46ddf2b7d1ebf67b0d0adac24aff8424c44da595cbe1ba9283651002b", + "8dbccc5d31240afad0fc93b29d26750830e490e5de42fb4bb56dfbdbea0a8384", + "6ec02e72bf3cf3621308f8224a9dd8e2cb040160b0ebba106194a304cc34a617", + "a77b5966193cf794f22c76af5b773aba88e4f2d9b88ff8441dbd5abc52fdd813", + "1a12a357c97b8e3461ebf7c0046784d8fc765a0036b3e8352c6775ca89886bc7", + "e89099ccdc15a780938aa8ea514c1f5c3127450335a058455ec780d622ed1e37", + "7506be0bb255471fcbe31ac6fecb7dd05b4160138ba792b7229ed864255aa31f", + "517224feef584301fcaa679c3546cf06b2171d4888167d106174debe6e6f863a", + "aa05083f45788ec6f843fb2d86e362469e7d2d5be3de5df02cc4b029cd7757c0", + "3faf1db176e22935972d4818ed28f682a8237b8886a026cd86980d7f2d5dc139", + "d36a8a980f5481c563c790cef61e63b084c7c07dca0ea87cc197f48838227f38", + "2741793b42ea00da3d95a8fc0d38b1486263d761fccec60233420267f713f1c8", + "518f59a183f42847b2761c0e80401c8e36bbdf8d08f2369fce9c58fb813862d0", + "de98305e187fb66cf11ecfa74e44273d05c7bca194be2a4dc96b9482779710d7", + "3c99bf55e3a14096739cdfaf84c606ea16840158a2fcc0bc47bcde005fffe8d0", + "fdfc876775f40cb5719659c1f4f779ba6c86af7042f2ebbf5cda90dc52459885", + "e6b99d85a9d83739016da2ca6f61e7b20a5e1df411db1dcb2705b05c78508c97", + "55d05b1277029bf9d6255df5d05257257e0e9703b7e448d3987f53d33d389065", + "5559abdc7612c87481eb13c36142b1466013a686685dacc694e81fe415fd386f", + "b268b5f436bd418bf1e12a22fb93c85f84d75f37f2444cd336c48999372ad47f", + "1a43b8007d2a53cfa25a7079413ce3da9aed08da05b544e27b9b0afeafc1c1b7", + "61958eb2cbfbc68f7c2a7043d75bafdf75d7dbdd6cf082156fe4417d87537878", + "282c17d18e0b935379f494264b73dd69058bb8e3e27c51e49e45c194e5e30a1a", + "c27620091c84e44b26c6ca3afa3de7e1ad12d0104e752f4c37c71ea6766afa35", + "5d6807a25c451581f2ae09eaf6aee359259c768821ce8b05f5cfa3acfe64d099", "ab20184d854755ddf48861460f5ea133e4eece934795c2b9cc9a9a2b974d5932", - "d1840b2a3d4899c1fbf44d908464a3eaa848f53494a66b305e49679de84d1b75", - "0479d19db4a161316a2f0480b66e9be82852f6aa9a610610fe9e6752fed21925", - "4846be9705c305a8accac085c2c6851a8149f57ce56f314e480f1cd255e3cfce", - "63f5c7e8679b923332afcfc5b9c681397214a67fc8d5effdcd35c49ec5d433e9", + "63a7d6ca9d1d7b838ea27d80a3f43b44677e244925b87eccb7fdc405bc799c31", + "6e99e2647951083b5c6d0b52eda6f4b8cf85604842633c15f7c3adfb9625391f", + "e9097524be7fcd7f250382fdaf57bfa38c12fc7ccc87bbf5e88f31dc253cd971", + "9d36772c12b3b96276c8608ee5323b5fa2e9bd809024a6d63a885e5400d3a4f5", + "42ebdc3dc91b842d1a43e9374d8fd6aa3f67653f9ddc67a54ee487d896468eea", + "0760f87998604df49fb3f400218935fc53ffff47e3db4f80613f0a6d3b6a626b", + "b0b24ac45f0701762e4f93a62487f9dbf3193e00e146d1aac8ad620fdfe6dde7", + "e8fbef2a85273b2d5caa13f3ab6a207970fa02f160fdf8b8e3ed3d7cba7b4035", + "25a5374ba2e9f7034b2aa9d86e2d9fb14443e61995583a1cb9c51ea85df02dd5", + "3206ccb0e5b53a443b7e410040504333e76c255a92035e32d516ccacde3f0233", + "5f25e097f6b5da52cf04845679decbd2b76ff644145fbcaf661a87b43db57400", + "e10e7cd71aec0d1a7bbd6b55bfd127e1efaf4a92b95f961fc8a2c7af679403c8", + "7221c34aae27fb314ff46562c717f8b3dcf01a4789fefd77c9c68a39635d9fa3", + "8b89c785a22cc3af1ea617eb1d3f3497ba1d31ae5c5e554f7e942a8c67a2dd0c", + "87946b6f8813595d56014abbfa2839c8d00b16c43e4a758b2ee5ad0f74efadba", + "e6a38fc534c3408e08418edd57e66f396d11c0fa8312b247859ae4b678347e30", + "bc8d4794f1fc42994e955b8bf2c8396e60a3f1d98643d069d06ba5789e81d53d", + "fae7cb306d7918fbe864b81586b618584fc80901aeb444ac40e096deebf841d1", + "5a88ba52ef6f987da62f6fe6e15c6536222c999074632772b694f60ec14f13e2", + "e0ba84c2e369f0eef7f171d6484ebf12c14fa266c671d7d8b97ce948841afef6", + "4eadee052430995cded3a4d31d75e89f32f7924ecdf2955b51255e2928bd550c", + "9cd2562070287428a6a2feb709398801d80f02052faa23d7e421142a35bd172e", + "486b3527ea3a96e2472a8d4bfae467fb5701d79e0d252ad86742b06c5e766d17", + "398bcf4d923de672e2ad62c0c04fc3e7e7a8625778bed39580638a2095f61204", + "d3313e7c2d84f510505dfb568886ff052c99d14305e0ba05a613b9514bb752f5", + "789e66299008ecfb98c049755337cfa8a29d84e7f8ad820144af20559566245d", + "d47b80ed8c67b430397e3838703b2f4108b524600fc0f9f48ec864cd4c0f0e01", + "d3e6a6490bb25a7ecdc0857c0e8f912c39f5391841ceb807d987ff2e273a36ae", + "8ad6a5cf4fcffdd0367e9e03e994ecdd0e04dafd876aef142cdac89af0783b3e", + "270283f328606176f044cd9eb3c06ae53ce3c50ac814005dee183e247bdec19a", + "5a45ada76ceaa00fb243d4317e12c42346b9f0d49ef321b8359c7aaf3e92f6ac", + "f3abac4dda0c4295e245210c5f58788d3882d0585eda248cbba53fcc5b6ad6cd", + "4464b3e62caac5f435ff55001bd0658defbc3f0a27198c695a776157d5a1c0d5", + "ccce5d3606a7d81bfe6e0cb84b049ba5475f055e2fc6cf8107dbba00bc23b0a6", + "b638294a95bf50a495ccd0144ccc96dd1e48261eee7408e250a85270acff146e", + "27134697bd39a84a130cffbb5208ae892dd161d0a850bd05c70d5702d21c49eb", + "951a62326f5ce89d27ebe1c665fcbf0d6c000837e71cc79de5e0fbad2ddf3cb6", + "9898ab7e3efddcc6bda9eae6143df5ebbdb7357ecf77584bbad279bb45c349a5", + "ca4ac8e4b2b152ff8ba782ead6cae3b7cf43e2edbe36324bffef99e743d5b5a9", + "c83c06c21a16bce522c414dc29117810f97d47e8bd174d008cb9097e6bfabceb", + "c744442e3c35489d027d9d958dd41d0a84377e4a72d21efb19d8b1579ae5b04b", + "8b3929388bfde7a406bd6b82014e65bb5a5bfda5bf730777fc8b910d12b9f384", + "458828806973e5d491737e0a80de113f116421a30f941263a3fa35e1cff1bf9a", + "32448b27a414496f9790d1b5e2ec62b2c1935bd730a19bb73d41a9ebf4641bc7", + "35813db4d6f04e77bd6f7d8f665802f0f08ab3cbf27d3e6f7247405a3b09ac10", + "98dece54a82e99a5e0f29c3da739f14cca9b0b60e480ab376a593c2629a364eb", + "b5421fe28d8a31ee595864ade1518e5cf4438032588e99b83924c4abc4e68d1d", + "0c3f6ca1b275fe12eee752aa6370da0c2ff58c9433268b5dd4c6c58c9a0c71ed", + "f2abc6720d02175ba01d916d6d9d00749a2effd929ab028ffc35f778a49f060e", + "561cd67cde548e134a7929fea0443b32642f5a4e3abf6bdcf7b06caf6b9ef138", + "335545ca1a725608a057a8562055749bdc19781006edb0c9fd0b304cdd418a94", + "7b47f71cd6f8515cc4f98fc70a7a7c76557872888b1b29c65f6f4840e7b435d9", + "3c3481193717dae696727ef972c35c8874a87a3335f2becf743cf1cde9158592", + "fb303c3bfef90424ed476b57c621593a2d964b6a08dcb82916d4d7b204291d5d", + "d849ef06658060c5e198c8552ca2d8f042054a10b4f13bf885a3f3cf0832ac90", + "69bb9f28e868cb27c1144122312f0f348a77a4cd4b661d2f2a44e90a3f4c82f1", + "5f336e262baf110e27c3ee16ec532f60f73c9b6a9841a889075d915c542faf1e", + "f624a4ace0bba7636ce0946cc90fce1ef4a92a12cae6ec3684a81ca6de3137fd", + "ca6fb3caf9041603deb15395997b96e98532e43f3e18d90ca51d3ded10bd0917", + "075c0fde7b57cdaf3d7cde6fd24eaa30ce9271695c46e9e17851cb20f3f73d44", + "ddc0430b3518638533356b69d58572fb3f0ab5b3fe576122221d6e7fcfa2c3b3", + "81e003e11e2af935e793e89c29fea9ec4b417d4a654548a1b4b699cdb0b75b58", + "c24a80861e2e7ca1b82ee4aa24da7591678b6ecd63e81af56d69ca3d76bf9f83", + "2929a67f8b3d1e186807675ea7d0a6cedf52cd7cbeb9955006db222b552af1ef", + "f999ee53582df9bc97c520a3cc4aca59eac077c8ef8f45ecb15ded9ebb4a856d", + "56fa3e42576e33eb85a27a7c016e3d843775236fb8e259d29b908528aa9e97e0", + "3094fd972cd20c7e7b39773b4fdeddec0dde7ab3b855d17081e79b9bd85698bd", + "34bf74edb080be997f040d14395743ffd236641e9ed3c6514e4648d6c27c57b8", + "04b0d3ec4afbcaa29432eed223d799c6b629a69c374dc4dae16bf794b55c14bd", + "433c4862939e4d9dfc0492957e2ac466ff26b5d711a75eecbce89d84dfe67ce0", + "94f9dcefa72c6e815a73100784c67a333540a2e3a2daafc71e6f6807b4402558", + "2812e3c1f0463a3a606871d667d5ce0d2b5202f662674e41d27be2c9ab767096", + "5c9dfe221ef0fb21cae912c464498887819944f4d3e877859ec4777cdeb1f59f", + "9204a1a2684e7c68843fdf7ae1c60c8be6a84c60658badc4056b273dcee9f25c", + "0392b9a670c30061fefc9ec0cbb00f1affc95c152a7205988f91db6e62048160", + "07380d3d667452adcd2c74d0d992e4cf460e46bffade286310b6c7bb6d615c1c", + "b49c1765721dbe22ea699df576d829962ff84d14f669d50e7ac617a8410483fb", + "56bccb3263bce9a0f2daf055b32b4bc4176676280076d9ddd5f0f9a19a41dd2c", + "1d3ba773d409d98f761f446cb5f2a768b916347fba3c2ffa1c68316403636815", + "12b9058fc11b6a83b47285a89e0afd022b7cf0edf474a8481cd03176e5c45768", + "56a5d304df0abe6be4e67f25dd700204f868fa9bf6e6e50704c4cbacd4b51018", + "df01d242199d1b4538f83c2b68407c4149fe54be2b1573397b182faee1e3882e", + "61b9ae0b6101d21088c5aef0315e8e705eb520b8409b5e09cac8afdaf643dacd", + "a0c7116e91ade0966a782d2885a7f1bb679eb08310f49cf3b1900abd3700c31b", + "4967084d8e28e78f52aebd1033c3e351052d1060cb7f267676f6134ea0c5cb63", + "55b2a6261aa995197f06f0e712d4ad8b74e53f5ab137adf350db7cd2c47d03d2", + "25038ffb1fc32437c18ce2bd665526907371f65f70292238d87b7120d2078dab", + "e8a8e8451bfd3fff751f0e7af8bd027faebade9b00984d2787cac9b75a8af8b8", + "fb9cd9337b624ac6ee90662c1c95457b097bdaa35289f504643430cedd7be4a3", + "9f1b2bad3f4b50d21ac79d8d25602bc3439ac0fe20256bcbbe2e68ae0c0179cf", + "2e3644ed871844cf45b35d5c596e595ab182d34ead1fec514e317a0102abb2c9", + "5c4dcafccae92ff8331ae0d3ed6f9543581a714767dc1a0184e98e2e82990d3a", + "24a2426762783ad73c87c319afecc9f291c1c6bc77e4a102ac6c7072711a7b45", + "a919d561e13c00290fc61c9fb30d19736a0da135e25df86525273845b9eb30d6", + "c28b5ed3263ebebf787529c3d0cda222713830c409fd7f19c0d98b012bd1181c", + "25592e02d01a4f7ff933cb9f13cd390e4efe1d1bc17ec774c319c70629feff76", + "2095627203ef576cbe4362e2051d492fbd6c3cfb386894c744d450c746867210", "610a7643ebef1ac39ff512f747868a1d9cedd0fe95238a907e652c87d11d0b03", - "1b53c152e77ce5a60f9127fe54927c562059b4c15ea27f489a3924a210725564", - "2cc1184cd5051556a4dc1c6f19e83622f8a039cf8b6030b41f330923ceeea87e", - "79cc9acfa9e23104ddccbe2870a31ae43e443603fa4f8174e15b499cea7e9e97", - "81530a5cf6aa551694b07a2fdc31c8cdfedceebb8c12479ed8f9321fa2e4e1d2", - "bce4b2bfff513a10ad20c95ba79583f4a1525bca7df9ae07381b15a34d0ca666", - "032538976faf1968a0b23c7335a6471cd06eb954af36ca5cdeff64866308facf", - "9d59308fb9564e1534d455fc2b715cbf982ccca9e8c80226cfd4c997004a1c78", - "1202cb9d406013c12bbc742154fd6931063f15ef0661185daee5a3f39bfd47d4", - "f4df831c03307bb7938b51c5769863cd013d1769d2b82671e86d2ae10d3d7e79", - "92fcfe95909152d08396dd7fc40cac2882f24ec71d299c1de56d3226285c90ef", - "c9e665f7bd5a35f874ae1dcd909290ef6efebc93316663c2bbb23cb7abba5960", - "6dffbdaa4f8d6925d9f914ef866108a79a0da0a17565882d48dbf82078f50f10", - "157cad9cbebb5ab9928d04a09a075dc98bab3389d7f5346b32a964aba240ae02", - "f9f60a5bd3d455680bdfcd91326b4a027ff945cb03f70b74e342e286faa2ece1", - "38aac22559142454b613bb0d05f8ac0b370dcb384323b797ddf687e39c5cae89", - "8c61320a2963d24be9f8bac774a4096bbc720d645e2d7a8fa9983ff03bc23ec2", - "b212d3e8737bb817946fc6218b96f0ed91d630eb0d619bb2c1a70018af8e1e6c", - "6cc733a91e5005a2645d80c4ccfb928de564264aeaa79c50e2f041b8a0de6b03", - "3c8474da48ed14810198a82c7ffd47c53bd203f5ecb650785b6a75bb77ae4a50", - "81e5bf2c593e46a2307c29884c0457e4866ba032aa8e3db72c79a47b239e2cd2", - "8524f0890b1fe5b02da21c0c2dd0090ba032fd116c6c463278ec5a6566d5b266", - "86d91473025938c807205ec27ff85d3a61f562c851933a354967ed8f167bf4cc", - "9c4f9016dd88d842682a24109d96c521123432790230b1485d481d4b92cee096", - "5b903e9cc75881e86da0e05369f2b957669eed40731620829365a909d26d5be9", - "71777c70f635001fbee6a3e620ce7597a027b3e5e314bf433e66a7e247d3df17", - "4b6acf9aef315267f00157645152d2d3197a845af60e9d9479e364c3a2de9adc", - "cf8abb67c74697d90a3564a90231b5326fec8a302d1a3099825e68e7209baa77", - "a60812e4793858f7414f0ae57225234979940792a0427863504b4f004f9df9bf", - "da986b9f2ea92309098f8d9f12aa2b3b145bc1a1d9eeb09c9ea8b10a072a22d1", - "0dea677023ba50e92aa3f1e55fc2a6fec4e3da9834d09b657e5d27ced0176b45", - "f0fc7ef1f73f0a86098bbba5c56c98a49c591cd65e16fe1de992d684c2f0515a", - "2f9645a4932d0076fa554e33bbbd380c591a0243ebce02f03509077fac2ebe76", - "a45d03a298f749bff5c387906020af808b22b4769d827a5c6dfa50153127678b", - "fd7ea5c44cad5f018b92744f3b27759bb3517c3eae58f4b33bc82097d1c5e3de", - "31bcd13472c505c703d7a923d28574e5014cb3560bac3a48edae0766042952fd", - "7b0fcf3083161ceea0b98a21feedf825a98f7a755bdbc23993de777c0e99c9df", - "6f5815b08d89d62be2754630d5388459b4053459997c3d44c544367c292da491", - "072751fbea0812e36a3e532cc09dd3120c25ab8dc0170e70404ad5fc009e1693", - "8e17801d61b50bce2a8c3003c8371ad657574e58ef8d96ded613be264c51a468", - "ffd35e4701398f156ee8ef2ffac95bb992d147ab77b5b38be081b9bc98014d1b", - "5511ebab7d22527a65d60df6e9e2e88b0c19c443f223a819579e88dcc40cb4b1", - "c9cd1804f78cf13ef256477a004686fd51a48e5ccdf0671966a7c6be0ae77ab4", - "a1c2f7c3f9cccacb1d0f49ea00eb8247324e4bac2e3ed308f7bb1a273236e263", - "314173851fa96753f5bb6951d249a9b1c6cf83fa1a210e0bc7b5b8f39211c0cf", - "b460bb8adf4ee0269355e084bdb8a6e3ec9c1134d8c811ed5765f329666f7323", + "6269aa1d4748cffe56823280803998856b26e2917c80a02530165836f7b9a23a", + "5792eb516c475943995c62b1bb643437ef355bb81784970c0e1660ee537339c0", + "40be56c16fc6a7c8f8138bfeef81d50dbc0f6ff8bb6d984c205316b21b99e993", + "cd6a29224419c7df5b3c52318f3a8583b0803023c6e71639a9dbf4f6570a3ad6", + "9ddd033ee68fc267d4a44518a3154fe33f80c9c14ffc283c846dfd6852abbbab", + "ebaf2ec0712a29c8ca23fef798fb53bfe211162eaaba625421c1afdd736ed770", + "4b7cbcefb2cacf2f83456fd2bf1cca08cd7be8bc2f0f57c46038b3ceb71c84d1", + "ff119dad9481dd0f5d3bed028a421dd8c2a53b7687344d6d0df2e20bfbd86b2d", + "9319d403198f947da7e09b2a9e135a0a60b97a7618424a0c6f4cc01b2b562d0a", + "bc0fd8ffc008eb190ccc402e4f385a6a3d6693c1102d9bfd3b0d9fff43733cbe", + "14720694c87f435a4a00a6fb1119cbc2e38d6b51e558d414367e416a918897a2", + "7d7f9980e26f7a3987172a00fc607f1556f0b0c1eecf24b478faf0e049c3578f", "f665669184d0305178bee1b54f9853481db52c137c2f2ba614b5e956fb19c025", - "738ebc2b1d82b2e6a96ecddf4a22c70755a0adfd00826274141eb6bb63275007", - "1bf2dcb0cf86271ec16bd49677e92ce5ef638138bb0f0f402102c1ea344b91f4", - "99090a063026a1dd5f8cabf5a25f0899aa07b204f5fc56af409ccf49b45d054f", - "a8a8733d03367f04f4d1cfbb528425a22c22e32dc2b093b15df33282d810622a", - "eb9a93a9c28cb7527b29d3ee78446e9b8723641ee2829a61dee517a4044d1e4e", - "8732d8fa859081b547edd67fe5ee50eb941e3cb0ea0e32c64d60d89e85b5a096", - "d85b85a18419a9a312ea8a18c79970990fc5d8ac99dcc5170097a45dc409faf2", - "7229ad59feb041afc228e5409011f54645245eed01d83089cbfefb5dd8ccf43c", - "cb21e7a422861321ba3b312db4b69437dc73e30cea350ed4663b14c39750658a", - "e85c30227d8f18b71974f69c65e34a9a2ab8f088bedca50dd9999e5d407dea5b", - "bd9b91071aed1382c97ef285185fb2f06a61b0245422fa85182d0e4c583dfafd", - "ea4d97a17288ef38a32b4e72ab103efee785ffef7fc7dfe530f087afd9501e4b", - "042c1f9c29251e67c546726de59b0146eea48b1c63f4ec9a1375bb016cbdf4c9", - "9b139d300af8cc930e467f66f3f618db7c7609f2b36fcb1084479763b77f23f6", - "eb96f3faefa1c2553baa13faab0b053ed947132da2f85c9a03b5e0b530d6a29b", - "ef6e96c4326231289c33c465c9deab43ef1eaa50d1aecce6a2d1ab3d0f4e12b7", - "58c21ea36e8e1313f573e8eb5f51a53245d053f9a2dfe397e7ef15590e56cb18", - "702448520a4a7f0bf04af188e496a14c7e01ca17e7a6d92df018a9c245454256", - "b7f33fe5debf2125c1609afa1626f52573257d50990c93c05d4e748eb8b3a4c9", - "bf5817fd48fafb09e9e21b04a59ae37629ee66a71680b7f33cf5a8c313b30de9", - "97411051eae641adbd84b7886c5f8f9db55b0c3e16dd8c12d1d007b8bb5861f6", - "6005896c10a80f03a6f61e36c72ffd19902b4b208da347b9b9eecb99b21623b2", - "757be0a7b1bdc02716cf1a4efb4662d0d83fe26043d8d8b52ea398e0291f19c4", - "65675ea12979fecb54bfe0e17385adbc871bc009f6516e254faa37777f62a531", - "5b02154b74ad955c6c76eabbabc8c8ffab621cff346f4d510d8c06d684f1db6c", - "8609ddf74ebab4ce881a453ea3143b515c30cf10c76665de47dee8a11e1ab00a", - "e16eb2660ab6ece15609de00d45266fdd1a51d0d899ed465bd8aaabaa5d7c79c", - "0d7c96099c6b36734b7e1e39a04605bd4bcb63011af1dae118a6f5f6874554b7", - "85830189b4ff56ea82a2e20c7c9d4473ed6d4efc8fa40c68a8df0b762f9e237d", - "024aceef1bc0558e2cd23877f93ed93be14ea0cd4a0ee75607accbe54ac9df0a", - "e8a55372cf638f92ebc1351046c6099b91bba6d5178633043bab8538fba8fe7c", - "bd73fc9a6b897113bd2dbd2f0b9b061a46166050d17cb5de9a4893f47125396f", - "322ae8fcd1f52c7ef116a5d5437dee0d66dd650f570f1d6b5d0a91a3a334ad7c", - "2d2747222c2d8d3bf87c61511a076594eacbb97139716556a29ee4c351a84830", - "e8a8e8451bfd3fff751f0e7af8bd027faebade9b00984d2787cac9b75a8af8b8", - "d071bbc56f778de48c94a831debce62fbb6f7519e47ced58a19bd9918a1bacc7", - "3c406ac84c68db46767e393ebb47d61459f837e518e07f30dbc7843c09854697", - "ea2ca122316c3136bae3b182de8d9196f300a663dba68416b9e20dd64c9fcbff", - "22daab4b38c0d2186fc466a987abf0189c20bdcce7cfbb1920fe7576d0b33102", - "a4b93e84a16db5127adf03b941ecee90ec757c81e16a23242196c673a9619216", - "9140084d11a07c8d2044e2f637cb0d9a002a09314ede4a7029d9f55ad9320f7c", - "b4f77faa8f1ddaea781595cffb56ad9758b9fc468cf9304f0a3111a0860605da", - "f9d76cd2d7cfd4d55edaf59d799551f5e0bcaaa0213af475eb13f75194d0cbf0", - "e0a68cadfe1a51aeee8c447e7ca98ca72c72927ddbda35c1c1a296715569af09", - "8ee7bbe0f906a6f03982a8c397dd90d3c3da8613ccc8c6fda063b330bf0b8301", - "04652b84e58db3305c4e059c7ea1b078ae6764857ed3b4f19efcf12357c1320e", - "c518e16cf731bbfb373e13b8da72658a78cb03959ceee7795f2f9eea4f9b3861", - "ce5eb50592100172e8a2788fc81c3a8bd41d88c99148a3df9ce0007857e7732e", - "ff0faf3a6a636df0f3bc8cc5ca6083e4e3c9b865f5590f8af06491cab3b26a93", - "5e987177db0934ce6c64cf2f6631a50dd3f9f3ea9fa5d180886792814ccdb769", - "876ded4c80860a64d7c325b9a6fda5b3488acc8d4fc2cba3a1f4ee37aa25300c", - "b96bec437993fdefb6db5be59e21e7d907934f734712b8828f712cb4e37f5896", - "eda124bc784f8910c592173d278aa7505a32755d919d9916a0ec78d12081aa95", - "9037043add6f24b50972795b56c27f0c3a4774e917ff137a7cfa6ed15e28165c", - "237f6c155009780b1fc8cb06fd208e3efa87eaf2e2082f7b56e9ab2066b06f70", - "015af204d3e97acb562ebbeb830548f8640a53806cb297bd46803edf53eae070", - "4cd21d371c8193ffa06b50a979e1025c224a52c3213170b2a7db6f2e554f92d0", - "5bcfcba38ad6d0dcfe114331dfdffbf5686dbc6d10e260473ced52486540fbdb", - "fe0e112810a0e5235a9e343495c2b6a0dc83ee2d1e6ae0e1b88264f7b366d500", - "a38b87285b2bcc53a557830ab11eac4e26cf540660f757e2b2abcdef1e6887d8", - "c463a52bb7e61135bef48b5f6d3ef2536f6ac729b733ca57536cc07bcd32675e", - "d5e4f83d7db68b2a1c27bdb36acf5fbeeecfdb5987f63ae5b826f6385bf3be8c", - "2300dc89b535430105975eb8cf9a0acab0208eaea958b3ab434060030e1422c2", - "b426f56e2448461184b98eead368dd239c11d291bfb27e07a680b49299a85a47", - "6d3200631d0960c0ca633c2f4d251c42a8cf7aceb4ba6d0dc573ca9f79099f85", - "34162a412be1624ee18a8c060b33f6270120390d1845bc0ff047e90ef237d4fe", - "348f79800e4fd872172de23cd6e4b8e7d2d5828d2f2ddc429ee6c6e3f4a1e6ae", - "3aed9b96472a4c516616cfc3479053ac31bfc2b4a72ae6ba930aba7420479c74", - "94b7359252dfed830f122de38417b6fbf91c85f54df951dd5dfb68b45336c1a5", - "3f9b05941242f2061840c298e0a1d7a88db27579d57b561eb8e1c837e82d9a11", - "6424c899b1af80c8189db7d4af0c349553434034340d5f36706dc0f29b79591c", - "9dab5ce1eca4f038cd138d248416dc76578046a0f626b081104b1378dd2516fd", - "23064ec6c4b0989e493135790b1db899fad302689b44738e78d8378aee05dc0a", - "baf5419638cd797784047f7e1b852fbf37f6e6535b34441378f2a3dc6e4df247", - "1d995f2b05f10a60901f7001e45bd1c2c1df9fd0f31f5b5af8f876ee14415839", - "a21bb69adf78b0d5eddb3397ede7615168c8b292f14ab3f8dee7496f502ee174", - "89916258c879edc79db0bd39f7a53590a1372033931e07ec281f1b20ac1af485", - "fdf19cd9b066c634722e1023b0e8df814c3edb71c8d6afc4e98f05d19707d92e", + "3cbebd9a44f719cf2ccf6e5650d7645573d3a84cb71e5b2bb7844e0e56627afb", + "7f13ad908ca71af64f12d140be615abc767512416b3de1bcce1a50d67a32cdde", + "f6c1c7d4a5cbef7cb4779122f75160daa11e374aa575bbcdc5373fbc95482f54", + "d9266d5ddddfc261bf4a4aa76262f61a5522cf14901b8b390514a5c61cfcb260", + "d8ad4bd47250696c2b82d9b0bddfaa728c646bc95ec5d337c7f8eb352d436797", + "b1ecf8760259dfd23b9dae16b98bbc35c190f3d9f5836a439e92ea80f9ae9eca", + "00426d75a29d1c412a6b2a3a5ec976889b6e38028c96917006d37762dd7cb6cb", + "314173851fa96753f5bb6951d249a9b1c6cf83fa1a210e0bc7b5b8f39211c0cf", + "e646bfc7817e210183c536fd889ee2e07141f88ef88c4d17a9083914d6187993", + "d61e4989090de493bead4770b40bfd33667953fd7722e8d7efbe85c6bf4bae60", + "2673b7a69ee3306ba3514056926bd0ac428acf1a558aa817ee0c30fae0594766", + "4326adac27e9832a22005b5f3de855324865cee63e49d9438ec9e25e29b06e8c", + "957d1c54ce3910861ce1a5992a9b019e38b5c464e8fdeb6c148337f1f1c1738b", + "c17e868fecb7869b51c43cc1e5efc45a667152d6e66d7dffd72a058460d0a54f", + "eba5bc627b7cff53863ec9f591e97afdedb26e88bb6c5c18b07990e90f299ad3", + "4d34cd964c2ae0c2e012390310382300e090d5a8bf973740cfeba6233f21057a", + "36874fa3221042d9c3bfe63c7d843c31d6e7f6ef24a6027d6f9e5fc12b73af85", + "532cc76c50e353b9fafdc96f930b53457b5455c51703399b3d589a66b667d14c", + "d13036eb72f7959b04a11f1363bf7f7f5bbf9aab41c51eba8f5d7a39f44ccd4a", + "55762daab6c9b8d33c2717a96e354bde6804011acb26b6e08a6385eb4113bca3", + "c1d91e14c53685414c5529e718d127993dea72d4a5a10c8a222147bc5ed43aba", + "690cedd12c61d4dce66079bf779b7fcf4e334861b88eb55c1f5c24e38fa65b2a", + "70df7abb685e75511b7cffd215d41d3b25160842e8b44fc828b0bcacbde747ef", + "bf8bc382e1276bab144ac68cd1aeb3af6ce4e991142b9e0ea3f846ffc520c3be", + "61ecdba01886166aa22e2a13858c1226a4c43bf31d9a41905df06fbb0a736235", + "4013ef536dfb0a07f9073ed8db72735786bea9962232dce2188801bc4a0c341b", + "28b1359dda04f05908d8e7a0efe9661f356a4921eb8a9602fc0a4bcf1df5892b", + "8757e7489e132db55d4e9600b6bde56fa9ebf7dfe912348f3838284e4a27bc01", + "2c8f2543ff2524a923d7f2a75de4942e58601d085fcae7e412c6685ec6aaf729", + "67944e37e8de925bae505c6ce67ade18794efd81f53aaab485b7e2de4cf6968b", + "bdcce4735a48542a2ae75077feeee1f4623f249f051e0e264c104f12f797df94", + "b06e9d237d18eb9169e0b55a3097c691c9e36092be38136d376cdd10aefa1066", + "671d2bd1dcd098817f5a1ed88435a3523a6d2db748e682131ecd8ba28447ad55", + "9e66cd9172948b23a1ee2907dcc4845e0c0cd11e6f19999383a79d29a74bc323", + "90d76b0304dee51fdf1193d50e88363969b974b0612a6f65e5819dc124657476", + "0774d23fe7027c1d38d457977c5994067922be4c9a2bde6a53f5e7ede62dce1b", + "e68e9d74194ba5da4f23e884f4e63f3dcb683367dd44f875d69880c97ca15f4d", + "133df08343b7951d5a7fa9d28679534d7474cde41256413e332687d6c8437dbb", + "f0fc7ef1f73f0a86098bbba5c56c98a49c591cd65e16fe1de992d684c2f0515a", + "06b7b945365b9bee6a62876d1da4f9d299d242b2e9fd3d35a4b8e086502469ec", + "0b2d37ae14fef5337ab5b40c5e184d7d4cbfab769684ae77d51687aabff82a69", + "d4bd8f0d3d07e8c1cda4bc6a7de857a21d2287341bfff061fe8959d958eee455", + "ae0bcc01927cc130f8b82e87259d0fe87c76255e2841921fcd761e64037b4d0d", + "de19d832cb2392ad17666bd4387d3075db7efa8be55bd2c46bbde522d0cd65a2", + "97456379a258283c71a75af0365407cc3d2f770ef0ea912e57529e06c0775f59", + "d49d6bc3e289ad34fdb828531f4190b011711a2495359788edd413b6d3f59458", + "d8bcd57fea031732c21b16c298a727e2f2bdbf231bce45eadf78bba154d4235d", + "10288f4451ed128f4f931e73dca81fd9cf08b5da38f015ee55f6d0fd74b82c2a", + "c38df1b688b6d5dd3adb5ddea0826d01f74a4db3b14ddc6a46a2188dc999863a", + "6f13516614e491042e8f16dc1f0cfc3619de73104a787e4e1f88aeefa83ff5be", + "38ace304486f4e038c48cb60fa9de7518282cc7ab02decac7c05d74bcce94dfa", + "0d79dc5dccb95a2e79d47385df0d32c70299a4a9652e1fdc206713b588eb70be", + "2254d413ca0a2905ac1003f2e5c22a44bd86ff5ad34f286c51321044ef819466", + "40c3a92c134512ecc49f9209b14762a21f34f4eeda745e528ff5a98c9089dc4d", + "3b034b4c05351b9ec15ed40c4b6e8977f22c9be80048c82767c9c95b45abfff9", + "1d83fd45d621911550858e7fec25aedcb5c7e90199a6911a31b5d4e93dce2ae8", + "f95abe66fe2016acc061f285c3dc991956f469f128cfc0d9691408d55c9eb529", + "d9a9f4881bcc61945c859c1c303822993a682e057e11d9089d4fba9470f9d060", + "2f5c94f47d24d69624ab460655a4495f2f08b3bdb5543546fc85e72e657d68ee", + "ab235dca27588789e90f329d603a28fa6b49d45d137305bed67af5fab361310d", + "b98c2c16d04680916406cfadd8fb63d67ef620f728e2e7488d58906025c24e19", + "7f10680ecc16485f2a26371587f3f6bbc86f254613952be88f64964320e2fee5", + "760754fbe0bc1ee2798fe29fc106bb074921a329e19ed4b03e3004b529468a7a", + "72e8e6756ac3dd096a104862e4fe1cc864d29054b600196084765de2f5e1dce6", + "43c96f25677f8b74dd80c9d2b654ca1066ab0a1863961687470239c99931efe0", + "9f4539002d46a10b0eef35aab32da039d57418c6fb95a9589bcdf56516fa0436", + "e40e84236ad5fdc15dbbfaa437b5f468db928d5f9822b98a0a8998366da17656", + "0305998e95368bd2196052ef1ae66b798012e4fc067e762b6d57cbcea3049836", + "74dc064b4d8e0d092ce832662e7cc0baf3e861d00b82f0cd951df40f5d296ca4", + "8da43691bb8d512ea6b3fa2d6daf492acff1c5c3a5d2cfbfcbacd04f97d16603", + "aa369fd4650abfe0ce9e02ba4cd2aad2878eabd1fb9568852986afe244ccb7af", + "6b39f04366cb47f226009d5302ddb464f759c32f6767de6e3857ffb51034885d", + "d02f57acf266443e1607861723c754d3dcf00bed7d37b8a28c6ca65bd104dd3c", + "77cd45cd40f3649a0f8f72a7dfa8f93b2334dfe3f7a22f2a0c3235e182596401", + "721f21b02dd679c24137a5ff3d947e46cddb045944476388ea10165f99def001", + "b1c2dd22350b36df12354e7111253f0a5c987d0209d9c7a6138d30f65d4cdd2a", + "6c4fcdc3674236bd4d9ebb3846f884fa22ed9e3fc925d6c557cd55cbc12dde85", + "a9b3b92cc9b17a9b95ee924b086ccca0e3c5407c8ee3cf26ba63088018109a82", + "d522044841b3770ac78c2f302fc12d95d761a05648df920239b37f4b6e74f3f4", + "333287f8aefde06dd8da7bb3da35c204eef699a8f13eec83a80bdc460c73b525", + "0a0f010c89f43d2ab02cbfa4f59ec25611876227abd6a5f46018df4e856dc057", + "67be37f9686497af408ece78acc8f586baaf4bfd40952a963d760e67d439b71a", + "a067d424516b2b2214987648b480094cd6d91896f8cbb655684bfed59a84a488", + "e63aace90dec2fe14702645cf20ae03d09b672dfc0c352a84709349e124093cd", + "70d7f245403625f44cf0c7ac5db4bc072c9dbabf5d884a10c22efd43447a7ed1", + "211a4177c8a4a047acc6146ba4940f224fc202f8e1333d9d8864ac277d4c3956", + "511912cb54cda24bcbf6c29523dc5a7d9d4d116e6206342df72dac5ff4d0f1e7", + "63c15d231ca1b2318315c563f38da7c533310a2195cf4e5b41bee4b58144f4d8", + "6bb524f2b9ec9fb3bd53abfc0693baf55ab1a9c3c60b6d316cdbf532a3df60bd", + "d0bbcaef3a232d96f6609f8ec93abf778c0579c9533f2f9be1c006d3787f0ab9", + "207c98949f7f8103ec65527a7e9eab4b4a37df29f9824a0a35a340efd7cf1caf", + "27ce55799f044670c7963f4755e91b027a9cc256e3f28b1a2368e90f167942cf", + "6efd1c991010bc45a8de242d0e8de87cd31107a781b353f2560920a02248135b", + "532909a31a3e88c339d3514a8abbf823c3e44f7da73bbfda6f45174a4ba7a974", + "8df02049fde34404f1b038cb42224b97b3b42c242c5bf018d1afcef2f999ce61", + "5e1e1d3fc539eac7f8e4931a68e605a7fc030afa60811f5e184cf5086be91c2d", + "1202cb9d406013c12bbc742154fd6931063f15ef0661185daee5a3f39bfd47d4", "18632f478fed6b48eb501d6c5d8571d7a901c4247edb0412c33713b122369111", - "3e92dc809a90d7cd59ab0159b9b479c3116070a32984d53d1b3361561a196b53", - "a5b04c6343e283fd159bcaaa11fb7c94fadfc52a001c5acb615c8cd86ce7f511", - "8412078e2370fbecaf64079e2f23a49df95166d079863033a5b0125e3c8205cc", - "d9781c3db7fe42a793fb9591dfd63334396126ca6d71c13c16a004e9e47b7954", - "f1237c22e019500747e1f63dd82db1cb3e576a124760a337997234f95d71871a", - "5a6e8ee9e1b05c144d419404c973855f8bdf086246da8c8e9e65ff79c7f1fad4", - "dd6163cd44bbbba868777214e9645985b5582e2e854400bd626cfb17773f8cfd", - "d5c613177a6dd81db689e6e85b4bddb53c212e11f1cfb0b2cc40ca28e56c21b1", - "57a9fbcd275a825eb06c86ac8f1da874692fdc29893857a0bd7519e1bfd80ea4", - "372b1859e673eeda87c31dea15dac396f62bc8905a17191da0aae5fd60f48fa3", - "cd3cf317662c1d934242c0d0927e8417e64ab879e591090cc4fd5287ac06b87b", - "da0e166c86288f7980986f0cc30e3539e1792dd02e31f22204d253cd35e14817", - "207f416b25bd3f9932343040fe7a32548f67e99d2f07d5aebc36388189b85730", - "c9084bc1cab17631711f813124050a5c8d779c92b991251a5e24a942e7a5fb78", - "b577cc618e4249e86ad6a4a17dc67d2d37e95e94e588685fc843b8c4db0b5d69", - "82e747b875a044815c9b216dbce3b28359bf2af340527360be5ed7af7ad40988", - "226e91b4766b7240498b7847b6560007979b27dc306ff0392b2eaa0fdf860ad1", - "ec50a3e6a729df31340b01006f9b857b9aa0b78ebe47c7897345e088542ed737", + "9087263c9e056a56808ded8f44470e37950fd1ecf0c2c1f782f8553ed4c54ac2", + "66705e69637139cea90f83cc83cd1adf31b3ba5732e66340b5d637f86ec94b43", + "cde7761a3e48a795d0d8c52c63447e2fdd608b617cb65d2d9b2422c35f90c1c6", + "bbbd9a94dfbb944a68d93fb2562da241a39eb4e5448000989d1a671610e087da", + "38f6a33f4852a31e057aa7e9c3e89f64ae317c3c0cea52c4cfdb288e8be0ba62", + "c7d7af3eaed042e1aeb38be6d067c5c67cb364183c10ea39cb00572645692b05", + "b31fb88995f806093509e6d2098f90ff86f3238a82d488edca79a8aef1b56d33", + "0d189d6f42c5c200a49d84a26a985b1fa9c47e8e6c05e88408e1fc7a338aff12", + "5d9623382c2e2b29bf85d592cd40194b6e8167bbf6da5b324ed693f3e39e24cf", + "48f5c2ae947a806cd05499327a1deb02fe29e4a1c6c50899ddd328cdb58a5227", + "0e8b644454b64abe6f5769f6ee3379a5b950680641e3bbecb7a3c924e95c4a10", + "9003bd9e1bbdc411567adf4c1968f771eecd38b5f7e4ccb29f0c48027642698c", + "c17b17b088d85b709511439a38026e48f94cb3cd10203c98497ba3b1256d503a", + "82643e2cf113d3d09975df6305c5901aef9d76b8c5f683911d0755d867c3cf9e", + "16837662a59b7e266eefe41ce9b064c0754a3c3d0ce3435319211028d8bc7291", + "3be9ec5d4b5b4eecb463f2b8cd5dfce4537e221e1aa501d4d28dbcde41858d16", + "2fdb13806bfd1cab85bcb617d50b867fcca17f993980aee3834aa65d39d71580", + "b6b49e763adcdf923bd2c89b1eed5921967a4cca036d7251480e3f0ed3a38311", + "a98333701d40a1e0d0ead1453169903917c90f8080823d6b754c84c14fdb0f37", + "fa86bd7b33c1c49d94d3fdc5b700412b579d4b5738537a3a17f7ad6e3a04c717", + "bc32cbd8f13b4ca8c10d9bc7844c45ea2cbe43c77683c5835032ae985b8c4416", + "b0a2edd99fa658db14d51cd92a9a4ec1d5849abc118a08ad88cf656eb055b609", + "3af7f5c312171733a190dd3bc1178813866ec1533268b6ebeac9d7304a0e4dc2", + "448f5dca54a90f4467a0610ca7fd9cb822b0a30a6ff900e2d45dcb1946891c75", + "b01d8e1be3240cd6a4955d234f98ffc818f6e17ff1971e0c534bd673cea79adc", "f1b91e604e9598b1c14fd899c947ceeeebb5500b32729b540553a3f982c46174", - "3765770ce777d224115660b875d62cf4343960a6f06d4a60001fd6ccd16c385a", - "51a874fb6958a29602408c1aab307ff10b0eb1bb91a182b75e32005ebb82bacc", - "fb8b42bfab4b49b27bc59bc813d1f58777a43dd895e958d84f329a386d1bf55e", + "346e54d3bc00fd51e70b1959b577bf83f0b10edb7def1e72db944edf4bb63df9", + "19be7de188454953a91457a0e30a3a172c3f4d9ee2afcae69a268e48b489f79d", + "b1964cd68a531b032ecdb8275dc8356969e771d0c01ccf3d6518942b75b9ccf2", + "359990fa3bab25d9f872c0a4b8df5d9d4ba863362b58fd16329385bfb8d3d6b8", + "1e3e9bbd462fd13db31c0b715d3b5d42f897aaaabbc9ce83e9bf19f89ea96738", + "58b6ff871ce9c78f17cb7198182f3b75cd00f2b3e7c4b6aebf3b881638e8e0b7", + "07803d6dc5f28a4fb0c43f487d36a44b522d764b29796714537cc9a40a130a7a", + "98f086d00082455f2aee8c399858210b6fee5bd282cb27015cccd21dbcd89a5c", + "dad85807906e40cbd2096846638aca272d975d111cca7da84216d4ea87471a46", + "42879c6a4b073a1737578d03db402f23d5b56ca9f5cece221fa54bc39087e22d", + "75597f53da1b5d8a95204d7391f71c8559e3a38773030f0728e22dd0d706ff84", + "65675ea12979fecb54bfe0e17385adbc871bc009f6516e254faa37777f62a531", + "2c3337c0e109772f3bc4e59639a045a6aa35726b0a962a93245e6ae0fc0c5267", + "0892ee129e02a4fd1a32c7fe5b9bb87cb961708830d58e451486c6e939023367", + "0e706d858e30988102daf4944d765079d17b6de19015d65a5dedd1978c0ce5b0", + "4edacbf396765724988dc633f72d6c61e7e2788bc86a6f3059ee2be4c1a172a4", + "eb8b07f5b9fb1323838368faa2cc0fdf6e818766965052e59e1dfbcc78ad57df", + "9e6c6674abba1645943b708a7f63bb5e502b7de02564befabfd24fc5a42252a4", + "34ed4dc1b29011ed08a14b9eba0ecb1229fada1d49e61a2aff556cd60940bb9f", + "d72ff2ca3f139d0cd6e68dc80f2b0b6592dde8423ba951cb0caf32329eb0765a", + "9858439220e720067f5c8bc3596408b364152f8bd7e82e150135c32c8ec2734d", + "1603f4b161dae20b4a56d5bb3022a4b7ed9f757c5168a82b7b4cff09c150ce69", + "ebf7e09d56378b8c94f47cb57d93875c0a16389694cbbb67ea7ad7bae09db98d", + "6d7b009f1ac86a0b3a83ce09d7ccd0c0a83ce42bc9c4132d09a9f13ef8579e3c", + "d4aff4810357ab7908294a18c2c23482b806b40a51856fa30c6fba1689bacabf", + "27de329ff632f4afe880d157c53de17bd60ec2d69ee6c8da20acdb748360f539", + "eee478bd5dedcc2a39a4bce31a256720f2c54e618ebd733521c06f851acf55e3", + "3cf2a389125272cffb2339c62e1c020db98e9f8e55ad90488de0443a8379a879", + "c5b257b83bf6ae009b1c77d3946da338edf7bb55b3ba51fcccccf838ef07e2a0", + "97abb2689f1a4af79d3a269cd57eec3062cb22a2994a5fedd9922bebd9907bc9", + "9a81c4f952b49170693d960b1ae33da768af2a7a6dc7c4b49b357cf8861bd2f6", + "553f8bafbe0683714fe8d5b6a45d96f577fef5001da7f4b8443e20d62df3b94c", + "9353e984348c3514f0b126d7fe33d269dacf5a32e6cb81e26c466954725f004e", + "46b3ad431877f63ce4f012c661f610b85af47bd20bc666a1d0fa7db6ca0c5bde", "7b79adce325d711fd568953a971599805ee5888eb7c7bb009aea4cd32053824a", - "fc526d03d31721f35824162e067de36d8f3f16dd96e67302ce86c8496a3fd6a5", - "37b2f913ab687884050525f2aff673ffc514a882752b84a54718323d54a85d38", - "3032c3e4b7491e233eb29ef960fb9d1c021b852659ac338be985c42a0ab44134", - "b9cafa6fce2dd9877764ecf333bed12b3a246d76dd429bb0e55355e6b889c0a2", - "18f05fb4071e6c5c527dab37c8b1eeae1ab8f2a557510f9237a926d39549c0ce", - "32af066aed929c0df85376f9bc2ac54164636b60f30a8919f9f3b4c0f6f202f6", - "420ab4588446d418b78af045f6da21153136bbd3a7e8d62096a49dafd46d787c", - "3c540d60f96e69c69d2529eb2b56a00591e676266956f09206a37d70ec6b3ec0", - "b309686e0deee7d71b62acb3b57f83a175e3bfafd50d1800591c665b0d8ab8aa", - "6c4688fdab5b02b2f655be9ea8e443d4bb68457f88f782ea30cb23ea52ae197c", - "e74cf83472e328da791857976f4160e8bb42d085eb4d491066f68338aafbeb07", - "ca52bcef69ffaa66b7b61478b394685cb962d09ef93663f6f26867ae9c115a28", - "da09bf265fdfce40160424d49689e2c4c316c2dc40b16a9edaab74ca18cf3dad", - "a2026f06546791f072c50e30685a75b27f9e9cbd86983a22375fd5b65e740553", - "26a206b0cf397170e33d3f0563a33bda7952d013869ce393e540173e1d4981ce", - "fcb12276e0b896cfce624abacd3ef907461b26d58cf39729bf29192ee2a8e881", - "02a9bda78c94720bfc01e89a4663ed819acaf5bffdab19cfdb197d6d1f9e0059", - "9baf527ed085df06f1a2840a9a4e1f4f98cc66a351aff6955289398242da53ff", - "e3489db5707b0c54c9190c0ddbb0b0f14db4c553429cbfd7cac55f6c70e4ac98", - "2d1a44bc2d8384793a430dde24fef47f28e49912180c001f8693ce1dcd153484", - "d2c60ddd81aecb03dc2fdc525b633ad8f028aaf6c90fb11219c2deccb08c0f85", - "e63aace90dec2fe14702645cf20ae03d09b672dfc0c352a84709349e124093cd", - "3991beb23e136444b9ccf2b0a9383cc86f6fe8ea037186d421d30528f7af5a16", - "511208e994f79b15a177919fef038307d6b221d57ae685f3231d8fc127962fae", + "0ee730a4c614e415a783eb03f783e4ec9f7d2d33dcd310173144ae2fd106ce93", + "09e2f72166773b1dc303b01aa56629fb86374fbcbeb195aab647460c4b9c6a91", + "1b64312ec849ab2d646860a52a83bc7b05f05d25a814dedbffdc875e5b5067e9", + "cba33eb9a3530e201115319fc0a26d708f61d0170c0853a267620eb4a73409d2", + "f8489b34e405a4d8ccea5553aadf384697b796abb0e25098ed68db3aae27c673", + "ee48c9c9add2bf460080ab5512d535caab8a522b951624ca372fe9d41e7c653f", + "f0a27bd84f139163a87b6f3401350363321faa23cdaaa3039c4a4ce468ca773d", + "bca4f36210ddf5c0231b46fc504310f97db68d7d826957c404696cd3e1da6cd6", + "39e0468bd8477c470a858954a053a56f249273cd93a707fa466b0756ab4a834e", + "6b948e2d252c32dcf5640134dd7cfc3f59ced5c1b9cddabc47e1582a6dd15ec2", + "67a1cab3f097758d7cae6728bc4c27087b85b8926fa58f7b383bfcb95da53e4d", + "f763b125c846a5c83a974a3a0d2640a34c9ca9c835e86a26825e3877c0568b06", + "f461a1ebd894e57f8da82b55c08a6f8bcbc78407f04daf266360ffd081e49eb9", + "eea130f592749fd3e5c71d3d77b1efc60664373494ef3dd35055f56bdd7061f2", + "db13a71ffc4126fdf8c7f416412fab45cd8783a9a8f2815dcbd4538122a43e13", + "2a0c05c539d82dc0955cc63f1d367ff369748d6a4f056a5aae3990e7703eaa52", + "193f3cb2085738ab9f6c7df378d91188efaa58fdd7b8e99ed532d55dfc8383e2", + "b3df925b8541f39163c134f7c48a95a05c3a2c79a9e63efaaa25409827ca7e23", + "32e164cc7e83bda976b93bf063e2960c0726c684ad4d7aac84953d8d4c68e73c", + "f991d3b9ad54d0fc4caea25f41c9cf9413fc3992716ed3beb5fd823dad131a93", + "46c7e8595f29a1da4f072c032b8423197827115c88f1613732ec41a3f9a40eea", "070e9c5aa912c5c873ef5364ed6b64a5ae95322542a6828c501e69ac82d40384", - "a0699e76ef27c3226752b29ab0e40756877ae0529dfcdad81adb4963f7108ab3", - "270283f328606176f044cd9eb3c06ae53ce3c50ac814005dee183e247bdec19a", - "c6c41bc10875485e273153a9bee98b02c849f0ba2e1ffeb98895173811d33711", - "bb3127491700fb5618363a6083f8328bb2f9040e6cc42a7fb026998077fbf839", - "6cd5531c5877f98573508771c2ec16c8857a3a87dc2b955512b90f6a0435e38a", - "3d0ab0e47cea53b9663e485d2b2847202047c7cefd1d125c22f1ff60e90bd48f", - "5da8ca774eed34c384a7044c9ccf0eaafeb6241d52dbccfb50961c12565c6aa9", - "63a7d6ca9d1d7b838ea27d80a3f43b44677e244925b87eccb7fdc405bc799c31", - "9b7701fd448d20451adfd5e0153bd2194da8ddc38e2a82b52c0787311be5f88c", - "ee3c8b96d827c26d78166204a3957ecdc44144daefa6c613c1df557fbdc7c063", - "1efc86ec61bea58711274789cd2e54fceec96d64c6e206c3830df953b7ef5ae4", - "ce4410c7d84b348f5c314a06e99b98e90e86f760391166663b1886020249f740", - "afda909ecce33426d0430b92fb8995d47b000f9be04ad3e0b575cf17443ea3f4", - "d4d214e62dbadb2c1fe1787531f64765d4e7062cb2029f3fc7285144983da78d", - "9e530113f6346550721a80fe90268e3ba3877eb00093397fd6f9964adab28dc6", - "6ec663717658b669dcc49d13c06c9ce4090b4241ab7ed0dc2a53e07590a61160", - "c1bb8fcb195f017bd96e376ff450ba4212fca15c87b638f9f8fba4bd16f96d1f", - "93cc3fe0655106c227c17b5b6ab301b1cf87a86a4a7d14418da27ddb79029340", + "50ee8706f6fd15f71601d4243010069ecce3c78a8389350e53901b226954c645", + "d650e0d49f36c8119f407392ff2665e252411833e3aa328fedc0b50d193d53ea", + "c925910f91222e4b226b1ff45f7d53f0a651a7c5d813a083d5f89f24aa83607e", + "bc7e7b841c1308b0f1a3621a1ebbc39bad3cf285275838c0d1ed61001217f50c", + "8e584aeb5d88f85b2354d3856538d0df0850cc391f0ef027cacfb477b4cddc8d", + "f1bf7fa75d823800f7e0a193a68c41e4c612d2b73bff9c6b153a401741f0779d", + "edd7c79797d883a9da5317b5a7453d1142f128975c852ecdc8c66758ed57181d", + "368cd876c0f4e37d5962518892e56da78d8480b095f75bed02060d50ace91881", + "908df8f772daf6d1cf68cadee6eb23e6fc671de8315e8e7bcfa9536862c9b45e", + "48ea12d5f1ee0be6e44df2c6fe7b186f5d7ef3237c5aa21364e66f2ad40c3c70", + "5e987177db0934ce6c64cf2f6631a50dd3f9f3ea9fa5d180886792814ccdb769", + "0d568af745d9490c9ab5937097b6546eb54595197a198166dbf61b0df35554e1", + "48a814e09bcbf44b88f93a0b31d7af1add40d8079b738ec2472266ba9745a324", + "c7af07741ff08b27e24fc895a3ab7de26cc6b3984d90db4f6ff1b22471492b0c", + "f87e304596519dd26cf10c919c0f0e9665b9758d63296c7f28d373ea5e7f2a0f", + "37b5b6c155e330c6973ac70c21a3877cd74461337be108a7267e58e2c5b87788", + "14b9ec7da7006702b0433f7c0ab2f8e60c67fc1dd42c9d17974310dc8cd6aa9b", + "b36d14f63278f6baadb593e2aaa646652e7d774c2bf15d92f656ef0532f535f1", + "c6f2202fc37a22249b83fc433fec7507a4ba1912f8df93215f7fd33a9d9a6472", + "23f192da585414f59d7f88cb6364432b05daf19c207f89f996dc19b14bd20c1d", + "8118ea27d34ccd2524e115fb4c64d177c22ca7dbadf6e819220d17143ee31901", + "f7da3d2573a3735642071a909d18fd04de7838ed4f0171b18ee9985efaf24fbf", "541fc2b9cde46a47ad045241210f02738ab53a3e7bb9374ee23b5228361e142f", - "d1eedca8499de6f4d73710d6bccf76aa6db59d7731feb23ea385b73bf0648c06", - "bfdd48b376d9e0cfa9d68d45cb16d8e39d28f2f22688d6daabb67c42cde6e700", - "1f393f07c5c15eae56abfd7e04267d87fbbf8bcd752470e3abd53b26731b52f4", - "9bd24f31f4e39aacb59d266fca3b82972d866d0123a68b108dca9ba76ea8b7b0", - "46c7e8595f29a1da4f072c032b8423197827115c88f1613732ec41a3f9a40eea", - "edfc0900d70f8f4821b28097b76990c9842c745ec4add474705947db6022f8b1", - "cd6a29224419c7df5b3c52318f3a8583b0803023c6e71639a9dbf4f6570a3ad6", - "af1a2c5cf9fe94e6e245d76d5c8982107cf47d7e8543c02c4cc7b874af3ab424", - "7489b60bf93725ea7299a4b7c94c71ecf1a4714bb9c7794674dee8480ddeb9ea", - "f002c8e390756809403d3827a0036d55c66f8e93cfe40a15ba7e8fb1f1f15d53", - "8757e7489e132db55d4e9600b6bde56fa9ebf7dfe912348f3838284e4a27bc01", - "5c4f05b0d045d4a2591afa88a2c00a10186585a42c06a8cb441d8b70e6d9f429", - "1b0dd831ecb7354f0d69ecc4f123a3de68a2e27134d9f8d9a8915f89378f696f", - "2e716fd0fc2bc249f8f40c24d0b92501fe3d61d07d39a52845d5bba6145b71a8", - "79cdf8a595968435de99880e8d2a3d1dc500f1f29bf33603de018b9854c5925e", - "7ae3335f2cb6c7c880cc355e8913bbeffa3d50e444b970793f87d221b3267809", - "14bd4256a12a89c61a9fc367bcb287fec2efdf603221a480e87891d352f251f2", - "710492390d5d3397b8b88dbfc7aeb2287625e7fe27e7509238a9608c9d11c15d", - "083702404de1ae39ab3c70ccaf9ba9cf8207eaa949bf5a9b0a951b7406df1205", - "5c6defecace46dae102e634f2cd2238dd59d079365c1a3b1c357535b567ea503", - "b4b3e7d1a538644cd9ad42a3287f51b4867e008405325fca5be1aa55bf0ebeb7", - "ae0351af5923ae46c2d2598ec20a52c25bc7b7dfdc78dc852e98d3ee9d19b118", - "092a862e6235d2b43c1b3f9269b3cc8fee10878bc02ecf583506aaa58a96ab6c", - "7d3767832f41dedabbad756876fd154d7993be065bfbfa2ef7e494303654096e", - "4b8cbf28c8a8bf6a74d34a79e6d7c2a45a6bd50df8645153d301a4fd4c9a8522", - "a62ac60f0de1fee047d66a9541cf2563ab0d517b2969705d2465334fa7ef0a3f", - "1c1fb31649921e3c8b72098bdc8b002bc326d7b96bdf026aeb2ee05d1608a6f3", - "7e02489f48165bf418e5fc3ca8818a0fc4fd8f761bd97b2857269d3938352220", - "36494d09eda9fb276c571ed74c4f6b84d416413fc4cce40391a98e608a637359", - "48d932e6490c66d5afa21344eae8531da1ee945d10f5fa483e8b9e30c00f5e49", - "b6d9597dfc4dcbb66d9e5943eeacf97d18313e08d02ee2e53ab8e39ba5d1b359", - "d91bc4c3f65ff54756c1deacba13c89813911cb0d5d703253cb44c25d4d629c0", - "354bd9a91c421ed16ecd8803f2bd8cc44bd8320c9979e3498550b84c023b4b92", - "7353fb43813b56299c02e9a82fffb36c8f8f27596317be64ca8fdda9534282fe", - "7cb6b1b5d81d4b7769523d80b4f1ac88da9e278a73b601ba1eac1212dcbdbd52", - "7e14154397c25a5a42ab74d264132407f71e0a79dfbfeea44456804c783866bd", - "7e2b9f0e249d770510be908feebc3c0802e4a95effefbe742548232035e0ea5e", - "175130c5dc414f388bae187df1d6ed73e75aef4e21a06bede0d6325202d6f116", - "69bb9f28e868cb27c1144122312f0f348a77a4cd4b661d2f2a44e90a3f4c82f1", - "06bd132bd3074330583a82e1f5abbd792576255a46e2eccc8580e31dff52d689", - "90292970fa5d193e009c27e3258a1e89bbbe91c744507843798c65da147fce92", - "343ba95eb76fa9432273fb52aee8c70b89297b64aeafe9b4a47ccdc575fd2bd4", - "99f2555ff39bd86cc48f06a08482436c0f18905a897f2553788b50d76e5ef8fb", - "1c9841e711f8fb661529c71692a0b784db4a82d9dad6525ccd6ad1e230947c84", - "271b5ccf35b22c4fd72a3686079f45cee22b6ffe816f8b47eecdc7fa34adf91a", + "e85c30227d8f18b71974f69c65e34a9a2ab8f088bedca50dd9999e5d407dea5b", + "fc530223b7a9c0d244d952ee82b6e2c81ec843864df3faa62e4805e7ced32a6d", + "15b9bc0715be412f86dd4b4484b2927b71b7d2cd6c15d40f05adfaca23304205", + "9e2c3cf0808c10b0580ac8f34e38d778ed1dbf60b524923d27c86a4e6c334df2", + "db188ae5e279f0bb9da9cd3a91b33737558b09410ed6bd4ebcc89a22cb63b976", + "7ba34f6b88d9273cc7a42ceee4055b7bc45de32ac6e70ae940b919b6706409a9", + "2726463e88ceef55a7019d3032ac21466eddb7814e70fb325766bd93e43a5b39", + "b557c096bef4cf6234bc3b7aaabae69cdddaa8733edd2c495a6767c5c0438cc2", + "bd505afb15257617bb2661a5f5e7e2f8dc3a46024290c801f386058e840a9a1f", + "db354f25cca426605e057111cccc8c9d3320aeb6a75e7dd6a7db06554aac764a", + "6cf417a47511c4c84d4e058ec86e21c29f9b67a07f5ab9a82d12cff9c2cf79e3", + "d82eb6c07d080df5dadaccc6ccc9c11cb3d13a14945f2743b6d2f48fd8c75ff1", + "097754afe79da6bdc469f8205a9757a563193dbf1a32744be2fe7ab080aa6e9b", + "ca28fdd42c96c91190ddc7019cf0382e8004535aa984c348d07be81e716c6a5b", + "54773439fdef2f43442c4c981c371137b5abf3a0c568f14bbfe5de2a0b7beb95", + "77023455d830874172187b7a2957b2027a500cc89b7709e6c12cd06ecb38c64c", + "e5c26c11601f18b8999c006f66c03bde01798bf2cada36826c16f1e9697e1334", + "20b782f158ee101452870a611403e4394144ceebe53e9e7ead0a61d9deb0d06c", + "8f610b8755ea8d53034bf51b1c2371746309c446f3e54aea686ed8d37516207f", + "282116335fc95b639485e764783d12725df082c0a8940e486dc8cabeccd02873", + "8ef74484130e2cb8999b2a9c05bfc06fa6830623d586090d23f1ab01b09cb788", + "eb9a93a9c28cb7527b29d3ee78446e9b8723641ee2829a61dee517a4044d1e4e", + "0da77b5fc72a1390cbff36d44ceaf1fe8bee0e23e36631df50405a05d9f5545e", + "6b7e2b0c8a4a7931c0710ed072178c4d240e761d7ca251b2a87ef25a90572039", + "7cdd65686ebd0275290c41eaede2b5df9dfaf841020b05d47692ca00d674cf58", + "2b413c44a51df625f905d9cc41e7e9974f2c052bacbea95c82c7d179046e6876", + "79899ce078905f509de3006046c367908879c90132561fd598dc5b275d66b7a8", + "024aceef1bc0558e2cd23877f93ed93be14ea0cd4a0ee75607accbe54ac9df0a", + "6ff3b0f8bb45db897933d0ff909c555badbae3aef72bf3406b7ca677a07015d7", + "a1c2f7c3f9cccacb1d0f49ea00eb8247324e4bac2e3ed308f7bb1a273236e263", + "e315a948719a0ba26859e7e07de36456f0d19b0a82c45bf949f2c96bc26c265a", + "fdc964dad4f71e5d8a74557a3beaa4d4dbd917e5e79a19bc43114a1ec3276047", + "690da01f2918633c85fade0e95f2b7e1ab0d61d9184bad3c8d0301e2349f7526", + "5d04d837736fcabb546a2f1f1cb0040be8b426fb1c98bf0ea26bb6f77d7657bc", + "7c72206dc7d17cc99ad35d11119ece90ccb60852ee6f5193ea85d5d42114794a", + "8a7db7d27bfeaec1b58e30caf38180e47dc512aedc98e27e96547843212ada50", + "cf3e0af1b115ff1f9efe2c7677551ca1bf378b7317f29ed9a3eeea01ed0cfd63", + "c9429304c027e0007e59a865d111e25f88ea1b546b8878f9508d3f721e99eb59", "c829590fc7d3c8d2911b3b8ee27834cf5319588b50879ee89bf7e2788cc46e4c", - "505cad2ab8be3e144cec7fd8221b61449d58bed3c610691563f0d416e820a811", - "ba6d56cbdd2766a0531c01f1ab6a8c443b4471fe1a8073f0657d4d47d6b68879", - "205c88fed281636189be560e8349ee1987bea84b2a9e529269db0cba7f3c125a", - "88ead0496d3cd487eb7148beea5a2911951299adce6f8c1cd0f05f68e9575dce", - "24becb95a31d7c407a39ad8ddac5a95ab444c2bc205009a09f5017d8ab574fd4", - "d2ab87088ac93793b59567c9387965251e5d7736f873499a609841e756b2574f", - "b6058ca9eea7c3b492bd9fb5c14588721dba3d3b481ae15de643f737dbf325f2", - "84b44d3a2d90bae12c3d9246bfaa4fad0f397b6b4f2350fb865137524c4dce6f", - "847732560938c1c93c84f6324652c3c6e6ff908b4fc04c55ce4154864b93871c", - "75f275858c12d58188f7e444f4fc758986822bbad995b3ea6cf2a7a313c4e4dc", - "cb18f10bcae23c7904002c0cde096afa4a7ed8851eea827834a69956777a7510", - "3e0c28c09c63a512e521ea290a980c1eb39e00f97248a823bfb19f09e942120b", - "9a81c4f952b49170693d960b1ae33da768af2a7a6dc7c4b49b357cf8861bd2f6", - "e12e7abf11e1b22719765d33153b8733aed335b2688d6947f9a733484b18592a", - "5610a113f5d620901cddefd3360a6547505cd4224a62f6ae3651fb26e12220f9", - "9fc741c374d68d4fcbe77bb71d5de6cbbd96104dd2f35342ed0b8a0dc3b6f7e8", - "a015d40a8dce5919bdc43e87a13ea92863317702f9ad243f389811999f348f41", - "f8087586d04d525df612a1bf8b5374a9e6b71c34838662391014b46ed100b8b9", - "2162a04d01ccf7d5840133a6c91cc5f76aea91b9cca0026140ac7f7cfe9de91c", - "43a9f6227e0bc5d47974bfb1fca3a5bad87f883ce4ec19a08075faae23fe4983", - "143d95f08acb7a211a728ad7b21df55df873585e7f90400320eaff7779d53d62", - "ac918c4415019c14baaba4568c6d8af5230582c7c9b5d9043903712e1dc3b611", - "e17bc8f65e7e8812645d7bb8167407389e49dff928d4620f6348087c2835f7ab", - "957d525e8aa523b7fd6c4236f66c554fa8add51089beaacb7f27dbbd3f725fda", - "5e9fae8e410801ab00657d5d72ea593ae95c94591dcbdb2d236d3e7f3519195e", - "d1bcf3655a4841ca3a839bbaf1af31a93d0d79cd6221afabafc2905597837bec", - "eaaaafb45c6b0b2c0bce09d13a8c23f09a0e6ad6a257c130984374eede76d653", - "b257e9ee53e7e37ed2dcda8d2e331a8828316e08176398b56e51aa37619a362d", - "a1fe8fa11ff1f0fc05c84a63d6315fb1583e39e7c6bd2c76991d4754d79d3512", - "8a6418b4596a8048ba50ef2a98c0e6a75c5cc134575253ab799e6bbf733b404c", - "243f13277ad70ea74a80e37efa3fe1ee405136282e08f8b25ff780e8b9b58be0", - "e8d6c1d9bcdb741f86cfe80ed9f7466c7fb752cd94e1879104bc568c1bcf370c", + "660e5c6cc56d36ed5cbb1388677e5de15975a9215678c9128b07ed37473fd27b", + "4d3273e3d504d5da278368e7ba52842c9eeede959d53fa0fcee712ee74d297c0", + "08bc3c26320b0e82a9df44693652405fac35ea3da56254939aff4a51a32939d4", + "32a1237cce674c7464e8e2e83cbf16bc787c97b0494cac4e81d01625b260979f", + "f2ea44f0231c7632c785abb5a2be0e1f6e1dfc73b6a44db0c5951bef407ebb99", + "3cceef37080f32e29ef5de759f5c540dc5a72cfa8909299114219a3da4cfad55", + "5d3340804060d5777275fefecca181c42715284692f31699b62b467177ae8e0c", + "27b152d7b5d431ac1dbd1dfb5001ee1429fb3a4f4671e365353dfcbb9dea195e", + "252b820322257f42dd23d2166fc0cf7dc9de1fcdec6b924b7fe97bc114ac9cae", + "328f36b5f14d0a808e5c0d989fdaeb2adcc5a664cb295b6813184aa1fb0cf041", + "c48ae7a165b7c0de4a5a6edf5a99f103ceb26a9339781285e0f1bc9c24100d7d", + "aac611f34cdca1f75e946318d764083ae5b96b89b191ea581f8eee9c137fb2f2", + "3c21149959213c42c2f513b184c37a32bce305779242c33b6f276ada10723b9e", + "fe93284a9a8a78a7ae76ba3421db6e1853f479f12b2421b7bbb6a79f721da04f", + "521cc0e7d1b78ff4455a5a3a1b5a86d87bd47cfa1915d065346a6effd7211d3d", + "404cd76add6111a906fd5fb805931226377e40defe1998a14def0b42e661fd09", + "f35813b067f5859e38027977eb0451ee57cbc2ae157f82c0214e3b5443b7c8ca", + "105ed1cc2d250590e3d0a982a3ed8e30e71d71208200debe784c040f37001b24", + "41b55b89297030a93aae34d8d41cc3f87f3989a11a2dd6df5caca23b4086a4ca", + "4725f7df694af17733dd18345b1493380904f9ddfc11dbcd0eed079b50cedf1f", + "6e6bc8a5966406be8edeec6dbcba841c561022e88c03347e0425e874b6277ef6", + "65f839a5d0520df781a22c6ada300604f046e60a5164b8455860ea8e449199b2", + "8b1e60bb938fb8881a3c588b57ff3b2e1d75c7a7c45dcb26087e09b4705f73c1", + "195db5d7f120c0f4e29a25eedae3144d26ee0b59a0226fca99b4d7e165c47b0b", + "ffba5c20cf564c9171a41d986ad3a399db3d3868868568d4d3c0a8d00f3cbc3b", + "52a6adc52ea90a2638b94206b1c8c38a06e760d709b23019381617e4f71c893d", + "dfb9f83e67828b6a26cbd44c7c23aa8f3a9c3234b14a634170f55d1a046708e8", + "eeb0ae3b46725e4cb2ad384db3eb9a3fcbdac0d116de1c88439b8dfeed93ba71", + "e04a627f457b1de5fc11fa9e1757d223abb31f02e943d1c8f4e43292ec10533f", + "93b78ffcc209f681d0cea4c4034064161a98cff0a0a57c31819f3eeb3588d41a", + "c48db90508c027dfba9704be510a279c99b12c929406948ed11224e2949045a5", + "f6432e3f742614638ecb3c70473e055f923d018533219f15f23e18ec054540ce", + "b94025f72a4057dd5eef9f5e4461a36154eb044877c635f9fd6ae66f79fa9a29", + "9724fa20e2ba896fdf6f59531211c8234153d1961e323ec546bccb0fad7ffd2d", + "6dffbdaa4f8d6925d9f914ef866108a79a0da0a17565882d48dbf82078f50f10", + "3c494b1155296ea46f0302c00b31dcbb10a52bfb255685446210cd9df1c8a261", + "fe5b10cb7f9e0c5d292fb2eae8c960e63c41e669c863d01979b52dfc00829dcf", + "ac295d008c2de3f9f2a253b9283ddc9f91a34371e249d1415a8803abd6fc4bed", + "ba22ec780336edd60350fe23290105b7e010ffd7bb578b585b214297a11c64e8", + "6e7264f8903d3544b236de167a11e023eac3830c8cd72c51b6c6f3bc6f71a9fc", + "eaf450c1da3b04a3c82a763d885744a7d762cb9c0f005050961b500a9a7f862b", + "8882e6c57eeee9dfd15a3f94c592dcbb74fe56cfb655957b528b3b71ae9e5922", "a484664254eb4ba0fe832c2cfd0fcc8aecd8a2a45b4b4942e192804b12de9287", - "e85d9fdca576181bf2dede813da7d18668270d98bf805f379d17a3b84b7388d7", - "eb8bdb80388925aa20c295cdf8f2c2e3dd67951c965e02c2ad02e9471bde22ef", - "14127471f6e2d62f053da35f034b8f16b0a926c96ff118fdeaeec691501e7ad9", - "e7f0bb9566bdbe7719b821d72a7892770f29d93ee7280f247d627dd193fbdb74", - "1e8fef4080381374c59f7649da8248974b6e5dd9e59b9bdbb36a0b9341b07d1f", - "d8bcd57fea031732c21b16c298a727e2f2bdbf231bce45eadf78bba154d4235d", - "82c5a2a7a30d43c54983cde90e4898acf1b647b4d8b623b3ceee851bdb8ebee7", - "ff1d80849e9ce6105a54bb043b6604c497dd2e2ffd5b6c05aa32f42e27a12fa4", - "663e31a4d70b0b531966f64796e9b488257d50a0a81c38c98bb43ba32fb40547", - "34b0f20d0e814387817d78d9503b8c596939468fb5b0c4110f37d4ec6a74f3f0", - "d84ba38899472aeda4ec230310b8b0ddb8f4e6f64fb732b68069ac14c86160e2", - "a058a3b9a845e7760eaedc79f0b2ad12100a92774f927881d27ffdad3b3a62ed", - "324bd470c85d4cbd58e7ca22066b8ade8cd385b55f69e632252622d2d048bb78", - "af0774bacd97235d890cf032208a048346e7f4f3a7ed3a34bfda605240aa2896", - "aa369fd4650abfe0ce9e02ba4cd2aad2878eabd1fb9568852986afe244ccb7af", - "ff53b33202dfe350d5c479a15ec20e3098be6a10f0b40e6cea62d56363e19b4c", - "2e397708a07d2bbc60f9a8e759b516371a3128a4dcdf534b1927ddc53ddbda1d", - "e685ab92dfd122ff9db0646d5812db3cd7f085a2cb0420323b426db678405bdc", - "541a69bf11776b9b6976c687986492dac4063bcc9bf46d4b83e03c9bde1dbf3e", - "d69647bd23b1402d026e3ba5489c741e081d9e3f74528531b3baea666613f771", - "25cf4445d7ea45f80fa8771cb5a7149144a02a7104002ada2ed460bff8c7c48f", - "7bc13084c7cd4cb116131960512064232d7b2cd175acbafbf26503ee82865fe3", - "115d6ff512a7fa750b97d733c45064e682dcdf791e373121f0a25187f1aa9a15", - "12017fee4171b14e805640acc647935be4a93090e07b18851ab1336c2d7fd909", - "d1c529d7da0660289b48dc11a39be5cec6c56e054b084890a3e0c1c05270ac8b", - "4cefc9bfe97cead2ca7ffdd9ad07247e4cca3b5302f29b1163b8d37a85355517", - "7f56a79c3824cf8a9c42c74c20bf9ae9afe86bf39cf1d63c6f867d57a805b3ae", - "4e1c9c0c91058ad4da79304534082fce0aa998e5125a8b8fe3682a2fd46cb289", - "d25649cb87a74d9cb1c1aebf44f4436a92dfc290616e9eeeeedc91f13ba7c3cc", - "cf8cc2d666d2e56309ce715836d4e0f25ca04053ef974972addaaedbcecbbf3d", + "45d204576c576b23014fdad862d38f99940cfefd0030d983c1b5241335f7e574", + "86ada0cc77bbe3a225668c79df4bc9b9a2af637735644e11781ee0c77524d9e5", + "30294b256a08e0378cbf2a865e64af544376052b414076da932c3915687c5cba", + "86a3b0f2f5078e646f8d5653b1975a038d433d6e8667b9575460a9d8ab46056d", + "f7eac24360b7de4af466d8f4bf1e7b7ca39f6586499053ffd291a75cd3a20763", + "39c22e3d79c3b6f3bb6740ff59d9429875d6dfcd2213cb5411c381495227777a", + "c0ba489968d7252dee9a17d3063e47456cab78414e6f4d38d005d7ca776a4691", + "bc5e30d191117c11a255a94e39be251f49107f487e7c0ead58bc06410a3d3c11", + "2e984b96556887e21dfce3f3d7dfa090421e4200fc7b132eae5e08e34eb370ac", + "48f2f695348cad11c59a5a6ccbb4ee8bafac09b6fb54d1e071229adc2f5da62a", + "a45d03a298f749bff5c387906020af808b22b4769d827a5c6dfa50153127678b", + "f9235d2b134f216682371628cd6cdad06c67281df478ec9e23f8b7796fa5ba82", + "fed168fac77c2bcd642a334b13ad3bf9662e33e21842c1cf2c4b421cc525f300", + "61f4a203bf75640c0c016ce7e759c1873d5fa4c7e34f4285b2ce1af2791a14db", + "9bb843c926ed83ea2d0d6cbfde711f773e69ebe87d14506e0e3870dadb74c72a", + "faf22f91cdce1efdf462120bdb5873b4eb4435cf086e42ab4eef59ba31f1c58f", + "36ae25d9e04f64d1af5e8f798e4a8d6dbca154c05b728acf51ae903ac897a7b8", + "4cdbb6940681cb8955440185764ac041f6837e900ba7548acd988528a70b8c67", + "a2641e76b8b0fbf058e2c059a6c5a0d2f9dd770e3cd7420becae77a67db03511", + "b2087795cecf51387443f3969346af88afef0665b5c7cee03287bf84d5e07c42", + "e74cf83472e328da791857976f4160e8bb42d085eb4d491066f68338aafbeb07", + "99090a063026a1dd5f8cabf5a25f0899aa07b204f5fc56af409ccf49b45d054f", + "3d658430861fd03f8dd766e76d1a7f0bc16585b60203b4469c9737d6f2ac86b2", + "27359ad258787f522c5181a6c248afb363d2229cea1e57253e82623add6bdc41", + "f0d05efe08a21c1001a9faaabe4764363d0c481f718693ba7e2aa80d0ae1fd16", + "7d66345eeea44b9a8f5362a63d77d4a8a21366e28166e2349e79839e10c7bde9", + "bea243252ee502c4532bc8a81e094cf9f12f9068b13313f05a647b8ef8c27bb3", + "9334a3b8cb2506e1524e0566a1ca5d7638859a862341e1c1990ea54da578fcbe", + "a93e7e4b2e61536add921a5ee47c79ba46008cf07ac70bd448415b848e6e280e", + "bc5b5ffad3f505c7b96d6ae39d65926d7f44e607832354f129f50161c2e21063", + "51aec379179c2aa766b998f5326658983d5a2bd5c993f19e183ce7e99fa3ea66", + "929628bbe8fa3f0673a7dd36f139fb5df9dc600469372878101958b9ae43274a", + "24d31931c24da00f533043baa5da2965cfa274b579f4482a0b39b9c97f4e1c0f", + "90b9309d570798e002dad71aa273635d3662cf5ea4fdf133a00197e3e9d46dcc", + "e54e7002e271a46adb933c4cc1e42b4e47aa9957d45dc3be5e83a3c54c7dfe76", + "f8156c94a629763840cd98abacb3a27ca19734afb59ec8ccfe323d0f058ace37", + "24becb95a31d7c407a39ad8ddac5a95ab444c2bc205009a09f5017d8ab574fd4", + "bb501d1a0b6c38f49bb38a96f6b13a901c9be954dea35917fa096c0c37fe33cd", + "77e2f5bc42c6bae9403b4d017aa17cefca1f09c4581120debede0e03808e64b3", + "2cac5802f9fe03f1e26b29345b0cfd04f45ddc5a6b0c73738984861ae431a96f", + "0b53da21a1c3af5da9b3c41d6bfa12e8aea8690bb84fb90d4f604b11dfc7927a", + "699d71086e4c7b324e2670266f196de73c05e88206aa64ee9d479ff85e6d0250", + "642d2bb1ba1ae9656f4cda48bb2957440ae694275c2ea1f51e3072d75cb3f0ae", + "de6449b8499bca4d5ae1fa907f43482574fb33248642f9131dac62a551b8c744", + "8dd68177ab03811eec567cd7feeceb4094e97a4607ee5e9c7478d50b4223f9e6", + "ddeb93e5a1a3192e838702667a3d0e7f6c5cf6ea3f19f474378098f4ae5735e3", + "311f8c3daf14216cb0ece44bf9ad4ca2e1b235ad2c4e77f69cdc8d2848a1b60d", + "62a526fb39696bd62c3157fcb85508bd11c6d6850ea188ad61089ec6f0036f76", + "9396d67ef20b6ba42dbe104a48ff31cb63976b13d369b980671e86c410028135", + "32c614725a492ed5bde80dd94269fc090f14746346eb0b46c298119d46a7655c", + "bbd089dfdd0f8685ac11317abd20b75e21c7fbd6517790ab325b84e146f00e0f", + "915c4f811b88a03aff22bf4d2e179bbcd51496e95f2411ea79b6cf5a2d4814bb", + "c121ae8cbf8130687c42b5863395b29ab23cf23913af196cb2135ec61374cb0e", + "8b781522f3e233efa32373ced643d616c11c401b879ddd784fb69a5d5e23283f", + "9cd624cf3071d5863e5bc4304ceb6f4f363ba4c8fb7167733406aaa678f0d0a5", + "0d1f7a9527cf559d53756901c885fc24d1d3b76285cfd5c277c541cb50fe9d56", + "d2ba0bf747087ac9e32fbd254a51f3489a4c3ba2ecf678dbc1e6a8eec6af00f0", "fd4e714d89cb1699918eae41512a75d2a9680ce93ad9261c870ebf2af6dcbb19", - "42738473419953c4753042f5b2993c91155b6854513409a745d0992f92af7304", - "b176095a0af26ad8236f74e9cb3616bafba12ab8732c264812b88168a301291c", - "ad690af3ce7a22ee7fb5638fffcbf7501953f56c7c3b42337867a3f8f110647c", - "d849ef06658060c5e198c8552ca2d8f042054a10b4f13bf885a3f3cf0832ac90", + "645fc6b4acb965fa8bf29c193ca42e8606eff5f9f9dd1dabae89f6ce7ffcb825", + "af80e8f2709e71845ec390e9c685d8a255f3899a745a038f6782bc558259d3f2", + "cc4bd0540882b4270b3b8917e020fd79e2f61706728be31af5fa2154aa061160", + "b5851acda4466cb042db31d72242c3d62c7044e9dae862340fd9177916cc7368", + "d0ae19c9654ca6ec33e5b37dc11ddf541981cd5cb9e89662143d7d0c21efd607", + "816b22377df83b654e2c936403438700f3c1436d108c9528dd7fc8c2612c8a8f", + "59ca71eac048736b6270f5cc907a6b95640eb880c18ed9da48a448053e3e7314", + "dcad9463e92e68fbc97f841048376eb8616fcb7d0a38bcdc8a75fc0b780dbca8", + "17014837dbb09a1a13f9fe6dda176d6300c332fa89935de203f1032d19823c78", + "b2723a0e07d80757846a8e76413b8dd43b62d88797f247fac8552e2b9ee2dcb0", + "05d756513c937b0dea3e925eac2ea91ff534f0fb844ab6c0647ed47913b95de1", + "ab5a62660ee3da52b03cb49b2183058e965b6e386cbb3ba173347cd7e06e78db", + "0d96e4e0ee8e0a5f407af14493d7a3395b258c39476be12965e089c7a3c63c6e", + "ce4410c7d84b348f5c314a06e99b98e90e86f760391166663b1886020249f740", "1c1dda158ec33508e3fc8e4c2684cbe4e56886e729aac7b004229f20fd051970", - "0d93550b2ad29cd294b9ad76048f86d234cc3076a8c7b7b33a1d984d1c413bed", - "7cdd65686ebd0275290c41eaede2b5df9dfaf841020b05d47692ca00d674cf58", - "7eb04c4a243759c8fc195a33b15c01c015c375268e2410462ff412e835c991bd", - "0a101da1955724e24b463ff02f8e079fed699e736f2ff4258a468ddd5e6a7dfb", - "8ef74484130e2cb8999b2a9c05bfc06fa6830623d586090d23f1ab01b09cb788", - "c5c190961615db6a111c16386b9b0a04e1d514d930b378bf3a0903281cec2e86", - "620f0bdd54fe6261d64b7a0646067c82fc9ae30fb91aef5104f50742d2d80488", - "c9e8b02d4dae6994caade8da033cbb5674e40210b151e39d63dc145dc51ea5fc", - "70d7f245403625f44cf0c7ac5db4bc072c9dbabf5d884a10c22efd43447a7ed1", - "1c236e184ee62b55382f3930a94a1d54ea3c237acd3ed526e32aae9640cdd5db", - "415ff58a8e82e96c9b0f1970be36b6df317afa124a4fd5ec733cb27fefd98c50", - "aef8c79063db1d8e8a3e036fc2c9d533c3b08c70bb19e3804c1d69f0cc6fe667", - "dfd6ea35c98845a9f0db8b59ee05f126d3d622016a5b88685fd7d4c87fd2f0f4", - "44694fe00494278175ac4ba3a4c0647199096f3a990f67bf2512b7b7a488e707", - "265ede8956430ef4f382d9f7add5405a47a8ca60f7509866fe3891ff40f3eb35", - "c4ff2c13789e31dbf47e9972a4af8313cf2f1ee345fa56ec6fe4474259f1752c", - "fc7bcc1c27f4307afe06f162af10fcdde7c2b2498c65d4198cbea6c8816169dd", - "62c9443f975871f06ae4803e93b49cfb010e53c041aee30dbba426cca953a750", - "68140f091fd7fb20007039a116ccda85728ffafc49b0e87531d09b02d4c52f90", - "d783032143ee2a061bce204b6f7cb9a3aef47f367cd1cc23a2e8c0767bb63157", - "a25c50a25e1e3858ed6a0de08044bcd93e2d81b11d1d6d44c2a0136be329d3a0", - "66116dff74ef4335e09101d65a3e738e3ee87515287c4aa718f6edd260db34b9", - "5bfcfbce6e4e6ece2cc8acc4418ce5efe39fb6355a32ac00fcfb745c6fb32c81", - "dc9d0eeaa77c80d024ba850d32fda4bdf46ffb509a40886eaf5564d1b2a77d86", - "10c384bb0e26cdb410773380694fdd1128bcdc2b275655086b3afaa79483de7b", - "4326adac27e9832a22005b5f3de855324865cee63e49d9438ec9e25e29b06e8c", - "fc7593355d12e83716f11e5197a236ffb74c49bbe17996c49434993c40fac947", - "7faab4bd141c984ac8261b19031bd2e9f53947e7ca6c9ab0fa5cc668d1b0df4c", - "12b9058fc11b6a83b47285a89e0afd022b7cf0edf474a8481cd03176e5c45768", - "59e5be70839305b4e1287ef6ed1e80a4c0ae1077eadb8e1d0347cd73e660f515", - "5a1158584f1cbc3e2990f35327c80ef0ee0d95e10824f02d5ea250bd3e4e5649", + "d3fbe50b164cd9ce30e5a3afda06ebd7d755826d53174a0cb5f04ac2180c555a", + "f2bceda1b4da95d9bcfe50bba524fef3ef98274466a19be8062b9fcb2b225d0b", + "138df56e342ded7e936a34975bfe856c40e27612d8d3ecce9b435b8ebaf9308d", + "e590c5c49e51cfb243d9f8cbdb6564e069d9d652deab836c0394912e94efcba9", + "98ad7e24193de32c9b75cc8ab0d51730fb4324b01e295e33338166351dacb16e", + "add0088fb425c80af5b9196b7c3d9e3ee80288e640a27193fb31cedce085edea", + "8cdacfdbf1887e4d161f1afdc515ca80aa17929968ffd84730228788e47bc414", + "b34375191aab9f94d554c9e7815720525a4f2df573d91af073ccc4ff379a9c90", + "52f0227938960d972571ba81d1d9ce8f3dd5d145502fcad23e3d047924921345", + "d1f7c0557274bdc6950c98c2378ca4292d3b567e448e748ba53093b5d3f9d2ad", + "05d6ff1ac54e9e65a90a7cce394ae6a0d6bf1fb2d157baa34e474fa8babe7ed3", + "196e8259960bb1eb1d67de9934dd65c71f097eed4e6e840640a2ddbd9a813900", + "0403a530e7709d5abb013450a8c21613cd1c5c7f4dfdc3bf6d912e5e3dc6ed25", + "8eabc49484c91b2a267409f5f5a276e140e560600ddf919ff18941b9b8586e18", + "79cdf8a595968435de99880e8d2a3d1dc500f1f29bf33603de018b9854c5925e", + "7ee8632a9520ab2b04c303fb8b5210d4ecf51e5a48c7b71388c8200138a330b2", + "46fdee6d4a543ab65000f0fe7bf223324c6011732b038f18c0bc96546f04db88", + "7e23284c0ef0d7cef34acf4801f848fbcad7964578ce112ce68b3872aa1c4dcf", + "61c04ecce9c877b3f7a9a02dff4b5176b3505c9a050a392b8d9f0e3406bd5f09", + "ae94ae374e2126d2873fcf558cf2ff26645762eca3f15649cfa5a1fa53afd98f", + "b176095a0af26ad8236f74e9cb3616bafba12ab8732c264812b88168a301291c", + "23f428573944867f9f77fc8b11292fc56e455234368d34c80efddca0feb1317d", + "1fc1fca8eaef98d49fd76f2d1b65994cb1d98ac6230094b83c14b5f564c78fec", "1851fe60ba1984700561eb22dff272e3379c7531b085eea41dba3184f63f3718", - "240e0dc1f2f462e04f60707362ab4561c5544c9c7d668be19f530aa6e61c5780", - "e8fbef2a85273b2d5caa13f3ab6a207970fa02f160fdf8b8e3ed3d7cba7b4035", - "d25cb72e3c427ac19f08972d1d7b0a29633c772263e7b106c4ce1de9b4fe35d6", - "f763b125c846a5c83a974a3a0d2640a34c9ca9c835e86a26825e3877c0568b06", - "c58acd86d2d1e0df779b60b27bd1feb454f0cffd6d1ba7a4ef0563f3a8228221", - "2b4e04a46ddf2b7d1ebf67b0d0adac24aff8424c44da595cbe1ba9283651002b", - "0200002c4c82fd2c66e23435a58b077fbbb778a14394b96149470b2788c78e90", - "88056b960487ac222d1a092027069eb0de5dcba08463f9ba9b1174a5f213a603", + "42a5ea15833276714eb3e6ee398b857b57883812b7bcc602d8e5b1df27f98990", + "b6bf8d5a8e0ac86d267242fc2a32d27bd687dff1db831d91492e10623ecacbc9", + "f2108fc1b649269a1357f1e87d52d4d36b64e80cd0f23afc05b4515d8350522b", + "e0a68cadfe1a51aeee8c447e7ca98ca72c72927ddbda35c1c1a296715569af09", + "f9020387e5199a06ec9082a2e9bd0547d8eba7f4601805ca69f0a8d529464431", + "52fe58995712e78602255c00a0695922f8912e407945c900a55994897768ed9d", + "0595558321cd9caeb201d4b79975ca582af46fb076aa2b286a3ca38718cfc10f", + "f6b196fed077eebce2ce189b46a5fd32fc3380141d64e2e399d05ddea8b2ffb9", + "8fa01e48295823021bf089e15a39bc188e8200ba9670650e834a4bb6132484a8", + "7b80e8437e67878105a4d516faa7fe680f17e5b8de378f54c25dc78ae8bef592", + "028dfef74103f812d4b659ddb1aab2235716185998ada7a6fb5f4e244a3568c8", + "f79c144d19ba0218b308e26b7e7db99639fe5b9bef2c50c46d7d9810815cfba7", + "197fda1e43c2756a088923e60eecdffaff248745855c7a0a41d640ded62a312a", + "a5919eaa229f07e1291f5dcaf730041f495550ac3d0bb3496e20193c98510c6d", + "a74676db4c63d73df97e998f85119feea1a328e5ee3bdf996e29ff99e109e698", + "67fde708757d6bb5b6fee99c97b311234ed6286c80943a3477d77838bb157833", + "5889deab3e6d85c6f2969895e076ca8d2b9b58c494431774c6cb21622e13e775", + "8947ef3879fe1ae4be375b211f8c1291a522ab5594a3347c85b91dddd4c41b90", + "0df9ca87c8b18d5f2e903878e25a7c4d03dcfb084ab929e309a05b3524b84fdd", "0fa232b59af2627a2160fffd1620cc4c9ebf3406711dee27fbec899c121cfea0", - "a209c56861df0e27ec412e987f7fc2291c4636c9dd21562524bd42f723ea56f2", - "e24b2c8c460363001e0e732777c3ee6b14ce3d50bccb667b6efa998a76d85184", - "d8752f919da9c52de8f5850d4197e73e280285287b858f20db97d43b4bf7bce6", - "20b782f158ee101452870a611403e4394144ceebe53e9e7ead0a61d9deb0d06c", - "2983b77265ed3864afe660935e2fe55c6165848ea0fa98b60ddd1721f89544f4", - "9dad54cdc27ffd06093579882522eeb83b57eeb6d424ea6519c741b3bfa3fde3", - "b450dccaff4ac76295f1d7e7bf1a25aa672c30ef72001500f343b053d525da8c", - "645fc6b4acb965fa8bf29c193ca42e8606eff5f9f9dd1dabae89f6ce7ffcb825", - "47b54a959963c3b83b4f7e4f75e7b6eca7f61d2576757d8439e81909f7424e5f", - "b65e00a34c5a64519d94ee200d062ecc9637e12664b2cdb537b79c907d7843d0", - "717fbc32231a050de665db2a4ced442c78511fb2385eb1ed7a1b6f011d926896", - "a1112fe7292886a25c81fcd5260eefe8aa886e2cb7a456f275a864bb5ba65e6b", - "387bc308ea79ca4ce0ddbeb2448a7002c393308dc1ed648888c76f0bd1dcb343", - "39b843ecadd6a7e811a7b1695352ae0258b86b740f823ba56eef875c6044303b", - "5fa84b847c49398b6979d4eef5e639b56e956c227fc766146d4bff37fbd33227", - "17014837dbb09a1a13f9fe6dda176d6300c332fa89935de203f1032d19823c78", - "fb6f2330b46cb1d8028a8aca1e00ac3067dc42266469845ef9600dc0a0e86445", - "77023455d830874172187b7a2957b2027a500cc89b7709e6c12cd06ecb38c64c", - "75597f53da1b5d8a95204d7391f71c8559e3a38773030f0728e22dd0d706ff84", - "4b75356c9f20f15947c192960eb1c16dbfe6dfd3490007b3e475626702acffba", - "0392b9a670c30061fefc9ec0cbb00f1affc95c152a7205988f91db6e62048160", - "23f79b84ba2a56e28bfdb2b9b7e94e07bdfd17d075c1fa61b08be3d1279e6811", - "4549573b755cf4e78033bf8f5c12fb3bb1dae5a5ad788520a45ed96baa63e55b", - "099bf1165d5f4c909f3538a11dd77927771f1c5bf95152deeca05e01a9f5e672", - "9519422422a309af27151deb38ea35646f950ac4c2aebdfd36be8f5fc2d731b3", - "451fc323ad59f81b8278d059a07d873d2002fc79a8bf49b5b94eb0c356c0d2dc", - "84569dea2e042a6e1b5c5fb3ee46cf3d234d38f169c2fa402d2eff78ec80ffd0", - "3a9069176630df2d75fd89f888d18744b98c68a1f31a199fce998939af146ab9", - "91ba3ccc10c777583d32b279509550cb9f28bbaae08c75b57b261dbb41a50695", - "c7d7af3eaed042e1aeb38be6d067c5c67cb364183c10ea39cb00572645692b05", - "efde04bda7758aa79c4e5af330a62325d055aba737ea524309dcf31d024629f7", - "4dc649dc9f9e23f53ddfb0040952ea22ed081cfd790d022a3a8a227dc18c0dcf", - "f4c09351213a966b8630b071eca16b1d6ee3baabc5a257c7c1bc8eea430ba2e3", - "2b9684ac67138333622c5fc28fac790b5bf153b9a0e2bfb8fe1ff56ccae10235", - "27b1b9a92fad3c4503bb624c13be5b1c1c60fee3441a91682970c8f37123d90e", - "6b4ab9418041f5a8c67c4b179fd7f80609fdb410a074f8e71d8d52848b7a76d6", - "16f661b183757b34f4a7370a442f14cbcc7afd1d63f02478b766b2f80f32b581", - "cbd29741cdc73f5ec4028b8e2b3cbaa51455d3df599ea64328ff073390778969", - "333287f8aefde06dd8da7bb3da35c204eef699a8f13eec83a80bdc460c73b525", - "e64990018b44469bd55892207ca7b603675d5a661cc68fb863bbc6d00ada16d7", - "37ef9d215d143e10d881533a7965b9dd3f5a97fb876c099a89d924f569466561", - "2ceddc87bd53cc2b63b0a40e74066d1f17cb0f6780ad1622a6ae9d253ea8b2dc", - "25bae733a439b4295c42c60008dc73f0a239d11fdaa7a6fb2d9600ae1bad0c3e", - "a538dd3a74396bc0a342eff590b52def891647dca6c93ed39062b715fd8eefb4", - "9b43decbbc78f64db434f681c206efbd5d0f0d6f14a0279c78d917b43abf0119", - "1e0e8658108f8af1bc4aca36e0d24106d2b007e61fa527e84b7a1e01b3c3951b", - "c3b886fb35de3547c2fefb1e2cfbcb43c182d0e35a692e28f8e5250bf3af9dab", - "14121018af44727f6edec16ace31a0caf6779c5c853b051e1b1b995d2e13125b", - "d9266d5ddddfc261bf4a4aa76262f61a5522cf14901b8b390514a5c61cfcb260", - "f03425ac0078e98a99050fbb9cc66472056980d87dbade7524ed8eaa50ded4ce", - "1f77328762c808a05f51a9bf5d271083a3a17e31414930e86c60f0d130188fc3", - "444529e2602f6137d02c42630f287f6b9f9ad0c1e7041955da13ab995d1f068f", - "dc5bb223fa18158d1b823496c603d92d0cb1c3f68a4ec7a9fd23e538f73921c2", - "e68e9d74194ba5da4f23e884f4e63f3dcb683367dd44f875d69880c97ca15f4d", - "8a18af0ad5c83320675ca176f7ae1f33666662485b6abf1d402df859d76318f3", - "0f7d85f086ef35423238e069d90304978a8fe197c98218fc431975181e7617a1", - "206a8319902e9de9654852adfb8e6c4805c933aac1240c60600a7ced3a73c17c", - "3e5a4ec994cdffa569df546c8b401f87999c2076b77a0cf22e86e1a6975adf52", - "5ae5730e537a1d636da11015493ed2dc1aa4dee6d43771a4e97b63f911def969", - "559355a6aee9a488a8d7f96bb38153b9cd31bec65cef312a3292ea2e619d4b33", - "e2441770a4b7d99ab129201d3e63e224b6a568ccea3a6cc1f73da72e9fa34284", - "a5eacdbb5b31296d6ec210aace8a17f9dd4069262f706b29ec8aa0c07c6da36d", - "f6c1c7d4a5cbef7cb4779122f75160daa11e374aa575bbcdc5373fbc95482f54", - "2b4a5d769102b4228c7688718b474fdac4b3360b294e45727aaeea0732095bc2", - "2f0f3de4b5d9bc7fd99a46b01cac408dfe42416ea50cc4af0bb8ce3db606e116", - "a43a1ef6de7c3e17c82881b6b0182e35b4e10e1f353b725fc749defe9799cf9f", - "c464ac8b8af79a293c39c6d9577745ea926c1fbd6a3058475293af84a7a8ceb3", - "7ac9c0377d107fff8dfda71c548584252c4c4138af0c7026036a59e8ec2c14ca", - "ccce5d3606a7d81bfe6e0cb84b049ba5475f055e2fc6cf8107dbba00bc23b0a6", - "d1877ebc1ce981dd2ff7922822203cdb7bc2a44f1ccca69016f06e4a6e6408a0", - "3c455b67d8b2290a6ca91a6777602a8c2f992e735a7fb04d198d47e3da137bdd", - "cce510c31d2f6fd54c9be8f15f5a46cd96e30213ef23af584c395c3042b0305b", - "f1f2e6f43e05758a1bf2a1d4d3a904b0e2f0312a57eea2a49aa1deaab31869df", - "d4f650598bfb8ffb01b600a343e8131f10140b4acd4209635e73429b5e700618", - "f6b76849ae341265deb295f715142e22ceb6ec8fc27607ffb86aaf0be32609d1", - "a14ec9f5521e5e49fb017124de80fbc461edf16c006de1358ec16ca752924988", - "be957d8069af8b36d627e867fd6f1b0cf807dc735a1427f41f2410de03379e18", - "d40b4f93454caa1281deeb70394c7fd368448e27b2e7f61b4390643c93745f84", - "9d5a9a1c4a1a3e9a2b21ea6f501f68f5d0f7f2955fd8b3b2ba5c580005a09b3e", - "2699025e78055187d8ba606b6ba23570eaab68bd106e833a569be80169978841", - "096a8cb4a40efb485d92b997495b18fb69feb815cd6c6a32afdfef259926c6fd", - "e21d85847ef158e35d02243914e845c9a1eacf77e872c2ad404661678df85567", - "4e0a4b3a68e0ccc04f730a1dae51cfa553505fc238db300a137fcdb068c556a7", - "9699145c38a951b204210ce608ba7cf12ecc9a0bb5c22b7a79a0128e89fab420", + "901aa3bcb2f19b1e552dac5641b522c1ab034cf2137bdf6fb2d9d547873bbd0c", + "6ad1a245428130e1ed1caeea06afaf3f67aed275c589a101d399ed0e0363d7ea", + "09d12977620b274d00185548e6c2a9194375fc0a2bf8feba4be6d22d78bbdadb", + "33109cefb65b5bbbe66f7298eb977e9f4c534e4263101951272518f4144bb00d", + "7ee088bc0cccf17729acafe4a02bf18fdc857d3ac69bb874e0a18e29b6d53c8d", + "ff1736259af3d69fbcd752095bdefa96c13bbe5ca88a52b724f791d67e3b02f2", + "38a666ac314c42861fa6de14c8f4bc5602d7ed4116c2d327fef9ea9a6e3031ff", + "b6e9c90a0f4249c5bf1eb17c64d1a90305aae8bf844d1f55d75336bf89240b3d", + "45a907915f958be801a7993d3f667f80d9c8481eddfac0f4b0776233b39e9cd1", + "6b43dabfc230f82aaabc1e8aaca0ff0759720bc11f6a28311d8db2b02b4acb57", + "928b1ebdfe293286ad97ecf4e351881a8a695cbc69fec9ba66765d1a3dc25af6", + "61327f00dd24ffa05e342c82341f4f949b16273a39296518064b73f0b57ac4d8", + "b11100cddd53470c3f77d2bfee1c380941b2d8a8dd541c85594abeba329ca75c", + "7f520bb1bc88eb8fbb691b83101a48500343bea1e086e69104cac733c9c45c8f", + "762c8c7cfe16c02f399343aac87de87e26b76a3c4746b300a9da6de69e990958", + "f27e3a4f575cd0d5cc9fe7a58271630750bb84ea647d979bf934e1b5d70b3a34", + "ae809b52127d274a6c061f2fb8a7ef684456705e57360148b17f2b84d12ab38d", + "579d7aee836ef911e3d350570807919f8953dd5b03367041f72f518df945a6d0", + "24bddf5068a2c60ce480ca5baa93870136a1c2fea4c23fffab947834497a5371", + "1ec2933f3cd2ce51066754e10cfbb89112e8ba2f4910629debd865674f26a313", + "99912737b5f67f7251a13cdd970137f49d0341e2e9a2a4a89238c7e9b1f5a323", + "56ab4bb78ea45edfca6b22f7f24a87ccbf58f8eb0debed638cf4d0bfeadcabb4", + "bd73fc9a6b897113bd2dbd2f0b9b061a46166050d17cb5de9a4893f47125396f", + "bbb9733ffd623fcbf6fbca425290de07cb8e9f57bc49d1e327ace8fdfe8a80f4", + "56e735984c497826a1feb361f7e75f5ab02d158ab836c8add1117b88d7f387c3", + "8429210daa2fff79bc5116b3a90f23495198e4ce3018728e5d4fe9ad06153b14", + "0dac9ee63919c5f987ce8d3ba97c5d5489296934077d8afcb2e39cc1ca76cb14", + "966d28e95f4c35e2daad472c3aa0efa840c81245f0583b4e50320a3522866859", + "dd71c1a3b76a66de4ade8456ba27b567920314fe19dbced3f4fa6b9e2298a9a9", + "b88284c25f00ed625cdb074352ea441b8f3029ff75053bb4960d4a1ed5c85d94", + "b7f33fe5debf2125c1609afa1626f52573257d50990c93c05d4e748eb8b3a4c9", + "9eb8b648d6ed2a8fab884d490203bd098a71a6efb6226d15c20518168deb8b0f", + "07a74d0d6083061f4dba0a56a229edb8764bb6511b2ab501a0ee7cd3538315b8", + "82a7a99f39e14c5f6533023d3ae8ac9e377a618b8c9fcafca73456e212a0424d", + "5cd105f7c8d29ecc06c61e605cf214878eee6358f1f6bd2404e77ec1f077423b", + "7a387e57c3e2d596bb81309b6cecbad605f5f3641eb5cfe8914140ec6c500677", + "78b25836b33d92260b6acb38d12f342f611899bfa1adbb7ddb9e5f7f86917fa1", + "eda4ed4cec5f178b33ec92045b82c82026187d8cb4dedcf534c26c9e8b10eaa3", + "a5debf34d05aee6504163271562d5b855ee0f509cf8878726e7ebc87b0172036", + "3991beb23e136444b9ccf2b0a9383cc86f6fe8ea037186d421d30528f7af5a16", + "d5ae9e20274bd170da7b37bb3c9ea6e980d666e866486b8479325e338b2afd98", + "79630bab9a43dd95092596ebbbe38625d4ae2d842fb004e1381a7f902cf1c745", + "4e1c9c0c91058ad4da79304534082fce0aa998e5125a8b8fe3682a2fd46cb289", + "0b2a4b4379f022b973c7afdd5c25453daa3a89d89d5b58bdadd80a3bb3f364ec", + "63883408082fd82e95a417a47a5418063da555b99f60de65177fdf1fbc7930f9", + "c763825393f7f14216307048a32f5fe078af5a65f51df70a86aedfd3188b0304", + "a63c4db0ea2073760632e49cfbf66303f2637d57ee1d73fdbad363b80a0a4bdb", + "25c8e4349fe58b08617285fe2bc64c8537efa0d8ece7d1e15a697c707e68f190", + "5117f27896fbca90bba0c515b8f5ebfa2ea618dbe7bc7458c1b58b616a8ce6a8", + "ec118ceebc0e8f7d04994d9957057fd8e8aa922faf98ad33a900c6c3f6edbd4e", + "0518223388c25add09bd4c4c872ac5c133db4428652fb2f06d3eef0135cee845", + "41c269fa56db5ccccfdd8d7ae8d32cf0bdcf5177e039d2a526b3516e8b788729", + "08bef830defe6215252881f5fcc789638da23396d0728d6a27de4d3b31bf7c6d", + "1f46b209b97cf056eb59e28141ee89d27bdb2d9d2184ebd7f1a4dd62f03702a3", + "53076b70cbfa642cf7e8c36690ecbfdb4c68825e7f85b2409f70bb01ce9e59c5", + "f083127cc1a17649818f3f8f4e8d926a0adb78d895f027d61647f6296b29ec53", + "188827c44ae53568c16e2a581b0b1223666042b7b4cca5ffaaab04c50f762c0e", + "83023f58eed647043935df4f172cd86fd114a9b126c05fb21d55a2f61a153547", + "b6d7c13913e65b624f1d2df5db0bd1602098a6ee408d1cdcf0886c80f9a90bfd", + "acf1a25f2b4094e6137259607b76cee9e4dfcae8eccf0990a7d1cd371b3be6cf", + "d253678daaff152eced4c741cc7e223b25a586e1092c251e02f7388a7cecdc64", + "028914766d7432c34de817e34d7a51daae43d590b4e281e0f0d1fdbfe5a6e445", + "259cbe48b11253b7cf5aba89778339a97d3799e0b90e104fb5889af9a0970d1e", + "30d323f6de86cc276241761f02abe85b2b21484446dd609067f2322033fa9975", "ba4108e337c0667746037ecd9ae06966d17f1e0ac5894f6d0e67d3247f2e0b4a", - "9e057982e001541ab7238a8a606a2b759e0f13297ea34e312cfa5dab62b55517", - "7ee088bc0cccf17729acafe4a02bf18fdc857d3ac69bb874e0a18e29b6d53c8d", - "015f31e83c3b708ea4a087371d90d2d527769c349cf4ab7d981e797777295e63", - "78bf714ca5238a877d8c09a563c3977f138d1e2fdc9e21d06d776f375adceff2", - "7446e455fff239375ce1f732e96f001e96e535289657207f346e78a77c72f5e0", + "42738473419953c4753042f5b2993c91155b6854513409a745d0992f92af7304", + "bb479896dd6a6af056bdda5c3a162909a6f756ad352269df02add8f766c7c73a", + "f7578c7d45034cd931c356e62f3e16b3645d481311f332a54d361cbd1cf73a4c", + "b309686e0deee7d71b62acb3b57f83a175e3bfafd50d1800591c665b0d8ab8aa", + "51f087fd2f702951d72a232570277510a1de66b30fd3d26f01415dfe7bb9c02d", + "c818bc6c8a6099b9f5edc5cb494e7e485c5cef225b533db1c3735fe53b4cbddd", + "9beb12f74241e657a2cf3b698d941c1483a7e37f2cf2b5d6503c4cb4a6db0cc3", + "32467bdb988c97dec509c1bad776505f01852a4dd1fada4356d7de6cf6e2e1f4", + "e4f2e8c2845dfa9fe9bdcd928b4f62738716b15bd94e77eeee7ef1e694eeb288", + "29bce1bf45301b01fa6940da5b2879a2bb10ab17d0fbe32fca4e14343f2be5dd", + "5c4f05b0d045d4a2591afa88a2c00a10186585a42c06a8cb441d8b70e6d9f429", + "9338ef09cc1092b5098e072b37127840d4d6f1afe920d6b07ab48c88ace3cc0d", + "afda909ecce33426d0430b92fb8995d47b000f9be04ad3e0b575cf17443ea3f4", + "630eb8946b983e7b612b7a44bb4acc9cac0c5107545d75bf0a93577be2f03166", + "7a5c02c99a413b3e9e26be1f2a766267e41a3055c433e50ef1e7f820a443e843", + "a55b65697a2f04710af2638f4769d6ffd804105420ce2cea52afd7736ad69dc7", + "f4d1f72945ad38c496a873e95ddaf51b0e7a6523c76944cef6ccdddb6a3b629b", + "5c6fd283afe0811d419ee0f18df577a4e9a178ac6aa745ad88d69c236e334da0", + "76ce45a6810ee97d6db885d53c13d96272bee4127b003d695f539949d2165f8b", + "6543352e49d75fba78681a7ab9262bfd70e8442cef512b07b67b4b8b1fe3c227", + "52ca70f743b2f6074d286156d3b6c461bc390339e6bd76afa3a5287d59f2aa21", + "1d3a50529e4707a68071b025d86eeb40ee07afa0ca00f8c5a36fddab562d742c", + "096a8cb4a40efb485d92b997495b18fb69feb815cd6c6a32afdfef259926c6fd", + "442b417fd1a4ab20a2d153914b2aba297ebb4a13e896f9ee171e346c68d92f3a", + "e16eb2660ab6ece15609de00d45266fdd1a51d0d899ed465bd8aaabaa5d7c79c", + "a01e1cd7ad61318c5d530a2020c32eeca6b1ad938408941437f8880d71766c0d", + "7ca6e855934dfc04023316e5b8acdc4b8a575e2c47d660ee32d9aef0e13749ea", + "2f4cdb495fb2b24b8e4eb1adeecadfb994abfd91d7b2771a940f1beae22f42d2", + "dc2e1de19a03a9e90f7f7b00486adadbbffa0ea330049acebf739135a863eff2", + "87f155999e0078a0e864089c6cb549066fce4a8df57a8c909d3cff6738cce7e3", + "4b75356c9f20f15947c192960eb1c16dbfe6dfd3490007b3e475626702acffba", + "9d5a9a1c4a1a3e9a2b21ea6f501f68f5d0f7f2955fd8b3b2ba5c580005a09b3e", + "0e83553f15840fcd8636958c5706a6b43f9f74f9aec4dcc6275870023812d08e", + "63f08640062c3f9b129be7aa34218ffacc091115800c14aa9cc41899ff9483cd", + "65ed537ebfbc8b933f38b801df81aa16f1f2a8a5334342b007746a582b32c2d6", + "ddf13dfd6f8a7e153f5a3917a82fd108f08464f3a7cf9a0176c2ddae8bdeac02", + "7cb6b1b5d81d4b7769523d80b4f1ac88da9e278a73b601ba1eac1212dcbdbd52", + "e50864c736823a3ad129cbd3d49a910e2325c3eb1244288beab72f061f2a0c83", + "86db02a74ddac3d4010ec9b6846dfdc0350ea8e909511f4f462b2b2750c22539", + "9a0c6b9fd844c5ac3399174a0f4ce4a7a9f7275a695c22104de744e8d2fa5599", + "9aa37b1db395401f895b655753e66284b35a898721c7b356aa2de3685d9c5fbf", + "d2e6fe4b0b9c806cdb5679c210c307d750ef49c352cd23b2573f84c452024e86", + "c497557c3b055ecf2d89726a08e73e475c8b69e99e578b795408f0d2ae5185ed", + "d214aeae374fa16b2bca5c091dfdef85c3ab2ad3e484dbbd53b0ba3af4475959", + "1a447c69aa675190a833c5ed37cedb6de6bb09fcd31b9be4b510956f0cfd55d4", + "8d44cbc983a0c97531ea7860f0d4c9f423555fafff9545a87219769d3b9e92f8", + "7db987bf79dfa57dc8d5f9e3dbd4aaed55c88891e81308e33febae91c0c7c598", + "ecc1c968bf2b0a43b64ea133a763a9db4ad4fff647e890a8932d8e17a6bb2774", + "79cc9acfa9e23104ddccbe2870a31ae43e443603fa4f8174e15b499cea7e9e97", + "36d7e110d20b149a4833a36c437f74daba417f7318fc448b045f5c41476c0647", + "8c1c6ade1f6ccbbaaed0887361e5751761c0cee6d2f2560ce746b479800625b1", + "c71be87eb6bd3e40618dba852593ff42455a09e2544ec676c2c42d51121282cd", + "7bc13084c7cd4cb116131960512064232d7b2cd175acbafbf26503ee82865fe3", + "398dd8c692a95ee98f45ec90d49e45351a7e97185be201a8fe1a09b2015e9e52", + "39dbcc5615e579b233753d4829cfefca464aa97eca6d15693915efacdaae887d", + "55752485317a3d7b0f357a9529ffa948273f2aa8085c349070af1c1265e123bc", + "7dc29db9042eb8a46c5b7b18cb2e825d24154013acd0fcad48b149cd4b85adb8", + "73acd232e2700eaeaacc5eeea64fa00e170436ea8581698b1fa03999817adf1c", + "bc775a9200259561b73a03440cd29d7a4cef5ca0bb16b9b9c9426b6f89e92b6d", + "e0aa067a64a8506f501a2f644b8058a434879b51772b0dfefd9adca16e3f2941", + "ed0c8f26d13a24642ef8cb9496f9bf6d8e984556e918c238afa7e13f8a22d9ae", + "b3193e8ec8710cb8dd4868d76bcb19e0dbf3d728e9ec928e0e3fe4b386f1a20c", + "d755d41fabd7f5445dbe96997ca5c33820280c6d1e474799463fbf9f276d8f9d", + "092a862e6235d2b43c1b3f9269b3cc8fee10878bc02ecf583506aaa58a96ab6c", + "cffde0e897e1ba6a223a9e31643c5a35d88670a2488784c6dd5b4316b8a7e504", + "1bdf89e7346d2d803df97859f0f4e716f7cbe3e7e915f5bd676ef0e03682ade5", + "d714562a4db36c6692e2336f1a56fa20e0aaa10855b12852e1282ec3b9ef7888", + "2a2a808b13a40820754d4d606d5674b12bb964b9c85006a785a36ea22922957f", + "adf6c5e10329b8ad69aa212548ef5b940dc98ed65ae2d7a5777a96ada908ead1", + "90e50d920a4a3a9c6e7e88d85a497e2b93a33d9db2466d0baa07acbc33d0d610", + "9caa70b23c8d8d2a36992d53dddf8d04ff8db21fc8e2f3d90dbb8f1e49b11c73", + "92b33a5f915194422570ac1dafb66cf6feba86151370a0783acf72f34cb57a85", + "3940e88f43a2689c5741bbfa1aaa455ca7b51411644423a6f6b26014a365379f", "dadbfd0a805d3489889f1aaf62f718fb734186525ad675c7e3d1a1e5abf42acf", - "e404a1dbfac8eee7db7276fd265eaa7256b3ba886e8198a2b46fab3a234663b8", - "4dfebfe4b2fdcb96a02dfbb24ea028dff4055fc26ba062af9d99027703089747", - "0dac9ee63919c5f987ce8d3ba97c5d5489296934077d8afcb2e39cc1ca76cb14", - "a0802fe8dc8d8d46cdea8e2f82c519e92fd2220ee1f62aeca31e153134a80931", + "a96612175985f39ee3a5610321cd0967c8e7e2f30eb44d6a0f32b089ea1bf79a", + "48a8058acc3916d82c0c425e6e2f6a4e296967090022a26e4cbe4d5449d04e24", + "e3489db5707b0c54c9190c0ddbb0b0f14db4c553429cbfd7cac55f6c70e4ac98", + "f77517ba3c5feb4122a22713fdd46c256860e8c85c7060b99f07e91b48b2763b", + "e0a9aedc4760d341d8adf58807617094260c513ccce4bbc7fc54af2e2bbe426b", + "1e134223ade2ab2317cd00c5a3153a92411d9f6c7f3b3a4542f947b279bedada", + "0d348e334184e62a36ff4e1dca807b194915b5cac48b492d0ead69a2f42d68cb", + "33602ca91cf65a437f0bf8eef1d8492cf5b8d506103cf4c7a5c352bda77073ea", + "cf10405c7626e9cc8133ea4cb76c5719ee522470e2c6954f32c4e3d39efb0415", + "841221793b996670800bcc96d11c8949627758e348aca7dd27ccb1b5fa3c516c", + "ffd35e4701398f156ee8ef2ffac95bb992d147ab77b5b38be081b9bc98014d1b", + "7633d94f0ccd49666731353bbba0b2e0f26244cb5ac206c481393d7f293ec9c4", + "71607d518dc03cdeecc015c72932bf1d23fbed8b76d641cdd67e08e45730b24d", + "5752c34f042112bd661bc4c32247cc0e993f3fa4e3170017b405fd0acfb47013", + "d1bcf3655a4841ca3a839bbaf1af31a93d0d79cd6221afabafc2905597837bec", + "c4ff2c13789e31dbf47e9972a4af8313cf2f1ee345fa56ec6fe4474259f1752c", + "a69ccf68499a544716c83f170e49419da8ebb67c9e244750b739573fabc113eb", + "28ef20af09a429e6e6b39a3cb8458ae6f48f89dd4b3e339cbddbc5aa176cce4c", + "24ce14852def4d1cf92edf6d78de0a90f8a1fd0f23a7105e67083c7d16673498", + "9b28c0f6ed28e2aa0f7557a3319b78104581b665f81b400e84729f3711c2267c", + "300ad7c03c062799a35e41aba179ba97081471c3860a1df1e2773450763825bb", + "e2c07f535a6783e44530aa13efe90dd5e596fa7210f6c2a913897a29adeab644", + "4a825a26b9fcddb6316a2cc8377378f5e54e23de6afc4d11f69a62b558944114", + "9dab5ce1eca4f038cd138d248416dc76578046a0f626b081104b1378dd2516fd", + "ac60db48653203440d229890cc62e3842a6bcd09676161f841c352e30055ea1a", + "f9d76cd2d7cfd4d55edaf59d799551f5e0bcaaa0213af475eb13f75194d0cbf0", + "aac529c3008c2b4869d310dc4b8f73a4b473bdeb0399abad1cf587b1bbafaa08", + "afb7b415beb4ec5dcd9c09a796eb415119096b1c4e2bfe37e31c30adc6dbe32f", + "65adff92f062fc62b099b8d7c53f4cf5587cbccef72c025e9b96a79e4b21cf80", + "9adc53b2f0b39dab4c9de8f4037921a91604d2bd626be19d3e0fbba849ea6b82", + "ec7f727c426561755b75e1566fb40627d3970eb433b56b05a2fc47e0486c2d83", + "c3d25b6851756a69168dce5c892e6dcf6ad82f73c21a71d97b7b893c1297ce8e", + "c67ee3760a3ff018ea1548efa822633c4090c70b2f736f0340ec0a813d3402ba", + "69d391f4433bf48452490ba36d608aaf9128f911768031fd45ca117b0e4613ba", + "62e232ddc39755239889e0b6013cab101556db674676a7c6eb6e74002b8db44e", + "770672beb1835172b9fbf091fcbc18b34eff39a78ea000cd140d2c4c0fb3aae6", + "36d15cf2a0da84e593ae3a3e348ae98d66738d6463db30006ec39d74d609bc3f", + "1c1fb31649921e3c8b72098bdc8b002bc326d7b96bdf026aeb2ee05d1608a6f3", + "eb76f9f21a6219e3d7108df88cf5b81a729b312d3a64072b83a78d5a64c87605", + "7f718a136ecca766c5be337bfafcc103f03783da73756bff666ac2d05e862452", + "6aca0a7d2e459b267b713afe61c4302731e75470ce7f26cfa8ba6bd977254296", + "521528bb44d66b0f6292f000b1578ec023b9b4ba239f11ac60cbfb5b5fbd68ac", + "cc2ab067791f573a9966e5acc93a66c8316ba7196a51950cdbdcd6b6032d8d6c", + "ac17f94afde180b3ba4f464f544f12eae763f79660049a458c9a5b20855e12f8", + "1d995f2b05f10a60901f7001e45bd1c2c1df9fd0f31f5b5af8f876ee14415839", + "903605a3cde2c6cdec81af71d06d684171b98b0dc1dd65bc1d5cb76961af4267", + "8b1dd1c1760358032e13b83dba555cfc73acb291d8e9bbf3db8b0c9297b92ab2", + "470dcef4cdb1399c262b2a3c23fdea2f3833adb1b8d9e601b51e811b6e928893", + "068ec18785ae0e8545fe3b6762bc0f2bb0a94ca5df8dce91e4b7b548fc589904", + "e5a7c54c80399358a8d2d57bf3f635092171567109568704209cc36e1f38df3b", + "bfaf77517219e88616a8090cf716ad5d1260ef17a5719deb0585a9360437c2ec", + "f89420e8644550d4ff69006c1d543ad43910819c76348cbf330089a503f9200e", + "df708c9fed7020160f05762da5b7bddeb3f32bd4fdbdbe90302e04b19ee6059f", + "7e2b9f0e249d770510be908feebc3c0802e4a95effefbe742548232035e0ea5e", + "8635c4245422b424589144086a3b5cb49296e70ed066cd56c888fcbf79e58215", + "ab6217b1a683cc8d3dacdc7c519b3b2fdac8ecfd74937c2db964d160c1ff35ba", + "5155d494f24ec55d12477f93f89ea5b209bc9978d75c55951b96187191103091", + "6256915338bbf9d0b1fd8e7fa0911fe6b92bf899f38bec8d97541674067fb4bb", + "ba94de5f1601a50c5798528b6b2e6f82669792e2c3ea107e2ae924ca7575da69", + "f2510e2ac1d39e1e03a0a762a4bbf3361d3829773266cb924408a67d752bd8a5", + "2e5c1c281c82a4ac014b36e9a276f7d66b107b23541ab878bdd5eaae030b8281", + "4e3fb94d726c59f6f6e91ee837f9d87eac38eb13b4801cb54cabe3d5c5ce3650", + "d25649cb87a74d9cb1c1aebf44f4436a92dfc290616e9eeeeedc91f13ba7c3cc", + "e2e4dccf0a6e6ac01f8e258f508fd533db632b2f66ee14de5b54291b4ac15bd8", + "3c540d60f96e69c69d2529eb2b56a00591e676266956f09206a37d70ec6b3ec0", + "dfccb8aa5641e528935afa3053bd8789af205c13fee0266f3b2eede569e97329", + "900a803959e9b41e26a970863f8a26d9fc2e584119e668876df8fef346ceba87", + "0d11edb94a043d4d5a92e52fd07ea58a6515266dc7a570982c2ef10e82a915c4", + "42cae7e85919e12e830420a27cdf8e43c30666b8953498f31de449e7f6f49cf9", + "ceaac9b265780d294b3a7603bc917cde8809f07d8aa3c72295d4fd475a9d2871", + "6267c7e9db6801f55a795c0e085c65b340fad0ee417dfa6216b4e24972ee8dcf", + "4a9ef42ae48c1e085341b617ad33c426e59b94fd3a81ecfddaeeab48ae108a81", + "7c6309c9f90f24372b894b7d9a63b8f587f932663d564eb32664e093d22b93c5", + "037e0ed281b3ba57402a95e94493dc598cf8a1cb401d9377f223848d96b8c5be", + "a6ba5ca4374bd7012811aedaf0a1fe15464622e59ea74be28f8222c0ebbd488e", + "346244af64521c79adfd7ebd57b35ad87dbb6540ad680bbb997fe5ccadcd9426", + "a28421bdbb6bc6c24cdeeef7180a91ae1432b4fcda0e2b2e7d31c584e36cb785", + "0333efd41834a9542ac1a5a967c323ada1910ede03911e0e6a12dc936b8bfb48", + "354bd9a91c421ed16ecd8803f2bd8cc44bd8320c9979e3498550b84c023b4b92", + "59e5be70839305b4e1287ef6ed1e80a4c0ae1077eadb8e1d0347cd73e660f515", + "783986ff162cfdce8eb36a7763e4a68c6231a6db19e5f4fd0a13287c921107db", + "fcd6a562af58c12c2020cc360aa90a3923639a83f57434dbb2649429f42842a1", + "16c06b53dc2460c4ba70e42fa1f02f759122cd7a5f22701a0acb42c56e9a8834", + "65e9d8aef78adec70b21f518a9cec1fe6fb06676c6c5146b58e10d34a2c77a53", + "1fa5168955a343f1dda45e8afbb34c698768db649f5880501d3db91fe967805d", + "dc59d77a11954810103225cf4b046374c5e660db00f4fb8253ffc365a076daac", + "aa513a81f47fc54baa7adc3d5fcf936866752ab8a2f32ea317586f2e87f91bdb", + "73524ec1fbff6884d8d33465855259e63cd3d86671388dcbd1190533ad38b483", + "82f567b5a6e9652eb1b314d1d2fa31b07926986feb9a580e31f39e1e73413878", + "3777d6e91e574a0ef7461960cf4f3472f4ad08b4bc79c7670fc233b5a2c4d3e8", + "2bc6272e7cc066e600e1e5dfc9c3333fc8f5ec3476927ceff64aa8e2e7933431", + "22ef460f98d4d52cbca04df92af1b0844b340c2d15c3bdf7acd8dc8ef3df33f7", + "48d932e6490c66d5afa21344eae8531da1ee945d10f5fa483e8b9e30c00f5e49", + "b1b1cf4b912be03767e072eaf4d9d7b0df14caa294113a1d71e89da1774b370a", + "bfdd48b376d9e0cfa9d68d45cb16d8e39d28f2f22688d6daabb67c42cde6e700", + "197ee1e43fcad2e6488be436caf13227eb0169458d5d34d15cdfd3cda49977bd", + "de53e65126832a08148eeccf144af9f3d254f2dfd048c86e6e11f0fdc692179e", + "66f7b786de30f3cd5ba1c8b0d9a0f4c2256fd09c8d918cf8bd32d5ed363ed606", + "9020142046dda4fa206d1ec08209c28d475232d6eed4764249ea1f0b56ff7608", + "0b4d719c825cd8990d7452bd35f4ed13ea54c7a0e059efcae4ffb3f1388c44f9", + "89de18d20ab7a5cd7265e8ef12cbdf3d25117f7b6230dd7b330ff18dce88fb42", + "654b27fab2e11293e21465b7015eb208b9b1dbc42c0f5fd01a3607925c1cde61", + "5dd61289065df45692b1c945d07fc51172616f6fca1316db259e0387fb417198", + "a53a582be265f267a2e4c85ce5b39fc0b3fbc644ddb545b3d766b2353f74d1e0", + "924de1f470fdc211ddb10c88065f8369b6f9af12267e6cfa3bed92b381d6aa85", + "eaaaafb45c6b0b2c0bce09d13a8c23f09a0e6ad6a257c130984374eede76d653", + "e4d070c7b3f28eff89726a2df3228415646f7b7b01cd90e5daf91dbb22337f9f", + "fc7cdab5dced8936794237e6630d2bdf6ba0876da9e0e180edbc7f1275a6db22", + "5e30fc27d92232e469e2b0a98692b4a7de28db5d27bdf886b60d04b836c76534", + "9f1df2530972f3fe7c85671e97a3c391919d5fc15d5cf663dffe4cf3e0422ac6", + "725870ac937f6ba3b24ab58a8d2ba085554a4c6952f1fe27c520efb686b7fddf", + "3d1bbc2c84154096b20c468ccf332697c17891cda610ddd64035ea3d5256dc61", + "3ae28db5361e59988f107164f11d88dd923801caa5cf1808fe7674ea257a00e6", + "e24ca281dcc373c104020d128d0c5a00a79eee357581e14c8a5cbfd0b72221ab", + "d7de901e2d9d89dfc5941267678e189c89f595eb2fd99e8ee419d47e36c2c4df", + "da44a557e77641d6838ca632ea0df38b9f108eb72a558dde37938dec4e257571", + "d4fd0ca6f861193e1fb632e1ac185b5158ceac3b761b48e5a0411e2821fa744f", + "2c4dcc2ebbcd7b62ee916ed360a170a7c1b47bb043bddd4eeefab67a3943d3f1", + "5e00868b54a1dbdd25aa06133a10067ca8563798d55e23b4be0ded2846e1ac05", + "aa35ea4ee28031a33f12923d652591e740c5367e33067836a5a949f246e203ea", + "7ffa26936746443ed19cc59ff253aca50f5427fbe14dd72d5d5bfb32edb48843", "975c298847b9d47e51829fe5c5bea5b9825657e5ebce4d897b014c23dc6777dc", - "0f57350393499cd61d1149b395ddc67e9e0e7bf4cd1cb97765c26efffec7e8db", - "e6bf5d704c5410de0d9f647769ad1a2963969ea67bfbc15570d01eb56bc37ba9", - "8b3929388bfde7a406bd6b82014e65bb5a5bfda5bf730777fc8b910d12b9f384", - "8f7c4398ff9b80ab3762be84a1016f42d8d54429b6ba485219bb1c333049e94d", - "ffe4c788144f17ef2f52b14ff0258dfb8f50ded884bbc7cb2e206600d174449d", - "df16175d100d3ca8dae7be938b81730f0d504cb73763ccf3f9862ca74bea0d9d", - "2b5fbd38da9e0a8f6b61ddb8e2fba9f41b84a19351e981e751879d3e6d57793b", - "2571071c207de5279aed628f30c70aa4205eb17f4640916930c6814ec03491de", - "3b7c03ed42e471e9d3fa96294612742cd63cde819c561d0d58b8e6de1febf65e", - "d4aff4810357ab7908294a18c2c23482b806b40a51856fa30c6fba1689bacabf", - "3909da28ee35246064478980e92d3681a2871775984aa3e363ddab1357973f97", - "f1fc6e988730b11a9cb5c010dbb49154cf04ad0036fc0642426b423fcb7cc4da", - "e3b6296587ebe8e840dfbaff8911ddb1e1a0da0cdca0a33923a08dc440fba62a", - "5c4dcafccae92ff8331ae0d3ed6f9543581a714767dc1a0184e98e2e82990d3a", - "57329127183155b6a8237ec48dce5c03f1fbe46474a01e5b71cb75ad467f9e93", - "2d5ad4801d3914da1dc6bfaea91bb2197978810762803856fd16f2a94034b079", + "e17aa0c5262e0826bb77d983b3e2b9399f056ee0db94c0fcc0a39133a4d243c9", + "0f2b875eea5e9727c95f99a8ff2ffce991e422de14d105598db63be9a3e12fc9", + "2732012cf9ecfb6ca0dab4aaa09d93071ed66ae2f09b7e7c274cdcefd8f2c555", + "e28f45be9f16e948aa127969d0bdfdb39ba9dc4124dbf2e3b4dc56a0b617350a", + "74ca64725ce77f87d8fa17fe6fe1473a322de09348c2eb062cc4f0282437e34a", + "b577cc618e4249e86ad6a4a17dc67d2d37e95e94e588685fc843b8c4db0b5d69", + "f4c09351213a966b8630b071eca16b1d6ee3baabc5a257c7c1bc8eea430ba2e3", + "8e2afaf278dde8842909c5c49fad1eb1e46b4b569f2ff315229de1e522195072", + "d8752f919da9c52de8f5850d4197e73e280285287b858f20db97d43b4bf7bce6", + "33470ebed5c5aa47864af9d0e73566df779addd48199d4894f79f08e62244fe1", + "2e8b439c910a76333aa0615bec7bc579d08f5fe9228a44586e908877314af483", + "f61a6c054e21d9cee07814dc52cf398c4fb3ec7c552457bf27c2c6f817efe85e", + "e6755d0e72be422b49b6bb1530faf9fc00a583b119342156148f13ce3e396cc1", + "1948ec267cde4fe77ff11c73ad40fa9f72491857d9aee2e8ec16e27b8bb897f6", + "f71590008c06882c5805c225a90908f558de060013c275e54a6a675d98594f16", + "cd99319364f5cb9a540983d0173ba20ef546ca4470c93c9b06f302d4f1d0348f", + "ca52bcef69ffaa66b7b61478b394685cb962d09ef93663f6f26867ae9c115a28", + "b8767c478ba40de6e0f24ddbb20a3aae9bf31b054adcefb7ecd629038cf378dd", + "bf2e3af74c876546d21c3f14042a8e0e65631d6268dd34d23934c138d8bba403", + "be07edcf381a02e96a788c07a0082ffaef17efcde16b4f4e74214b1da21cbe09", + "dfb39765b1695df590f25059cf1d908f4bf0de89ce4b384a7b50f8bf1cfaa300", + "4be93702a6a0d2b8a1a5904dec4b34e5d689a2a2ee0fef5323813e5daa63546a", + "77fca8673db28250fe985a695e669ee1b2354b04bb38152a355926d3075b6020", + "858d4070c2ace31a8203ef296f453d991aa3e96f05fc41d830232edf4fda1d13", + "6091c69e3b085c4d38222e3dba3087ef98aadf77262719ef74ed24ff6c86892f", + "fe92919026a76f460f8727991c6b7d215c4f00557d5e18c853513dcc62c9ec57", + "c588f71042d935d7b1dbcbc3813d54a09ca6c271bd4087bec50251b4bd6ed2b7", + "ae0351af5923ae46c2d2598ec20a52c25bc7b7dfdc78dc852e98d3ee9d19b118", + "6ea610326013e9078766f7aef34203746d088a5c9c9c483be93e893bd4f31e24", + "eadffe3114db5e47ed2c1049b749b330454b326ecffe473d9162ae5bd7810fb3", + "0e6886d857416f8889c72576499742c7e0d6f363a252726b55d32d77e49bf1b2", + "663e31a4d70b0b531966f64796e9b488257d50a0a81c38c98bb43ba32fb40547", + "40a2fe7334ae72a76df01ba52e2733bcb6d3c7ad2f415b2df5e543735c3baf8a", + "126023c4bac110c76d5a52f0df307c85e26b55286b19a2212d6173a4cb58cd0a", + "287fa89f028e9a40ae8974d020bcbd165f3616d605336f50e0bd38120dad62f5", + "02b9bef8e3b3daa61ed040fbacf9b6702a4ee770e9a79803de5abb7e5ae53981", + "6aeb82d80cfd6ec00236692432b2cb3b328db82e612a787cfacbebccf14de310", + "5511ebab7d22527a65d60df6e9e2e88b0c19c443f223a819579e88dcc40cb4b1", + "29d84ba46cbbf0d14d932b292a92d563cf3734a5bce0a76505b609c17669e48f", + "955ef9d991e563fd557965463f74f8206f2bef41155a81b457e13954b6ebe1fa", + "eb8452fafb1aa04d0389a2909af1a10ffbf9ced46e57378df2adc95c334b76dc", + "b49f09e999021496f8322e712087778229f2f295fe3001e3efc66e7441647b67", + "8fa6caf47438e0f4da29d1aeda4a13e06c1fdb8a94c6f492826e481c0ba090c5", + "58b323e6759324cd9fd40181d463814e778dc3a634931f9847eae61b1f273057", + "a96f1f3d444fce52a2a2bba7aaf1e105be5e71836539c17f90dfbb86d48940e9", + "62c9443f975871f06ae4803e93b49cfb010e53c041aee30dbba426cca953a750", + "4a3440a95434bc6c776d0f1e9ac0eb1aa83f916e556325bc731b9a44fc62dc7d", + "405d7b084e604273e8a724499a84c6d804aeace820b9fec313a2702a301b519b", + "1a4bc7d4f3f7086cf6126ec6d14d7b3a2af1d2af02a6d8eb928da233b6ae2b66", + "4b1f6f27c28d8139f836dc5942bba7184643809f002372f6cf0941da9357741e", + "ca4c066a791f3a3b545c06e0fd41d9d7d40871255e8feaa3878a1027c4222438", + "7ded86877ea618c8b1dadfd4e9dcc8b9331bb0a92ae0ff161d1751dc0254986e", + "03395fbdb7de44c982436ace54d7522e62b2fc76ea113afbc1831ad77f1e8ec4", + "9e8a32337407fd1ebb3e211baca5a52d8f11b76e4bc16ad486996256b5830e0f", + "266311def1e1dd5fa1f58652adcfd57a3e891b05f2f234f02af09d2ee9de0ebc", + "9977b72122d25e21b4350f894f0dc2eae9a6faa598a863c095f6a27e44bb666f", + "3ff1e4cf77f208defbd80d37c35e541502e922ac0861db2e15ca9c4b99e16d91", + "0d7ccd1a4b95c083d6317e1a0bda4f88321c232c0bce4e5c3e39d753f68fdc5b", + "e3e258442a4af8ab945c3445a0351a35a3797f35b437e1812ecec68e9a1c2ae3", + "65b0d410f76d34bbd221d1b85e3a14916a3a179dec265463044dc68474639653", + "fb4b785a94907294387f776a3d3bad78d746097fec8026028356ce181d8fcb76", + "c3df87ed6e75f2d9a68394501997f4fae5a9f54f0451b980330a6059d39dc751", + "e77111c9bd1e657d55c1007a03108f458ec128e8dfe465b43790317f7b9ef90a", + "559355a6aee9a488a8d7f96bb38153b9cd31bec65cef312a3292ea2e619d4b33", + "dcc47da423e6c32dca76b6eac9a9378e02d5e86d519e9e458ce30c1ac16c8635", + "995a2a16cc7b91c92bcde962e8f94d7cf20ebd7ab149d9ebd3505e6c80ef6a18", + "c39b49488a21b72f219e32c4cba37587fcd368e39cfa225767a7d3e8caee526f", + "c9cd1804f78cf13ef256477a004686fd51a48e5ccdf0671966a7c6be0ae77ab4", + "9ea81b70db68cfbbb856bd05d8c5e15e5b7e9136021625b431ba69bc79eac0ea", + "3c5ce4f647055b0ed7e2cc7887ccf5fac2caf45a28f014d792f71304fce256b0", + "01fdc6c5fa3be12030be97672a7f9c98c9b42162f0e4a7355d0b8831803c714d", + "c33b2f548b13f85607827fce36232aca61e5c65456434362e7c1e55a6858afa2", + "feab5c54b74adeb7282ec7269ecbe69a75be8110ce0747b451c139c45eab96d6", + "a06a3c6ded7fec001cb46156bbacf617b1962fb3cd90a828d1623e429479deef", + "028ee42d4513512872684797dbd5bf9c20591754864d95b3e359aa7ef524a5f0", + "d9f91ae377a5c3466c3254296debdf9c5b3f7e962294f0e5317fee70742ac9f0", + "fa6c89b26f2738aa12efb0b76f4423068a52cc3df1f63a97ebe8170978180e0e", + "a1112fe7292886a25c81fcd5260eefe8aa886e2cb7a456f275a864bb5ba65e6b", + "761ec8a202f265b387107d357377fa4e76db16f0db7c61b0faa48e15ca3b56f3", + "80fac7df842f73cf444860bc067d0fde2688405fa86ca75173dc4d039d796b97", + "b6e440eb99c6a4f5f43ae780d813ca4ecb4fb5cac0146319034191e1f37b36bc", + "7228a8e6b21d184781ef76320f597609abd822ee45285c2b36341daf212ce7ef", + "1ca2e249e080b9217c3d85bb6575c8ab59fa3726607fac8f98b06ebca5fce368", + "aad6df027fd24daa4a2146a4a471ecd18523d00253a690151d7aad02a0cdf6c2", + "529a661f9f25c8067fb69b800ac42e66a3a11344c45e25d334af0f922ebc6214", + "24ae5b21c387633dd00c32dcd7f13dd25f13198506666404e4a4297d76f71303", + "2e7bc4e078a1d993a9ffb389f8094715a152b4137d319942d7b21d84c17c99ab", + "3be98e3772788f9418a92fa64718fdd12de85021a174d94846b529ae03211353", + "9bcde8944bc46b5bdad0a58a0889ae5f83fc7387f40e131a110d54b8a0195cf7", + "1b81839f989d1ad1c1abf4d9a1ce05df14c6c53ec4734b1dc45b6722b1d591b4", + "d4d214e62dbadb2c1fe1787531f64765d4e7062cb2029f3fc7285144983da78d", + "0b7962c654aa5d397e5fb314f8f6d76451f987a8db134b5a6c81959d2ba2f278", + "fcf8dbc0c4b216087d1b788230ae5383ee6c2bbd23a9a53ae273b6424ae3cc14", + "6261d7b53e178b20150a71c3378fda95808a2e673441e96151f9fc5a41a06b92", + "4e1cef63f6b4653948db900c9d0219787dba48e64bc480b7254f63b19189fc88", + "5da8ca774eed34c384a7044c9ccf0eaafeb6241d52dbccfb50961c12565c6aa9", + "8653a6ea6538c714be35c4e0d5fb48fdd61e8c87623221e0cc54c0cb2d965a65", + "7bea712fff192a7f3aa6b9912034a45105df5c463776352dceb6e4028c98cea4", + "9fda7b6ebc0daab7a58608add00f474d20caf20643d5194fe7afc4041d918ed5", + "eda124bc784f8910c592173d278aa7505a32755d919d9916a0ec78d12081aa95", + "8a280ce9e22844345fce8a6d322a4b2c5a35325b8c2f465cd3a023f4c32f5751", + "44dff427ae569868ed7f7d8a6019f9acd6eeeb0419f46a255e261f10a9ce7fb3", + "a60812e4793858f7414f0ae57225234979940792a0427863504b4f004f9df9bf", + "3856df3ef2d2145b4fbda46eda95bfe38c52f15576bce403bf3f074d51cce8bc", + "9cb517d90e222bb973b1d7cd23b5462bc934064c33763faa713f8a506ee66536", + "2e44adec46a035275000443c8b5823ae4fbeaea9dac5809ae7d1177e6593b72d", + "ce0486410cfb76699b53727fab459cd6f53a9d1a666aaa55d479cc15cc2c2326", + "f814d023b3f37b8d52c88314dc3ec389d3c9dd5efc219ccac078e1d53e5ddedd", + "2c17f4a27c8fb27d9163a49debe768550152d41d3bf99fbb74242ea4d6d96736", + "d8d0ad3d69787f5c0d15a130107d2fcceee9d374649c0f13fff66963b3cd6ba8", + "33249e435c044a6b3660f95dc40f4d4bf33c3fd49a2d2b76fb847f8a11c4a3c9", + "4b11fd79e527bb16827c2cf0d28343ff6d857dc1821e15d41540e599d7971600", + "58dcd95e18d69ce5ae183dbc540ecd0f247e21e7b5df5284b8d3733a95b2b427", + "a6380f945b07b68fa56384ec7ce4a6d72832352f669985b3ec8d7507583ca1cd", + "d4327184759d0c2f6e7cb6c8c18661d2908fd63d24790e956ecddc2efc1d6d43", + "6feca9afba385b50b1325a51044666bf61979dc958d72e3228b157939245ba4e", + "d1c529d7da0660289b48dc11a39be5cec6c56e054b084890a3e0c1c05270ac8b", + "7aca12e4c020c483b74a4cacd6f2599a182c48e5650df1016b9bedacea1b38af", + "c909cd6bbac64c8e6e0b2754d8cf584b4487eea9d8c2b511a0240bf1298aa0f0", + "1ceb33b9fb23fc1d358fedba9ac95294c3b4f119b2a187cf153e67fc3dfb9fc8", + "e5018debe54c77f98e7a4e31bcd5766cdf7669044890b68671202242e04dd06b", + "e0466a952408b959c7f874e7b6c3b7161361a92aae99b37ec7e363899a6abe0c", + "1023daa051b07903debbf56e09c043ac6ea8e77dc34583c1221b38e35f0260cb", + "f68d644cd8f46420f36cb6d70cb83756c6935a85949caad3092fd1f4501b67e5", + "454c658aaee51baa13d3ef441466709b55b9807119fa0a74ad6353b2a893f485", + "73b2a19b4056e3e0df94b1c6e24cd28fbd9a00cbeab8039185712eb2221e3914", + "0e869b3ba6cddc9a0361e7eecd509f4fb1a0cd217c1e9242ae453d21c45b45e9", + "b6dce9e87ee708bf166d32fa29700725fcbe8227a75d26e52a99d4f05a668e23", + "f40b3eeaab01b34aba3ee12e1d1884eda95216d935614eb1f9d2ff0014d855e8", + "63ff7ffe9262cd7240a24f24fc15455e9b847478040404b627485dd464fc061b", + "8b9310a40f5573137a63d3532c94edf0dacd10e9bf5d83c1705be14959c8b3f3", + "c9a83ff03e59e6ff2aa2beb8bc4369ba157f5cb73172dcd0ee646166aee64d65", + "a5b7bea58ff79e9d052667d440ed966232d63afff0bbc530829eeab8e302912d", + "032538976faf1968a0b23c7335a6471cd06eb954af36ca5cdeff64866308facf", + "029f99d5c0e9615c07ee0a864230962b2d0b4ef7970c36ebcb58863f37ebdeeb", + "53ecccc2528b95450bf6c9d614212f57fa5bcc79101c26b79b307b6b3ac19918", + "b21b1aedc2e32c2e7a81b454467270b9e79ad15c782407153dda981262a782b1", + "78faaaecdc3439ce2a0b6f46f911373fb73fe833fffa64d321c1ff05de6cddee", + "41b6ad6d409ff91f524e0c2e779789c3f8d83f1e487e0208caad2598eccfde76", + "67046cc25a27b93e10757ae9204bb15c9309320979ea87edb9a91ee79e318c46", + "8cb1c91305b3811ae811b1b08eb66a8ba744d6b46f0d3fc96621752868676fd2", + "60ec2f16e298dbc9e73480688b13a89ead9889ad3c5bb2787637a050b091bc5d", + "a4dcbb5cd984176b407d452aa30f47d2c9a4fcfcfc514dcbd7db5de2d832c39e", + "4fc022bff2995e4347fa523551af54ca5679d619a38a4d4b05402d899628d446", + "1e8fef4080381374c59f7649da8248974b6e5dd9e59b9bdbb36a0b9341b07d1f", + "18806387c65333f7cb8bcfc78b58be647f3239c8202116e46349267a958051b0", + "594417fed321c615ebe0502de158273f420817617e8317ef542e976b04431394", + "88ead0496d3cd487eb7148beea5a2911951299adce6f8c1cd0f05f68e9575dce", + "cb468d319bb5559a39322381a46b5cc52d94dd873a391e1f725786c7b387e4ac", + "778042da82b1c0d7068b219627d1a6f3981d1ea29c31b386a6de50a0bb7ef4d6", + "b90de98159fa65d7ae51eac931e1e14d1a8918b779c820af6465976c4d2a3f18", + "fb0196d94b6253b46053c9e2e90f0cb5fb210ea812a61c09aa24cdb4c3e49742", + "1e0e8658108f8af1bc4aca36e0d24106d2b007e61fa527e84b7a1e01b3c3951b", + "3ca7cbe967e4de9de54af9ebafe3e61a92d97f8472713dcb7bf960bded037312", "e64428f7997abcb5b5d77e7b8c77841dcacd5b1e576cd5aa4496a3e94d59a16d", - "0539d7019e4d47822e92121b1c1de118890d844ca89064938df8b8d86ab18b2b", + "369324b6b754bc2528792c507cc30b3c787acea33815c2af6c6d0002a1358600", + "741883fed4a70e4d024696e40bad9303231f27576316167444ab23b1e638c3ff", + "eda971db9cd248cbad38627a20aa1e0899a65936397729f65e55ccb064e44bff", + "4ef8e8676f51cbcffb10f1cdd6774a6f14f737ae61f32680c2983d912370e019", + "97e576b1c0811b79195d96027f86df1f3d2ea9bfd35b075033c7996a01f514a8", + "cd3cf317662c1d934242c0d0927e8417e64ab879e591090cc4fd5287ac06b87b", + "94b7359252dfed830f122de38417b6fbf91c85f54df951dd5dfb68b45336c1a5", + "31a2064c7a46cca5d9a31b4ecb3148fc22d41a512cf5bfbda2d764feba663f97", + "197d772abf424d5ac89e0aba2a18e0334801e3bc4a9ddcdf66c3a3c56e625266", + "b4af9eae95bb5a2de122f8cd9dc2822a730ab269a6c11cc992e2510933853d62", + "99c51b1a3f423a90051e6f6716750b9400d6c29acc0f5ab52d30b35eaa7bfd37", + "65af655c484d79181e5fe36e304ca9729acaf013f608aede4665800af37c4f97", + "a7038adca3117c3104bd51cb5f4e3620ea1cc677bc41f903b329039d3e518cf5", + "7c30f28c5e93e92b04a787f1b4e7de815525ed9c5c2e40a124767fde7a932346", + "1bf2dcb0cf86271ec16bd49677e92ce5ef638138bb0f0f402102c1ea344b91f4", + "e5fd0cd1a0d3b066e3b9d9b43609f271c39dd44e33225a37454fb5226608d7a9", + "85b2d949166f4f45cac91d3c3da9d372c3517a5c137c73716287d8b7a2921e3e", + "d730487bba438d57e93b81ff2de947aa1c1a78b506ed3037ec40c4f66a3c736d", + "889bbfa5236aafd0c1378e50cb253aaac726193c319256c3a4ec1036ea2984b9", + "39b30b25007d2a2fa7390dbd0262042970a0d799e3b3a855ca8e81e6805624b1", + "c99eb02b7ec4c992bc620936075bb01ae9e5cda91902333ffc2f5819b1d6c55b", + "db4d1876ac38e9d1982d13fa68d895731d12b41452dec832618db485577dfe17", + "348f79800e4fd872172de23cd6e4b8e7d2d5828d2f2ddc429ee6c6e3f4a1e6ae", + "57a304aeb3dc71cfaaf65fc4cbc03d8e0a9aa6f713374bb77778403d0fd0d140", + "61ad1aa0aba5955c58ee93b80beaf219b64a25cbecb5d5950620623413c0d94c", + "b9cafa6fce2dd9877764ecf333bed12b3a246d76dd429bb0e55355e6b889c0a2", + "3096fe6003c60d31fc0b22612240729ac17c4fcddb46a5e48c85804b1549d115", + "0cdf8772dd2d6439e05fbf8f5de04d1a71612e2fb2c64e70938cf815438704a3", + "ccc0cda28baabdfcea04a4469e33fcbad428ca5588168c5ae4724ae1410baa08", + "5d8d421462f8aaae31787f49bfde9d56116cae7cc1271731fff8385d040fea73", + "0984181b5c87f473e7e0e8ba9f2a3411d9cb320d1d026f732e2b066b4e845286", + "5fea956925bf5f49189ba0cae98584b7a94ee593495a2b6e8950468d8bc7d927", + "88056b960487ac222d1a092027069eb0de5dcba08463f9ba9b1174a5f213a603", + "03938d05d6e689c20c30c3c67757d28e84665e98483e8add7b2709dddbb86634", + "021360ba27c1b7ab2ecaccc537ba8c2d65f707c244e05ff84a579eb447f00a17", + "505cad2ab8be3e144cec7fd8221b61449d58bed3c610691563f0d416e820a811", + "fc2dfdf7b4b4cfc8944aaaced50bfee45e8375dbeccbe56590aa41134a1941c5", + "c464ac8b8af79a293c39c6d9577745ea926c1fbd6a3058475293af84a7a8ceb3", + "5e2a889be7293fcb0a2e639c4564d6ec609daf5ead4385198a10d6a381d4125a", + "add4ce9225aae2e7ae393731bbd87fd6c8679fef9e079dfd9e99068bec45aba0", + "1dc4d3e19ac54bf20a7bc4884a7fb2c5da766c6254a0d5ed3ceff6e972a5badc", + "92548a97d29c0c3b6c934db0996bdc5e601abc4ec04f6aa1d2996c8a1d7da103", + "1ad6d9052872281c6e03320b29b03cbcdc99f8e5f52dd35ca938cb050086aba2", + "61b0779116497b4c0bf963f74937d9d7bee8f0377c86185b7366ad8fa7b2dc8c", + "dba2343985bef1754de8ae64e5c534453bc136d265b9732ff66d5324a803b1dd", + "03969ddffcdc2c1d36bf1d1ad53f26b9ac218addccc271e699587e6b95855b10", + "f8e4b1efe03c9f0210db4e5034243801a2c48a5a46e9eaa9773f97050d7e3628", + "2b4a5d769102b4228c7688718b474fdac4b3360b294e45727aaeea0732095bc2", + "531d47ad37a60a9aa889653df3383f2e7f5d697d0ed94a295b620315850de974", + "5b3e809f5146142fb1a7341165dd72a87f5e4ec0b3f807addb51c62a15423924", + "0e73d480761aca7961b4a284db3839cce95e21cffc703784601328b206fd2bb1", + "1ec3b18d93cf2a19a65db9a03806df7e8549fd6bdd57cf21dc01fa1f1558e2bc", + "8838adf2be07699c3b3c79d86869ca024f5bfa83ca08943964500f3545c8d712", + "f68d82ddc2728d4e35d2e493ba546d30424ca950680778852b16c6545350e5bf", + "de87ee56f58e3ab4042d0f49464a2c8fbeb16f41b955183158507ce2085cfc25", + "19100463a955b0f84c03ab40c8fbcafb74369ba3cc0b5b438c2a19b3d9c5ea6f", + "fe029c2c7da088f9d3b88043852b1a6e1fbd8c963d656866ff41fd726624e425", + "69c04bc9d83b06d52a88996286fd47366668d2f3957adf6186d225911a97300a", + "00b9163319af184bab4c9de88695b44498af12c8a400cadeac6c115c8cebdd63", + "50418f430137a1c0629719b855d2b35ad92fd0994a87bb5e97a95e7629f60fb5", + "046ff529546644352a9ff73c77271b23938ff18e0a0705a05643b9826127b3ad", + "0c69615a9cf3515db538f0648a1c00b17bfef7458b7f5ee0e19432b91a651e67", + "2bcd19db30169c578009ff172619b9982bcb427ba63ab4c2a90424f6e05f561a", + "75f275858c12d58188f7e444f4fc758986822bbad995b3ea6cf2a7a313c4e4dc", + "b4c21864550f3a686c082f87d6398e71a642e34ce16621bb1b02f6aedf757beb", + "d6a10c859e44d75d3495da7112fe6f989ba523159490f6858af918fe2aabecac", + "58c21ea36e8e1313f573e8eb5f51a53245d053f9a2dfe397e7ef15590e56cb18", + "d1c655a2667c2b85c39a88097b98f8e1f2c2327297324373ade5d64535df387c", + "93eaf193cdd56a493d4e647788ccb78ae2ceed97a7a87d5683519847eeadfaf0", + "c38992e93a7b047b950008c2f397f5be10757ed281e229b3839af444a05f705b", + "daa1d50067d3f7a35a4c4b4aef611cda4ef25cb2264fe406a493dae6275fc104", + "7d42ad9cb15ab7fa75c96b4591e6164ef5314dd8b65e61bd921f306fa4fa27f8", + "2699025e78055187d8ba606b6ba23570eaab68bd106e833a569be80169978841", + "588a525ed2e6b059a6d553cb6be74b4445e59dbe5c122809b7a56ae4c88b7ead", + "f7c46353a077138c0f0a252c9c83a7bd776ea1b2dd3fee8f02cc3033ee3556ab", + "21ec6f9e9f1a2b3b1bd5940a7b5ee2c88f9cd98f7074e46ec30aa9bcfa2d5acf", + "7b0fcf3083161ceea0b98a21feedf825a98f7a755bdbc23993de777c0e99c9df", + "8df2346679f6d59271180e74c67e797f11fd9bbed8a838977cabb0e1253ffbef", + "91f56f47ea5a762ed90b02e497014f86e4598ca4e05d253a5b37b79627f3c8b8", + "0701e951cc16057b03b098274f97ceee6c308437a380a6458b1caee3fe86d255", + "f9716d0f01a4b284864c6e88a7decdb9cde8cac21d0b76c12fee1a817cc5aceb", + "24521dc72e67698c03e9e2592410c77a806d3508308e6704b59515b81a44a2b1", + "cbd29741cdc73f5ec4028b8e2b3cbaa51455d3df599ea64328ff073390778969", + "63d2b337b35674f7823f752f3314164c37327bed6745bd4b6aba5cb322709e6a", + "2f46a5519c6c6a101c57f1fc7f04c282cf1c3b3d980e1c89f9e5002725213f02", + "76997327d86762d2e9074a82c5f068d49c63eee3fc9ad27fd7a2d0ef747f80f0", + "bc118aad247d5dbe56db85316e0ca34470f6b334176722ac4654353889c2c72f", + "9140084d11a07c8d2044e2f637cb0d9a002a09314ede4a7029d9f55ad9320f7c", + "deabfeea7b009e723a1f55d3303f581abaffa881b6ea373c5c242e56da63aef6", + "311c5091dab1cbde86a593f4e30b7172d814ff38785488c3e611b8ed35a657d4", + "54ec705e69ef9e37a01c545622ebc87b9e32141e58c06bcafd3837008b08331f", + "030976c39d0c4c3eea5ba24dce4640dea7ba4d95ee2b63b3befe0b55a48fbbe6", + "55260ac2f1710ef595188c9ce178684fa15f9e5cbd918325c14439e967b093f2", + "271b5ccf35b22c4fd72a3686079f45cee22b6ffe816f8b47eecdc7fa34adf91a", + "1c23fb8a74470398e62da51bf8ff3f6c78ddc3441d3a50312d90a184a0273fb2", + "93a84a7703b61dd405719499d833c4cc065b503ac1e1bbdca69f88c810771721", + "f195507f950c8ba8343c3e623644108d6ab3ea4802571ad9f888ae1e81632089", + "fe0e112810a0e5235a9e343495c2b6a0dc83ee2d1e6ae0e1b88264f7b366d500", + "9065847be0759166e2c96a7bdc0eefe9e189b54795d2773a5f96d71396a27992", + "defefb71baca3e7bb4650c1b948ea8e436b4938d57aea82d021704fadb47954e", + "794a564270da178528a956fd041737bcc80cccee54033e894109fed6c321b532", + "609d06d9827c058eb94e6f81c493d62401599202d25ef4b02bceb0048367e335", + "ee3c8b96d827c26d78166204a3957ecdc44144daefa6c613c1df557fbdc7c063", + "78abab345e65ad3d36c91f493a6a2ab19c535546a502ae6fd9b5af463f36c296", + "e9ce297a6f5bf5a1dcf10a08f6f3f6c84b2cf6bfb173512f9756d00273fc588f", + "55dbf97520c67e09b53b1b5a0bc3ee320cd4e2e9f9b6ffbce28a11f6e49373fb", + "f29c772d6d0d4d2b1c64a5209ecfd289fe9af8509a460a84f3c2553f5f4c942e", + "5c3363abdc1b883c88f526a2daffed7409ad12d1f386fd0d9d67064a37fb91b7", + "eb8bdb80388925aa20c295cdf8f2c2e3dd67951c965e02c2ad02e9471bde22ef", + "e8bd7b2cb99813e5c57bb6b08696db71a01092fa100ffec609b83db535f0def6", + "6e74ed36850bc318712cefe7fded67e0a1f038e8505c548c2180b9d397d79a12", + "115d6ff512a7fa750b97d733c45064e682dcdf791e373121f0a25187f1aa9a15", + "3eff7aaaaf4c1e790b8b58bdad6a08b8438db38021452c8f381138b4c80b5fda", + "c925bf0de664b8b996d48a3cae873dba7d1ce0a795b5375c496a86d5b12fee79", + "a2a0d6d8f4f1e5f4f1075e407b458448f0264e3f20308ae5079307ea16fc2032", + "451591ecabb33beb826e129dd0df7f12b183635fc89476d4a56076342f98768f", + "a392d90630e8e3241a1f877e82b4157844a8156b8d484e0086c5989abd36ab79", + "e34239bc64f360077df32003dd9fe15934ed07af3452d8f2af1a2bd5c2cf8baf", + "7a1ed0d921f5902e69661932a925cd92ecd13e8ec313275a10ae2ed4575e70ab", + "12f9ebdf88090b9749bd7450c60ee8b6dab04e777bdb8b2d4341e37580e7398b", + "544f0c8a2c6b5ff8bcb9cf521726115d6dcb59c843e00394c75fff6192a2564b", + "4270f1438c0bc2da121b25455f606ac22167d540e268bafe1a229804d146a617", + "5ca7e4258892f103c5be99fab8837782ce6e5655675b3eee6c30fa09759e57ac", + "5f3c09937fbeb6047cfe65069b4bbcd573155569925b8a03085ab6bf7bdac830", + "a6f53d50a17856ca67bf01b346a244fe3da1f0d91781655b771676f3ed446245", + "74b63c0af96113fef9f9990b16f6f5917248da3c03d980f0f5afbc1736191376", + "e7f7bf9be34d1d72993ea60d489343e7673faefd67e6cf3b9541accf57897cba", + "0a90a2f200ba69e4cc0956977f19e3a10848a0ee7b5edc51d55ccf1c017931ab", + "16e5f6c093a1d94fc955efbb8a711fa44c1368be75b6257d47f3fed560cfc8a6", + "fd901759795e126c0e5ab27f3ca066f712d9dca9031b0874da74515bae34412f", + "1bc145c2d4653df22a7aa3b588897ff5e4ca4524be41d430ef4209af7fa364b4", + "6eea41de0c8909f225e83dbb289c4176682916f6d0c5edae979e5fac7ccd544c", + "c06f1eee467f0765fd181acf11c69aaeaa1765492a3e2418d8039b4f5bf9c128", + "f6c81bae46f6352b9f3e2a0db6f6a987008ddefdb91a6488117ef160bf14fdb8", + "4cefc9bfe97cead2ca7ffdd9ad07247e4cca3b5302f29b1163b8d37a85355517", + "37b2f913ab687884050525f2aff673ffc514a882752b84a54718323d54a85d38", + "8824cc38a3cbd18f8e0cf6d41500f134ccdc1550d3ba48857312acd9abc33c56", + "b1827f3a57f204898aba6763a1615929bf7a4d019080d59db2cd9be0ae1d07c4", + "d89bd6e39ecb003af3755786f2fcec685acee689cccbdc890a7bc8941b518d51", + "89905d021785853f3d0d170a2ec49ece6f7d7b8a204e2025afbaa7bd71a65d69", + "226e91b4766b7240498b7847b6560007979b27dc306ff0392b2eaa0fdf860ad1", + "550e4a3a92a37849d11e9e5fcd6448d3f69b59a6530d6a36127e41ac4b2dbdf3", + "6e4bfa174d23213780d1709f76066ea3532697fc32f03a3b85cb5d4e24a5d5e3", + "2cc1184cd5051556a4dc1c6f19e83622f8a039cf8b6030b41f330923ceeea87e", + "a4bde268762c8e4f8a9db39372d3e5d10772e6b5ea591a253b5674930062b92f", + "fc3d970829dac7850ffba64937126bad4a232631e681dea94d271614662817f5", + "ef573dfefe33236a13f3d91989146451f922fedb0d04b830c6b1512e4b77424a", + "f4e7dde438f3a6d8e90a2d8640fe54f8d84579927c6af265b7274e4b68f6d5e2", + "88ed2e3b212514c0dcc7338a338b943de557bd902dc47df739ebf6d704b6ef1d", + "8f8b9a317ef64b42d9959113b713630ed94fe6f62875d38b69a5e037a0dcb38f", + "bc91cd337949bebae3b7815a213a27a43d832fa06b5bff74565fd24a69013684", + "620f0bdd54fe6261d64b7a0646067c82fc9ae30fb91aef5104f50742d2d80488", + "f38afc81a9eff441c6032a5b22745ff248722a7dc69e41cc799cb78a59fa7b2d", + "554d1d14128c062195f83e3d449668025da490801bc063b149ab7276d790ffd0", + "3d5b86dbadee4c09b217edc6c4d2633770d49e378916b96d24b6aa7f70479109", + "4e1446791916274b787c16ae484ba49620e75aaba40d904f7015964957a30f24", + "b7b22d28cb308c74b883b925518142a342b2c05f09118c90363504498063b071", + "2c60329e905ac11ca44bdcf82337d3e5409f63448b7f3e8a48927bedb9e8df0f", + "1ddd3868d7d726b33af7d906d1fdb3504c7075c64381b3b784406cfa52fbd39a", + "06a9c68acea2c7b396a95cbed59395d975216b57fe7acb7d696fc773b17faa42", + "8a6418b4596a8048ba50ef2a98c0e6a75c5cc134575253ab799e6bbf733b404c", + "e4c1143309319ed0af48c8ceb223302a222a14c29080be242460995c35088f9f", + "f1237c22e019500747e1f63dd82db1cb3e576a124760a337997234f95d71871a", + "35a0ccb2f474fd93f9783508a46fb6e6f66eea3d449ea26311affd7b059f1c89", + "2d4f3320a3a9a9dd8f2117191d83d4d2cc10a01687c812e307b2b52c8ef8c809", + "a65e61f14023534446b5ef2d88564da1cf42d08b66f11aa55b965c6f5ebdceb5", + "192d54f1535058d04ae39a455994d04e31145d6daf61444841694c98694e8591", + "1ec1188ffcb26dbe697f17908ab6d9ed4d2e59099accfe7406f643eeb13ea5f9", + "477aaf021e13ca9e9edaa7bf69ca18ce35ccf29fd5db56f27d8570b490ebadd9", + "e98dab0f23e23136a856d30623bf1688e5f84da619f24e1321d39204bcd9dd31", + "c57cd25f7ee7c047fc19a6bab4eb4461b8a646b8edd5c97eeabb03857578f47c", + "d4f5d5ab8487c0fa12f43ba341e4fcc1a9b0774e8a50f31bec3cc0ba19a8f893", + "7b6b18024c360423614a3b8679536b07f846558dfa777ca23388d81146097001", + "3e92dc809a90d7cd59ab0159b9b479c3116070a32984d53d1b3361561a196b53", + "d482a2520f571f0b4ea87dc0614217e04c805797167a9af4d395259f89ec4cb3", + "497c4a50fc9035f5a1918bfea1ef7dd8d562851ac5c543394d1019b9373c4d26", + "de629c948116d5bbb7c22cb34fae302177c1dc285a05d436f9f03b7455930800", + "ef854cb958efb8fbf50215afb0e297e218094f9dd638d040fd0a51ded763ad41", + "6ec663717658b669dcc49d13c06c9ce4090b4241ab7ed0dc2a53e07590a61160", + "3abc189f7d7996970868858004e5f457ef0a68d0f16740b1e7d4dbdc7f7b7b97", + "f484f27e94039c70312bf9232ec2958086a09ae9bb234a42060151e1367bddcd", + "dfc771d6cc9194a8adad2fa493b23481d359e55efc7a3d0e6381de76fbcc0551", + "d1bc9b36f0be33788c873b46d32b89c22e439786df5d0a4c8f0b1b10329d2f99", + "e06054b2586c00c8df884c29c6ba1cb44f8420eb774b423f28655dc7e844a642", + "fb4ad272e5b215f8aa316babdfdd059ca87059205c1c06d64c58988bafdeac62", + "9ebfb1a90c896cb969052c170981496cab87c5e2b70085ebb7fe9b5df5ff0374", + "2239051d42c3c9a0344402e18842eed917a0dfdff52209cba8d48e76b491c1b0", + "80ba29915476aca182789215a8223ad9748d3acdfa3dc336d47b32c21d9de20d", + "84daea976640a1d3cc62dc708aa9d33afc684e68955952c810c2d8ac87037cd1", + "f79ead260c162513fe7c1c79debaf646f5aa8193e5ac70bbc416eefdd5ce1de4", + "5c301f93488523bb200d68fdd7341fe851321477afd7d67b7ed45e53051b4517", + "d4542b8e2c82afbe2759ff5fc458328c171cbf386c0739791817602556777890", + "243f13277ad70ea74a80e37efa3fe1ee405136282e08f8b25ff780e8b9b58be0", + "0be44e0b98b8e2646e7f9b74420d75e4cb237621cb43054611e664d1446820d8", + "0d93550b2ad29cd294b9ad76048f86d234cc3076a8c7b7b33a1d984d1c413bed", + "97ca76dc6b46a2fcb1863974d47e99216c9f66af44b5962e92fe34ffd0a370cd", + "a589e87ff165ba50f1feb140e8b57fa95979abd55f8cdcfadc31d6cf48aab4cc", + "f8d2204c38bc0e57a4ac7fcbc337ecda23afb274d30fd0c1234db2c4851fa71f", + "015af204d3e97acb562ebbeb830548f8640a53806cb297bd46803edf53eae070", + "4c233fa3835ace45b6ccd93ed571ddcdbd58c8115f036f651034b098c5629440", + "c7a46390162dd76aa5bb5d101ed28a778d4c756ae3dccbfd4e6501c94d0e4e21", + "eaa13af5de38e6884e02ddf552fb4496f499117c3797cfcac60d4b3e3510d9f4", + "036ee345f2e4b61b8fef1b5b64fe6a84c03fe4ebe1639004c29f2cb86731812d", + "7536ca2210ae84f6e788fa8243063ddb66a9b05f6e88c1a94a23338940d9f557", + "ff576b21edfd9ba6a1109e5a509b2167f8a8dee8c221c6285767594700e18fdc", + "64e3fb7cecb3f198fe952b2a19d82c4f01edb3e48a2f73f1eede49f146dfa0ff", + "e6b9ee673b637157a0d79b31d0903e2791937fd548ea09b6e2c3e55071f1d65e", + "b10ef217e7ee00f2d31eb82a56c614ed5421584aa347dbd39c8375f61bd45491", + "9ec79c84f5df29e0f7aeef36ca8262063db6ea48100bf70a1af24c3b8f662a51", + "ceec7c318f697269aeb9c019eab84f94b4297397f231d2899aca996e8f01325e", + "5802fde3696ed8181b85b7d4df7ca20b58ac6180372246291128c3089a08b142", + "ab8854ddfaaffc972531640564c400c46b49441c58c9580cd0db54d59a83bffb", + "4caf1ad209e31f2b6d9e4ec75a5a5e3a6711c768f4f4108c0c70bd6e7934a933", + "72ffbd3f946d92b6fff9d81ec446899185ebb5665d3f7836f77e0118cc8450ca", + "35eed8612af542b5dd61e60916fe41ad44e714ae8166fe946c6f780395ec4e1f", + "883aca3e7e5f1eaf423317bd9d2a288da23dc1f11f10f5f405780055ffacf940", + "331e791fed6baed394cc93372f17f449b761d6f1b69e77d5fa5b41820c3b722b", + "d1840b2a3d4899c1fbf44d908464a3eaa848f53494a66b305e49679de84d1b75", + "54ea65637e0a63be410a3c1affcbefff88ea066990f16e3823a8487c16ec0a69", + "e5a9e8a3e9873b469339b429f4ce5c1b4ab0c298b26486441343f4c83aabf1a0", + "b96bec437993fdefb6db5be59e21e7d907934f734712b8828f712cb4e37f5896", + "5ec20de3c5daeede2ed99679f50f0decb347ed662c78aaa348804fd0ceef6225", + "86cae61081dbdcf6e6100b7477e4910901c062c2018d3e325b66524d7bd21c08", + "3ab451e867845c43ecbe25e41ce46512680ad39af828b9afa8d10048d8dbb448", + "c97e20ded7eaf8374a86b96801546892272f280c6cf5eb79333759a620cd11af", + "fbf754f098f38f8330abccbb52d08dca633b5563dd2d3cfc92d2e035cb24042e", + "9c602bbc36b4ba7a87cce64fd78d25c09829d3cd9ca49062d03576e3d9a5ee0a", + "6a79b15e9673609230c47698ddf88773221b6f6ba5597b4a1aff7b03ca99a628", + "5eb0f7409978198397f4122a6fdf5ed30180404f7e99c4e647c75b0f2f0f6ac7", + "c172354584d21d8df4c566e3f3209aeab91b8eaf177ebfa60c7868dbb0259209", + "ceb224d4429165f4bf3cae3114a4b80060fc9be3cb5326a8aec3fb481c7ac8ca", + "8ea1aaf7a82ccc13ef2ddd08c69967a845d36dd66cf3dac1536bc77d47e910ff", + "46dc51fb2722dd8eaec67a02c68b66277967a80e425e053a1fa7d7ef8f632781", + "2d1a44bc2d8384793a430dde24fef47f28e49912180c001f8693ce1dcd153484", + "cbcc3dd763e2e9c022950d5fb41888f3b83a3d9354ff533ab33a82a068e2a746", + "81e5bf2c593e46a2307c29884c0457e4866ba032aa8e3db72c79a47b239e2cd2", + "a6a8e1f8e515ac34f10ff017e88cc3a3f0cb60b1d90c110f16470a8b7aa89e21", + "8a41589d32e869d30f462cc28e9f508345a100f04d29734c0bbaf0b4558853cf", + "4db67621ff325505d93adea22b733e91c24389b35407b37840cad1fc992e1aea", + "1f2dda64ef3b71173b60859c0309889db6e22ab0a3ad6207b8982c483b973f54", + "9fcd7f6f0e1228577698367efc8564b956b74738571cc773f952f7226b906c84", + "5d2ef2e424f1c31c05af3cfd8b474efc245a166683dc66fe5baf72705bf8ebda", + "3ed0a728cdaac4b038d9bd8f55d3edbe7df8a42cd7f158401dfd7f639f9f840c", + "ace8a2c9be68ec84e2a42275c7e94fa9efedc1f53d3ddc786a992c4d5fe548c3", "5d948bb5164ffa7a59fc792646c1e9ac45e1bf5026fb99d070193fcfa97f7239", - "300ad7c03c062799a35e41aba179ba97081471c3860a1df1e2773450763825bb", - "6df7219f6da1ead40dd2ef106364d215459f051df739a81ca5a537befbfe4a80", - "38f6a33f4852a31e057aa7e9c3e89f64ae317c3c0cea52c4cfdb288e8be0ba62", + "8c1e1c5073bca3f08477abbf768a75643ee7bd005725d4f45d827bbc6f41c2f7", + "5bdbe750083788636ebdd5894a06c2ccd05da5942b791c5a5840c47a52674f83", + "0328f305d3c661f8c7742842a7b830d47aafa998b1f333e249f8871c5fb4b3bd", + "befe24e5985a08cfa4dbc7fc7818b361942409157f94357f0faf8a824f4f8fb6", + "9b3477830b9a7170b550750f26eaf475d10fcf9a3bf08d4f9343c725d2efcf06", + "abca79b44a5013922a83e257092160389d77dfe5462b5f16d936d8fe9be2acad", + "9df4285dccbb039393ddb8fd845b364a0f0bf6f9f3143909f6855e97a6a13177", + "d292175bbc55e5dab1843fb8a70bdef02d44e468d6233a1989c4bc956ed47d87", + "0add51d6700023ac5410ea8e51fb9196e64a761c4ec0db92e1e2150b0f23f0e8", + "610faeca6abd5d55323ac9acb39a1c0f499fa58a071814ce5b8b768d79ef8a81", + "4d5dbacf6f4ca99bb472fd9d3843ee24d6209e3acea517426c86b4d6d65a2096", + "367306f71ac377334b36e9a69e8146aa5a0ab2b54856f1ca5447b6bcf799e940", + "9cf233e16abc417a1a3e73eb792972db5fb84efe4ef42979fb6f06609f9bce13", + "86553256e9afbfa3f14646cad81cfae2c740020ac08716fc1b7f223f8f12b959", + "c349ccace509ab0197ce8479433c165c96850768111c4010d064180df5b7a2ab", + "829f102c973854bf45ac3283e3e99f6b6b0799454dc8d6f7ce296462ad360af5", + "655ae8c318b7bcf254fec7b1e3b20319f11bfd08a675c3256bbfc95bce6668a4", + "fe2b838130d24c6293b79560f40d64ca01f6b62a4058399a96ede31c21f2650b", + "33965655b08acecb901a4cf7a1c30e7c67aceeaf0671b5ef8f50e4feb62231ba", + "57d46b29ff6e84d0e81f7ded9a8ce791ca1b03ee937c4d6e6446f8115605dfa4", + "cce510c31d2f6fd54c9be8f15f5a46cd96e30213ef23af584c395c3042b0305b", + "3c455b67d8b2290a6ca91a6777602a8c2f992e735a7fb04d198d47e3da137bdd", "1c426e9841e9d4961dd7f639ab3fb8a9793b96c98442b799ae5c5a1443ca5807", - "37b5b6c155e330c6973ac70c21a3877cd74461337be108a7267e58e2c5b87788", - "029f99d5c0e9615c07ee0a864230962b2d0b4ef7970c36ebcb58863f37ebdeeb", - "85b2d949166f4f45cac91d3c3da9d372c3517a5c137c73716287d8b7a2921e3e", - "2f4917c70c3de04280ff21a9d052c6234faa0a962605d450f591a949722ef37e", - "8c8611a2be6b830d5e9924deceaf6105ec63ee7ac7a510f59ea965b9c2241a13", - "9ebfb1a90c896cb969052c170981496cab87c5e2b70085ebb7fe9b5df5ff0374", + "d0677a8bacf44c8d9b562c3b09b4416f00039a34aa42828da06abbaf3da74232", + "a73fc58072358cbae50ae292c7ff4b66763e5a7511c6a0ffe219f3f6a4b7dc9d", + "1ba2db2808f4ac2577ded1554bbb7bac56eb9dd531492c5fe78239606710eafc", + "01f5cb7dc8eac71c8e5e6b5d75daee81773238d241f3df7a0f21d8238a2c1866", + "0ced85ba0a836411b23d4ab80bec4faa039b1a3488bef690d89d9a07a94520be", + "56b27fa81917399a4b52c5f2df7edba8f2cccdfb3a730e6c30af3b81a2f954bf", + "517f8d352f41534a7556ccfc1354f754e76095b129ebcfe737c939dc18d058d6", + "f9e592f2c6cdcf429718dd0aea3d7ac22daf0533a0b6831f00d7a64837a0b569", + "44e4019f351b48d37a3f1fd2b7082859c4ef0ca550579365128257d573e4c505", + "da986b9f2ea92309098f8d9f12aa2b3b145bc1a1d9eeb09c9ea8b10a072a22d1", + "d7608f4fed5dce08f5a124b4731587e73279baafe5da8217569f244d3589d233", + "19d0b5abc624836dd4e12338e5fd5c7283b426660ad88313ed9f9c80f40e9aea", + "a460bdf143bc86aba013a568c677e9fa14585652d5aa2ad7e28ec50baa083ebf", + "f633a849403e05a09d450a1c4454cef816dbfcd99303cee920c6aed8e9f92a24", "d23dc826040868cd7f9ae866401a21170fed6da2ee874cd94cc3797b64797c45", - "8eb6905c8745867e38d145621aff4e8d3cc234b4747a58407374e5eae0ce9c42", - "87c815b695f5fd65e6a52662e61390a521fe493e519d8addfe8f1635dd706688", - "8d0d14ad13eef4a99a45518e8a33a8e634c60c1a489c61945a564d9e6186f27c", - "cba33eb9a3530e201115319fc0a26d708f61d0170c0853a267620eb4a73409d2", - "76ce45a6810ee97d6db885d53c13d96272bee4127b003d695f539949d2165f8b", - "2565095bdd107fe67cf81ee78ffdbbc0da8a2b41cd2a398ae878f8f5d49d0234", - "c3ef60007f71cc5c47c4e254929608b227b8a4eccf38cb5faa56227471499a9e", - "7b21c6477b003da97b17f74fea7fe4cf1965714d5583652ce67fa2421e16181b", - "5920843a393e1d6e33548935c1a7683f45c7008283228c05e7c999be6d7dbe50", - "9334a3b8cb2506e1524e0566a1ca5d7638859a862341e1c1990ea54da578fcbe", - "9ea81b70db68cfbbb856bd05d8c5e15e5b7e9136021625b431ba69bc79eac0ea", - "ebaf2ec0712a29c8ca23fef798fb53bfe211162eaaba625421c1afdd736ed770", - "41b55b89297030a93aae34d8d41cc3f87f3989a11a2dd6df5caca23b4086a4ca", - "1ec3a93c5937a41c4220fab70f7998015326197de6fc15cf87dd206d83fe8970", - "a53a582be265f267a2e4c85ce5b39fc0b3fbc644ddb545b3d766b2353f74d1e0", - "fe93284a9a8a78a7ae76ba3421db6e1853f479f12b2421b7bbb6a79f721da04f", - "c839942a1ab82bd684d5cc3974e99e32703a495101452a44e494608e09b815be", - "f805962a4ca947fff4530beadc5ed924a4775263e3495fa0ab1730b3569f9701", - "a65e61f14023534446b5ef2d88564da1cf42d08b66f11aa55b965c6f5ebdceb5", - "99c44d1b7c5dc70ac9c6c238048b11b35a4beee65f9490d9edac3e69d858a270", - "1c25fe3a4d0e136936cfbf5b368b5e0f196d4bcddea88c9d165af123606b17b2", - "2cefdceb5f3d6eae90108ef18e155e960bb5d4994c53e93e6d17c09792187069", - "52f0227938960d972571ba81d1d9ce8f3dd5d145502fcad23e3d047924921345", - "f5b25af43644ec6be54fe6436a4e2d566adb984e50fde33f9855a519fe6ae45f", - "af80e8f2709e71845ec390e9c685d8a255f3899a745a038f6782bc558259d3f2", - "67046cc25a27b93e10757ae9204bb15c9309320979ea87edb9a91ee79e318c46", - "149d2536df467fe27f4d43cb049ffa135d65def147eac860f9919e1feb1fde33", - "96d7be8f8757a5870d9825b48aa2be9c33928bd4ce1a8cf77856ffd461b63fa6", - "93fc0ae77b64af788cb87a196412cc6ae1847e4bb4336e2d34bfc7426ae118ba", - "c3b43da197e879d672d29f73dce7af20ed42bc0eb362babd0718cbc29afbcba1", - "73b2a19b4056e3e0df94b1c6e24cd28fbd9a00cbeab8039185712eb2221e3914", - "bb479896dd6a6af056bdda5c3a162909a6f756ad352269df02add8f766c7c73a", - "b94359d90b4479e5e0898440b2d61572082646ad450777fcb10e0bd730a66358", - "f6c64acf511defc6a29d9adf9427745cbc3e3ecd2bde82d705081fd1d375fd9a", - "42e0938be484a074742f067d544bf4cd4ba2037d2ef9a969983c1d456a55d76c", - "06797182e119f2ed10a561dd4539b7e0af3fe1e4be05328f2119adda98387e8d", - "e2831d46aeef97112cd38c940ec25fa5a1aa6a151c4993e5cc56445ed97b7423", - "2d825027a11f448a52d69fa12d21a37c0cc588ab7492ca1339f5eeb3db399c73", - "37e86744e89d5dbf399fc7f75cb8bd19514bbb61004b30b66e78b116a2bd6a17", - "350ceec4c87bc78d8d44881f385b27bb09075d84e4e8ba17c5367de952e1483e", - "0df15c1bdd6fee342d706831997bbde6e80b6a2a559007eedfd86b8171955cc3", + "2a169049e5b39bb2b3306cc60f6cc3d1d0e5e3d6e02d1e071e875c2bade741da", + "c2b38cb9ae4e4f1fcf6727f0b50bc8fbc23df490bab3934a07d0b42e5c7c5edb", + "a058a3b9a845e7760eaedc79f0b2ad12100a92774f927881d27ffdad3b3a62ed", + "23fc7704fb92aa7b1e897e5f21e4ef9cc9da190384d0c4c7630bb0f26139ddec", + "073f1483fdbde87c939cbc18740535a6dbb5bb5d604c5462c59384aaa75ce7be", "f47910f0be2a1d4be157f759e24b4b250c0ea3b730337653359220f5e820c004", - "179906c1f26a59c5a6b787b3e150628b8eaaa18785f3d7996d4d0f64da66699f", - "d94a0b2f2737eef69d32a395974ac2010c1315f0ee001b146ce7028b09c8a3d0", - "c1bba0da524c3926cf1d28254763163cc3499b27673c255bf9e19bb6a9b2a663", - "33470ebed5c5aa47864af9d0e73566df779addd48199d4894f79f08e62244fe1", - "ab25cd5e8741ad231f6bbc30a238cb3ed78d88c307de23bb2e27debd2d0c9391", - "b51b59f60f89ce1a9a46222fc5077b718876052b027702a5748a337a72e25e8c", - "2bcd19db30169c578009ff172619b9982bcb427ba63ab4c2a90424f6e05f561a", - "a0c7116e91ade0966a782d2885a7f1bb679eb08310f49cf3b1900abd3700c31b", - "44d299f54b4fedfd083e793934879a9c28cc58aa6c2e8d61437c26974a511ce2", - "4b7c2c9dfb6d0ecac33806d2ab47bd3f28850f1b7431a29b01656375bfa27605", - "1eebf4b74f232a59375f2ca949d4a7a30f98d9003c9554d05d0295e4ea1bc550", - "12f9ebdf88090b9749bd7450c60ee8b6dab04e777bdb8b2d4341e37580e7398b", - "44562b926cb36d56d6360f3ccf2d9ec0c263af1355cb6e4ed684dc65c7f2a5d0", - "7e36513dee1f1e045b70a2afce6e8da36d72e52251dd7bf3c68d57f9ade3c276", - "4c97a86ad229dc3da66510f6bd046669dc19dec1256986940259353e2dc281aa", - "dc59d77a11954810103225cf4b046374c5e660db00f4fb8253ffc365a076daac", - "c777c59f299a13990643821aded8e213b8bf14d998ebb2d2fe90f77d60d9512b", - "7ca25a135de6db15fa27448d971708c285c1f628e9552c4c1bc4b6f105400d77", - "55260ac2f1710ef595188c9ce178684fa15f9e5cbd918325c14439e967b093f2", - "b8cda795804b0e1efb249c3247c032fe135d881159a202a5a4e23e91da1ad607", - "7011b7a1643bbf6b4d1f368bf20108ccf92e2f5cad600e287d5377baed991e67", - "b453f669d1f621ead3de799e18d113e14e5afd4089d46248d36b60200336c881", - "659cbb3b90b3824e15b5f21dbcaf6a53e04d629cad62ce1633ccec87bd2fc91e", - "934a7f4b342927add00892592eb7e34521a2e4a5871dd7a5c52a4502bb9d6d04", - "2cea53b03b73610a728e1d4fc198a9de7c0b89bffeb85e60969a3dd5bb1461af", - "a0b1530dc7a0fcce4bd7b2ced3454260dceb6f4338e2b5141ff86bb24fc0a1f3", - "3f8a8ccb663b855fc664a780ac4d7dd966d8b4e550c6ac581799d3b2e1a16225", - "d138ee34e1926441f905daa58aece2045f4cabffd35f645edfa6101f191a1d1e", - "2309609c1a9a969482166faab37e4ce4afe6a0a143091a8611924facf210a317", - "197ee1e43fcad2e6488be436caf13227eb0169458d5d34d15cdfd3cda49977bd", - "b076d383a3a1da8779de4933bcb026b63a747bc4b7331f1822139dab3f9df945", - "dfb39765b1695df590f25059cf1d908f4bf0de89ce4b384a7b50f8bf1cfaa300", - "de0ef822f5b94a70a16d73437f4e9643c92697e102b49f0392329be74bb0ced1", - "30b4fd05b9d54bd5355c82fdc649676aeba7f928e7a03d3c65edc8ec603fe0c9", - "fc81ff4981c88cfa2aa4a525ab634d0496414257e39d4df28f70ab6eaf2c7741", - "f31479f6de4dd3051b3b9f5974461c5f7d22d068e54a3e7306d6872305abd075", - "9df4285dccbb039393ddb8fd845b364a0f0bf6f9f3143909f6855e97a6a13177", - "252b820322257f42dd23d2166fc0cf7dc9de1fcdec6b924b7fe97bc114ac9cae", - "f3e4d5cf093a553b17ee430c4f0ff5e823722890ab1e5a7db6d43e7f0cc7a383", - "ef97c8b3aec6e35ee9e712d75e0021e51c2397fadfebe707f4ca1145e5bfd1b9", - "9433645f919258600f84f5061d6d76d5e4271df70c9bc84c934dcf5b0d448556", - "cdbb9025a55f231df41f57f389c77e18f5e05db4cc71569e91969d28babefcde", - "f71590008c06882c5805c225a90908f558de060013c275e54a6a675d98594f16", - "70fa9fdb0d6b1b49bf64aa7379036715d387ef74af51a7f5b7a65b5ad2c11f85", - "889bbfa5236aafd0c1378e50cb253aaac726193c319256c3a4ec1036ea2984b9", - "e4d070c7b3f28eff89726a2df3228415646f7b7b01cd90e5daf91dbb22337f9f", - "d349f7da795ac6d6452631ef2c259cf9c7478f58a2e9e4eec9734a3b33937873", - "7a5c02c99a413b3e9e26be1f2a766267e41a3055c433e50ef1e7f820a443e843", - "d8f476f32d8fae21f7a2e35db42e12dd00f22c0aeee4401e02f4f956a66dba86", - "895e2c3a76dd56815b717974a4b1ca0cf4b64731a78ffa12ce35d624c7449b0a", - "735c53ccc4f378c61f9cc510467c5c8ebdd835cc1b38f2e58debf6cd3dfa36a1", - "9eb8b648d6ed2a8fab884d490203bd098a71a6efb6226d15c20518168deb8b0f", - "305edef1b0cf670e81ab1e1479cb9c4b05cc82561b15b7b6b84e25140c88f808", - "78be0f78afb3c8516567f574bd7f9affeda5313c077d95373bbf7bbf125b8e5a", - "cc2ab067791f573a9966e5acc93a66c8316ba7196a51950cdbdcd6b6032d8d6c", - "23b09ed0fa27598b3823947543ecc92b6004cb7308b3a54cc741454758e70718", - "e73ef11e2f9ea1d2d7a11b1ef2eb8e823c1fff592afa7f0d48f63b8bd2d01819", - "0333efd41834a9542ac1a5a967c323ada1910ede03911e0e6a12dc936b8bfb48", - "5f92acab79d0676c0d742e5919000a81c5da9dfc017515729ff9c1bde351bb8c", - "0d3fe65d690694f2ee40b1a11062aa2cc979a3a17ea3e200626e72444429ab0c", - "905da64c60aaa2a4ee99f785144c06728a96406248714242df945993c39f13bf", - "8ea07ceae4da2020dd0a5ef4f85b37217a5f7d130ce1d9bf82a27149c4ec47bb", - "1448f45f133897a6a980b4857a8c6fd7f86d01520a80f0079bc0bf2278384c1c", - "85f1f718d3b19a882d5d06899ff62d42918575f6490478af699b50acfbce12ed" + "a25c50a25e1e3858ed6a0de08044bcd93e2d81b11d1d6d44c2a0136be329d3a0", + "b692259c89ae47186d0b0f7dd1763fa3c7c7d175c7c476612ff439454d20be13", + "e2441770a4b7d99ab129201d3e63e224b6a568ccea3a6cc1f73da72e9fa34284", + "37ef9d215d143e10d881533a7965b9dd3f5a97fb876c099a89d924f569466561", + "2ceddc87bd53cc2b63b0a40e74066d1f17cb0f6780ad1622a6ae9d253ea8b2dc", + "ce4e0dfca61346acd9773fbf9514248077b0c362fcd4fbd385a89ee02119524f", + "b1a84abf768d28aaea594d9a74e42916c12932d71bf705788bf15ca47f0fd93b", + "81530a5cf6aa551694b07a2fdc31c8cdfedceebb8c12479ed8f9321fa2e4e1d2", + "87033849db9e554e3805bdc9ebdc1143424765e793b99b1c85e8500084a10506", + "f8678deecbea70be3e4abde3e6639f3facd3d37cf100e9732539c0ef386cc420", + "0084b0db24c0492ac9f312274c64c91119859c5799d93a73028f841d991074f8", + "063879da97c1c0099d909d8d662a3aa507b3764bd5974cdefe3bb6ba7a789157", + "ecc1d0dfc017ea648aca65698b63fc39583acb7856cf8719919f09eb043e2826", + "444529e2602f6137d02c42630f287f6b9f9ad0c1e7041955da13ab995d1f068f", + "03f933e9581d15ac3449fe7574bf79598737d7f7863e8570213027ad5ae5a92a", + "8e776de0bdd13afbc3f39dc779c82799bdf23b6e2ec6729eec6bd90284a2ad74", + "7a7d213c82041c35d0ac663d83bc40aaae442a5974fadab8501b67ab54a48e3b", + "2df5ab2ea21ee547586fe362e82c6d38c9a85300ee4a6b08ff09f2980faa5458", + "bc9894d9c54b831116d3473e264d4c4634df5b18fc0f2f3d5f4917d60197ea56", + "ef1b12aa1f14b9a34bfd83ecca713cf4918e8c43fe97dbd5bbfc329be5d60331", + "e1ac7b72e3615dac44337553abf6121499850ed4fca277dccae709690158ee25", + "c47c4a7e7aa104bf982a3a7cfd5a7db2d71b1d7a87de3277947406d8028c4aa2", + "047c3e6e224e592cebf9a54407b6d3c84b9ba7fb0b0c801f3742d65bfbc6dc83", + "3b41fa645c2071dc1d7c917ff28b4a7561a2e6a34e521a96f8131a285794fef2", + "205c88fed281636189be560e8349ee1987bea84b2a9e529269db0cba7f3c125a", + "3b7c03ed42e471e9d3fa96294612742cd63cde819c561d0d58b8e6de1febf65e", + "eb3de0f368ebbb76e3b9ae99b831e4afe947a21f1a3fcad21fe49e7c58c35532", + "7eaf04ee49d9ba5d69354cf11bc0c1e99b7c598042d7a1c84bc746e3e369e5a1", + "9ef2908fac0f85707d3e886228aa4b791824ce5eabd6135d30ad63dc40efe78f", + "56c5fed416363040190963b0ed68ccb0134caf7895ac3cbe9bd40bbab6a38968", + "8bdba2d3c56c07b0a3314e7520ec32ceae575537702a347b170c868caf63de80", + "d4b36bec03f928025119d1907d7669377f31d0276337106fabf8bbd6dc49aa2c", + "8ee7bbe0f906a6f03982a8c397dd90d3c3da8613ccc8c6fda063b330bf0b8301", + "9f239d814086461dae0773009983b82981e6f8a8365ba8df1e81332c8e3666c6", + "bc6aef3126d9367ef75c50335aeb0be5d094c233424021090dd6f49129f374c2", + "87c9a6edd8fa3576423a761c768ddce93155d208f2fb75ba2ae1b26daf3211f1", + "8c7fb38baf012e8cda2780979f786a0694f8efb3c8a9d4bf8a3b298eab4b5d48", + "ca26d4afe3c1e2dd391ec81c1b59bb8a4a8107ee7713bf81b566e5f226ee9141", + "5943d4100b5ca8e9fa54fcdd13fc3b2dfea27f65d096c3faf1145de8be30c0ba", + "f74976bff78e8570e59eb65c777dc1d2d3584486cbe0f01cf9a4675599fd3174", + "847732560938c1c93c84f6324652c3c6e6ff908b4fc04c55ce4154864b93871c", + "6d0b584a452665db3ac726311ee0efdb31a330f994479475b877332703f09a23", + "1748ee775f38e1cf43a225f1b252d0760ec240323b7ec56b7c8811a97b01d2a9", + "d16b1ae35311e0da9f79c2013d105f750062e58f5566ab7b5ba15084e1d56109", + "a25a94e5910d167017f77af107c448403a8c9c95a0a2f691b26a2db4540bcbed", + "0c50c8d5c7e00f70a69250d1ff2a5a56dfe8a9f0dff6f698cb5b71dad2750157", + "c831ce2f4528d9288471666bb99fa8e739801c1a93baf9977c3af9909751d583", + "2f086f6f8835c7931925469ba7d849d9316ef5ee6ad51fbe7d2f56755b73038e", + "d6a0303eb1a61c5a1d01e1b4c5c84ed9e823104c71cc14ce50c0a9f8387c10f6", + "85977cce7bd649295d0c40de2cdcbbb6aedb0f5968e0970a2fff9a4eb44504b5", + "21a8a8f2252a392869f8c999595b2fe4400af3aac4dcc7e7c93b12d6279ca630", + "4a4680f5b2a670250c8e3a17ca95aeac421842653676bb9ff35cba4a239fc6dd", + "c7bb0f8cd7d497c52eca18d070e0887a3df9495848d9749d05f35b6b748fc872", + "7f646b5b4ebaf6628925acfb67bc0e33b805643cc5e6b2785f5b7299eb1b65fd", + "7c940fa03f179c47b89fae6580b6e50b254bca7536116b979c059747481fc192", + "aef2ece18ecc3c74685e435ac9c3a3817ecac977278864d8d39b8146f4f91055" ], - "last_check_ts": 1783601745.1315062, - "last_data_change_ts": 1783601745.1315098, + "last_check_ts": 1783734277.247423, + "last_data_change_ts": 1783734277.2474256, "last_failed": false, - "last_success_ts": 1783601745.1315057, - "size_mb": 22.99532985687256, - "time_s": 10.31207823753357 + "last_success_ts": 1783734277.2474225, + "size_mb": 19.565000534057617, + "time_s": 13.773402690887451 } } \ No newline at end of file diff --git a/data/llm_stats/amazon/nova-2-omni/17d5844c-8089-43df-ac17-fc7d32d859df.json b/data/llm_stats/amazon/nova-2-omni/17d5844c-8089-43df-ac17-fc7d32d859df.json new file mode 100644 index 0000000000000000000000000000000000000000..29751ee7ad67e49136f4c3dfb2991729092f49ad --- /dev/null +++ b/data/llm_stats/amazon/nova-2-omni/17d5844c-8089-43df-ac17-fc7d32d859df.json @@ -0,0 +1,774 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/amazon_nova-2-omni/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Nova 2 Omni", + "id": "amazon/nova-2-omni", + "developer": "amazon", + "additional_details": { + "raw_id": "nova-2-omni", + "raw_name": "Nova 2 Omni", + "raw_model_id": "nova-2-omni", + "raw_model_name": "Nova 2 Omni", + "raw_organization_id": "amazon", + "raw_organization_name": "Amazon", + "raw_release_date": "2025-12-02", + "raw_announcement_date": "2025-12-02", + "raw_multimodal": "true", + "raw_provider_slug": "amazon", + "raw_provider_name": "Amazon" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "bfcl-v4::bfcl-v4-nova-2-omni", + "evaluation_name": "llm_stats.bfcl-v4", + "source_data": { + "dataset_name": "BFCL-V4", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-2-omni", + "https://llm-stats.com/benchmarks/bfcl-v4", + "https://api.llm-stats.com/leaderboard/benchmarks/bfcl-v4" + ], + "additional_details": { + "raw_benchmark_id": "bfcl-v4", + "raw_model_id": "nova-2-omni", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Berkeley Function Calling Leaderboard V4 (BFCL-V4) evaluates LLMs on their ability to accurately call functions and APIs, including simple, multiple, parallel, and nested function calls across diverse programming scenarios.", + "metric_id": "llm_stats.bfcl-v4.score", + "metric_name": "BFCL-V4 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bfcl-v4", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BFCL-V4", + "raw_categories": "[\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.583, + "details": { + "raw_score": "0.583", + "raw_score_field": "score", + "raw_model_id": "nova-2-omni", + "raw_benchmark_id": "bfcl-v4", + "source_urls_json": "[\"https://llm-stats.com/models/nova-2-omni\",\"https://llm-stats.com/benchmarks/bfcl-v4\",\"https://api.llm-stats.com/leaderboard/benchmarks/bfcl-v4\"]", + "raw_score_id": "bfcl-v4::nova-2-omni", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "covost2::covost2-nova-2-omni", + "evaluation_name": "llm_stats.covost2", + "source_data": { + "dataset_name": "CoVoST2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-2-omni", + "https://llm-stats.com/benchmarks/covost2", + "https://api.llm-stats.com/leaderboard/benchmarks/covost2" + ], + "additional_details": { + "raw_benchmark_id": "covost2", + "raw_model_id": "nova-2-omni", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CoVoST 2 is a large-scale multilingual speech translation corpus derived from Common Voice, covering translations from 21 languages into English and from English into 15 languages. The dataset contains 2,880 hours of speech with 78K speakers for speech translation research.", + "metric_id": "llm_stats.covost2.score", + "metric_name": "CoVoST2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "covost2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CoVoST2", + "raw_categories": "[\"language\",\"speech_to_text\",\"audio\"]", + "raw_modality": "audio", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.407, + "details": { + "raw_score": "0.407", + "raw_score_field": "score", + "raw_model_id": "nova-2-omni", + "raw_benchmark_id": "covost2", + "source_urls_json": "[\"https://llm-stats.com/models/nova-2-omni\",\"https://llm-stats.com/benchmarks/covost2\",\"https://api.llm-stats.com/leaderboard/benchmarks/covost2\"]", + "raw_score_id": "covost2::nova-2-omni", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ifbench::ifbench-nova-2-omni", + "evaluation_name": "llm_stats.ifbench", + "source_data": { + "dataset_name": "IFBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-2-omni", + "https://llm-stats.com/benchmarks/ifbench", + "https://api.llm-stats.com/leaderboard/benchmarks/ifbench" + ], + "additional_details": { + "raw_benchmark_id": "ifbench", + "raw_model_id": "nova-2-omni", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Instruction Following Benchmark evaluating model's ability to follow complex instructions", + "metric_id": "llm_stats.ifbench.score", + "metric_name": "IFBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ifbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IFBench", + "raw_categories": "[\"instruction_following\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.687, + "details": { + "raw_score": "0.687", + "raw_score_field": "score", + "raw_model_id": "nova-2-omni", + "raw_benchmark_id": "ifbench", + "source_urls_json": "[\"https://llm-stats.com/models/nova-2-omni\",\"https://llm-stats.com/benchmarks/ifbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/ifbench\"]", + "raw_score_id": "ifbench::nova-2-omni", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "maverix::maverix-nova-2-omni", + "evaluation_name": "llm_stats.maverix", + "source_data": { + "dataset_name": "MAVERIX", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-2-omni", + "https://llm-stats.com/benchmarks/maverix", + "https://api.llm-stats.com/leaderboard/benchmarks/maverix" + ], + "additional_details": { + "raw_benchmark_id": "maverix", + "raw_model_id": "nova-2-omni", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MAVERIX (Multimodal Audio-Visual Evaluation Reasoning Index) evaluates multimodal models on tasks that demand tight integration of video and audio information. It features challenges like situational awareness and social sentiment analysis where the answer cannot be reliably determined from a single modality, rigorously testing joint audio-visual understanding.", + "metric_id": "llm_stats.maverix.score", + "metric_name": "MAVERIX score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "maverix", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MAVERIX", + "raw_categories": "[\"multimodal\",\"reasoning\",\"audio\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.666, + "details": { + "raw_score": "0.666", + "raw_score_field": "score", + "raw_model_id": "nova-2-omni", + "raw_benchmark_id": "maverix", + "source_urls_json": "[\"https://llm-stats.com/models/nova-2-omni\",\"https://llm-stats.com/benchmarks/maverix\",\"https://api.llm-stats.com/leaderboard/benchmarks/maverix\"]", + "raw_score_id": "maverix::nova-2-omni", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmau::mmau-nova-2-omni", + "evaluation_name": "llm_stats.mmau", + "source_data": { + "dataset_name": "MMAU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-2-omni", + "https://llm-stats.com/benchmarks/mmau", + "https://api.llm-stats.com/leaderboard/benchmarks/mmau" + ], + "additional_details": { + "raw_benchmark_id": "mmau", + "raw_model_id": "nova-2-omni", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A massive multi-task audio understanding and reasoning benchmark comprising 10,000 carefully curated audio clips paired with human-annotated natural language questions spanning speech, environmental sounds, and music. Requires expert-level knowledge and complex reasoning across 27 distinct skills.", + "metric_id": "llm_stats.mmau.score", + "metric_name": "MMAU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmau", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMAU", + "raw_categories": "[\"multimodal\",\"reasoning\",\"audio\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.753, + "details": { + "raw_score": "0.753", + "raw_score_field": "score", + "raw_model_id": "nova-2-omni", + "raw_benchmark_id": "mmau", + "source_urls_json": "[\"https://llm-stats.com/models/nova-2-omni\",\"https://llm-stats.com/benchmarks/mmau\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmau\"]", + "raw_score_id": "mmau::nova-2-omni", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "multichallenge::multichallenge-nova-2-omni", + "evaluation_name": "llm_stats.multichallenge", + "source_data": { + "dataset_name": "Multi-Challenge", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-2-omni", + "https://llm-stats.com/benchmarks/multichallenge", + "https://api.llm-stats.com/leaderboard/benchmarks/multichallenge" + ], + "additional_details": { + "raw_benchmark_id": "multichallenge", + "raw_model_id": "nova-2-omni", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MultiChallenge is a realistic multi-turn conversation evaluation benchmark that challenges frontier LLMs across four key categories: instruction retention (maintaining instructions throughout conversations), inference memory (recalling and connecting details from previous turns), reliable versioned editing (adapting to evolving instructions during collaborative editing), and self-coherence (avoiding contradictions in responses). The benchmark evaluates models on sustained, contextually complex dialogues across diverse topics including travel planning, technical documentation, and professional communication.", + "metric_id": "llm_stats.multichallenge.score", + "metric_name": "Multi-Challenge score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "multichallenge", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Multi-Challenge", + "raw_categories": "[\"reasoning\",\"communication\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "29" + } + }, + "score_details": { + "score": 0.755, + "details": { + "raw_score": "0.755", + "raw_score_field": "score", + "raw_model_id": "nova-2-omni", + "raw_benchmark_id": "multichallenge", + "source_urls_json": "[\"https://llm-stats.com/models/nova-2-omni\",\"https://llm-stats.com/benchmarks/multichallenge\",\"https://api.llm-stats.com/leaderboard/benchmarks/multichallenge\"]", + "raw_score_id": "multichallenge::nova-2-omni", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ocrbench-v2::ocrbench-v2-nova-2-omni", + "evaluation_name": "llm_stats.ocrbench-v2", + "source_data": { + "dataset_name": "OCRBench_V2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-2-omni", + "https://llm-stats.com/benchmarks/ocrbench-v2", + "https://api.llm-stats.com/leaderboard/benchmarks/ocrbench-v2" + ], + "additional_details": { + "raw_benchmark_id": "ocrbench-v2", + "raw_model_id": "nova-2-omni", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OCRBench v2: Enhanced large-scale bilingual benchmark for evaluating Large Multimodal Models on visual text localization and reasoning with 10,000 human-verified question-answering pairs across 8 core OCR capabilities", + "metric_id": "llm_stats.ocrbench-v2.score", + "metric_name": "OCRBench_V2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ocrbench-v2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OCRBench_V2", + "raw_categories": "[\"image_to_text\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.582, + "details": { + "raw_score": "0.582", + "raw_score_field": "score", + "raw_model_id": "nova-2-omni", + "raw_benchmark_id": "ocrbench-v2", + "source_urls_json": "[\"https://llm-stats.com/models/nova-2-omni\",\"https://llm-stats.com/benchmarks/ocrbench-v2\",\"https://api.llm-stats.com/leaderboard/benchmarks/ocrbench-v2\"]", + "raw_score_id": "ocrbench-v2::nova-2-omni", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "qvhighlights::qvhighlights-nova-2-omni", + "evaluation_name": "llm_stats.qvhighlights", + "source_data": { + "dataset_name": "QVHighlights", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-2-omni", + "https://llm-stats.com/benchmarks/qvhighlights", + "https://api.llm-stats.com/leaderboard/benchmarks/qvhighlights" + ], + "additional_details": { + "raw_benchmark_id": "qvhighlights", + "raw_model_id": "nova-2-omni", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "QVHighlights is a video moment retrieval benchmark for detecting moments and highlights in videos via natural language queries. Given a query, the model must localize the start and end times of relevant moments in the video, evaluated using metrics such as Recall@1 at a 0.5 IoU threshold.", + "metric_id": "llm_stats.qvhighlights.score", + "metric_name": "QVHighlights score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "qvhighlights", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "QVHighlights", + "raw_categories": "[\"multimodal\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.767, + "details": { + "raw_score": "0.767", + "raw_score_field": "score", + "raw_model_id": "nova-2-omni", + "raw_benchmark_id": "qvhighlights", + "source_urls_json": "[\"https://llm-stats.com/models/nova-2-omni\",\"https://llm-stats.com/benchmarks/qvhighlights\",\"https://api.llm-stats.com/leaderboard/benchmarks/qvhighlights\"]", + "raw_score_id": "qvhighlights::nova-2-omni", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "realkie-fcc::realkie-fcc-nova-2-omni", + "evaluation_name": "llm_stats.realkie-fcc", + "source_data": { + "dataset_name": "RealKIE-FCC", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-2-omni", + "https://llm-stats.com/benchmarks/realkie-fcc", + "https://api.llm-stats.com/leaderboard/benchmarks/realkie-fcc" + ], + "additional_details": { + "raw_benchmark_id": "realkie-fcc", + "raw_model_id": "nova-2-omni", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "RealKIE-FCC is a key information extraction benchmark drawn from real enterprise documents (FCC filings), part of the RealKIE suite of five novel datasets for enterprise key information extraction. Models must convert documents to markdown and extract structured fields against a specified JSON schema. Nova 2 reports results on a human-verified version of the dataset.", + "metric_id": "llm_stats.realkie-fcc.score", + "metric_name": "RealKIE-FCC score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "realkie-fcc", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "RealKIE-FCC", + "raw_categories": "[\"multimodal\",\"document_understanding\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.598, + "details": { + "raw_score": "0.598", + "raw_score_field": "score", + "raw_model_id": "nova-2-omni", + "raw_benchmark_id": "realkie-fcc", + "source_urls_json": "[\"https://llm-stats.com/models/nova-2-omni\",\"https://llm-stats.com/benchmarks/realkie-fcc\",\"https://api.llm-stats.com/leaderboard/benchmarks/realkie-fcc\"]", + "raw_score_id": "realkie-fcc::nova-2-omni", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "refcocog::refcocog-nova-2-omni", + "evaluation_name": "llm_stats.refcocog", + "source_data": { + "dataset_name": "RefCOCOg", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-2-omni", + "https://llm-stats.com/benchmarks/refcocog", + "https://api.llm-stats.com/leaderboard/benchmarks/refcocog" + ], + "additional_details": { + "raw_benchmark_id": "refcocog", + "raw_model_id": "nova-2-omni", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "RefCOCOg is a referring expression comprehension benchmark that evaluates spatial grounding in images. Given a natural language expression describing an object, the model must localize the correct region, evaluated by accuracy at a 0.5 IoU threshold. It features longer, more descriptive expressions than RefCOCO and RefCOCO+.", + "metric_id": "llm_stats.refcocog.score", + "metric_name": "RefCOCOg score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "refcocog", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "RefCOCOg", + "raw_categories": "[\"multimodal\",\"grounding\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.863, + "details": { + "raw_score": "0.863", + "raw_score_field": "score", + "raw_model_id": "nova-2-omni", + "raw_benchmark_id": "refcocog", + "source_urls_json": "[\"https://llm-stats.com/models/nova-2-omni\",\"https://llm-stats.com/benchmarks/refcocog\",\"https://api.llm-stats.com/leaderboard/benchmarks/refcocog\"]", + "raw_score_id": "refcocog::nova-2-omni", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "screenspot::screenspot-nova-2-omni", + "evaluation_name": "llm_stats.screenspot", + "source_data": { + "dataset_name": "ScreenSpot", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-2-omni", + "https://llm-stats.com/benchmarks/screenspot", + "https://api.llm-stats.com/leaderboard/benchmarks/screenspot" + ], + "additional_details": { + "raw_benchmark_id": "screenspot", + "raw_model_id": "nova-2-omni", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ScreenSpot is the first realistic GUI grounding benchmark that encompasses mobile, desktop, and web environments. The dataset comprises over 1,200 instructions from iOS, Android, macOS, Windows and Web environments, along with annotated element types (text and icon/widget), designed to evaluate visual GUI agents' ability to accurately locate screen elements based on natural language instructions.", + "metric_id": "llm_stats.screenspot.score", + "metric_name": "ScreenSpot score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "screenspot", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ScreenSpot", + "raw_categories": "[\"multimodal\",\"spatial_reasoning\",\"grounding\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.854, + "details": { + "raw_score": "0.854", + "raw_score_field": "score", + "raw_model_id": "nova-2-omni", + "raw_benchmark_id": "screenspot", + "source_urls_json": "[\"https://llm-stats.com/models/nova-2-omni\",\"https://llm-stats.com/benchmarks/screenspot\",\"https://api.llm-stats.com/leaderboard/benchmarks/screenspot\"]", + "raw_score_id": "screenspot::nova-2-omni", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "tau2-airline::tau2-airline-nova-2-omni", + "evaluation_name": "llm_stats.tau2-airline", + "source_data": { + "dataset_name": "Tau2 Airline", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-2-omni", + "https://llm-stats.com/benchmarks/tau2-airline", + "https://api.llm-stats.com/leaderboard/benchmarks/tau2-airline" + ], + "additional_details": { + "raw_benchmark_id": "tau2-airline", + "raw_model_id": "nova-2-omni", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "TAU2 airline domain benchmark for evaluating conversational agents in dual-control environments where both AI agents and users interact with tools in airline customer service scenarios. Tests agent coordination, communication, and ability to guide user actions in tasks like flight booking, modifications, cancellations, and refunds.", + "metric_id": "llm_stats.tau2-airline.score", + "metric_name": "Tau2 Airline score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "tau2-airline", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Tau2 Airline", + "raw_categories": "[\"reasoning\",\"communication\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.688, + "details": { + "raw_score": "0.688", + "raw_score_field": "score", + "raw_model_id": "nova-2-omni", + "raw_benchmark_id": "tau2-airline", + "source_urls_json": "[\"https://llm-stats.com/models/nova-2-omni\",\"https://llm-stats.com/benchmarks/tau2-airline\",\"https://api.llm-stats.com/leaderboard/benchmarks/tau2-airline\"]", + "raw_score_id": "tau2-airline::nova-2-omni", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "tau2-retail::tau2-retail-nova-2-omni", + "evaluation_name": "llm_stats.tau2-retail", + "source_data": { + "dataset_name": "Tau2 Retail", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-2-omni", + "https://llm-stats.com/benchmarks/tau2-retail", + "https://api.llm-stats.com/leaderboard/benchmarks/tau2-retail" + ], + "additional_details": { + "raw_benchmark_id": "tau2-retail", + "raw_model_id": "nova-2-omni", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "τ²-bench retail domain evaluates conversational AI agents in customer service scenarios within a dual-control environment where both agent and user can interact with tools. Tests tool-agent-user interaction, rule adherence, and task consistency in retail customer support contexts.", + "metric_id": "llm_stats.tau2-retail.score", + "metric_name": "Tau2 Retail score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "tau2-retail", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Tau2 Retail", + "raw_categories": "[\"reasoning\",\"communication\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "26" + } + }, + "score_details": { + "score": 0.783, + "details": { + "raw_score": "0.783", + "raw_score_field": "score", + "raw_model_id": "nova-2-omni", + "raw_benchmark_id": "tau2-retail", + "source_urls_json": "[\"https://llm-stats.com/models/nova-2-omni\",\"https://llm-stats.com/benchmarks/tau2-retail\",\"https://api.llm-stats.com/leaderboard/benchmarks/tau2-retail\"]", + "raw_score_id": "tau2-retail::nova-2-omni", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "video-mme::video-mme-nova-2-omni", + "evaluation_name": "llm_stats.video-mme", + "source_data": { + "dataset_name": "Video-MME", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-2-omni", + "https://llm-stats.com/benchmarks/video-mme", + "https://api.llm-stats.com/leaderboard/benchmarks/video-mme" + ], + "additional_details": { + "raw_benchmark_id": "video-mme", + "raw_model_id": "nova-2-omni", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Video-MME is the first-ever comprehensive evaluation benchmark of Multi-modal Large Language Models (MLLMs) in video analysis. It features 900 videos totaling 254 hours with 2,700 human-annotated question-answer pairs across 6 primary visual domains (Knowledge, Film & Television, Sports Competition, Life Record, Multilingual, and others) and 30 subfields. The benchmark evaluates models across diverse temporal dimensions (11 seconds to 1 hour), integrates multi-modal inputs including video frames, subtitles, and audio, and uses rigorous manual labeling by expert annotators for precise assessment.", + "metric_id": "llm_stats.video-mme.score", + "metric_name": "Video-MME score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "video-mme", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Video-MME", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "17" + } + }, + "score_details": { + "score": 0.779, + "details": { + "raw_score": "0.779", + "raw_score_field": "score", + "raw_model_id": "nova-2-omni", + "raw_benchmark_id": "video-mme", + "source_urls_json": "[\"https://llm-stats.com/models/nova-2-omni\",\"https://llm-stats.com/benchmarks/video-mme\",\"https://api.llm-stats.com/leaderboard/benchmarks/video-mme\"]", + "raw_score_id": "video-mme::nova-2-omni", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/amazon/nova-lite/87698ae1-019d-4ab8-80bd-12f77ada1225.json b/data/llm_stats/amazon/nova-lite/87698ae1-019d-4ab8-80bd-12f77ada1225.json new file mode 100644 index 0000000000000000000000000000000000000000..0d2d2340d572f74f411943e38db8f8f513152a79 --- /dev/null +++ b/data/llm_stats/amazon/nova-lite/87698ae1-019d-4ab8-80bd-12f77ada1225.json @@ -0,0 +1,1138 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/amazon_nova-lite/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Nova Lite", + "id": "amazon/nova-lite", + "developer": "amazon", + "additional_details": { + "raw_id": "nova-lite", + "raw_name": "Nova Lite", + "raw_model_id": "nova-lite", + "raw_model_name": "Nova Lite", + "raw_organization_id": "amazon", + "raw_organization_name": "Amazon", + "raw_release_date": "2024-11-20", + "raw_announcement_date": "2024-11-20", + "raw_multimodal": "true", + "raw_provider_slug": "amazon", + "raw_provider_name": "Amazon" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "arc-c::arc-c-nova-lite", + "evaluation_name": "llm_stats.arc-c", + "source_data": { + "dataset_name": "ARC-C", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-lite", + "https://llm-stats.com/benchmarks/arc-c", + "https://api.llm-stats.com/leaderboard/benchmarks/arc-c" + ], + "additional_details": { + "raw_benchmark_id": "arc-c", + "raw_model_id": "nova-lite", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The AI2 Reasoning Challenge (ARC) Challenge Set is a multiple-choice question-answering benchmark containing grade-school level science questions that require advanced reasoning capabilities. ARC-C specifically contains questions that were answered incorrectly by both retrieval-based and word co-occurrence algorithms, making it a particularly challenging subset designed to test commonsense reasoning abilities in AI systems.", + "metric_id": "llm_stats.arc-c.score", + "metric_name": "ARC-C score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arc-c", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ARC-C", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "34" + } + }, + "score_details": { + "score": 0.924, + "details": { + "raw_score": "0.924", + "raw_score_field": "score", + "raw_model_id": "nova-lite", + "raw_benchmark_id": "arc-c", + "source_urls_json": "[\"https://llm-stats.com/models/nova-lite\",\"https://llm-stats.com/benchmarks/arc-c\",\"https://api.llm-stats.com/leaderboard/benchmarks/arc-c\"]", + "raw_score_id": "arc-c::nova-lite", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bbh::bbh-nova-lite", + "evaluation_name": "llm_stats.bbh", + "source_data": { + "dataset_name": "BBH", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-lite", + "https://llm-stats.com/benchmarks/bbh", + "https://api.llm-stats.com/leaderboard/benchmarks/bbh" + ], + "additional_details": { + "raw_benchmark_id": "bbh", + "raw_model_id": "nova-lite", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Big-Bench Hard (BBH) is a suite of 23 challenging tasks selected from BIG-Bench for which prior language model evaluations did not outperform the average human-rater. These tasks require multi-step reasoning across diverse domains including arithmetic, logical reasoning, reading comprehension, and commonsense reasoning. The benchmark was designed to test capabilities believed to be beyond current language models and focuses on evaluating complex reasoning skills including temporal understanding, spatial reasoning, causal understanding, and deductive logical reasoning.", + "metric_id": "llm_stats.bbh.score", + "metric_name": "BBH score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bbh", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BBH", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.824, + "details": { + "raw_score": "0.824", + "raw_score_field": "score", + "raw_model_id": "nova-lite", + "raw_benchmark_id": "bbh", + "source_urls_json": "[\"https://llm-stats.com/models/nova-lite\",\"https://llm-stats.com/benchmarks/bbh\",\"https://api.llm-stats.com/leaderboard/benchmarks/bbh\"]", + "raw_score_id": "bbh::nova-lite", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bfcl::bfcl-nova-lite", + "evaluation_name": "llm_stats.bfcl", + "source_data": { + "dataset_name": "BFCL", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-lite", + "https://llm-stats.com/benchmarks/bfcl", + "https://api.llm-stats.com/leaderboard/benchmarks/bfcl" + ], + "additional_details": { + "raw_benchmark_id": "bfcl", + "raw_model_id": "nova-lite", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The Berkeley Function Calling Leaderboard (BFCL) is the first comprehensive and executable function call evaluation dedicated to assessing Large Language Models' ability to invoke functions. It evaluates serial and parallel function calls across multiple programming languages (Python, Java, JavaScript, REST API) using a novel Abstract Syntax Tree (AST) evaluation method. The benchmark consists of over 2,000 question-function-answer pairs covering diverse application domains and complex use cases including multiple function calls, parallel function calls, and multi-turn interactions.", + "metric_id": "llm_stats.bfcl.score", + "metric_name": "BFCL score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bfcl", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BFCL", + "raw_categories": "[\"reasoning\",\"general\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.666, + "details": { + "raw_score": "0.666", + "raw_score_field": "score", + "raw_model_id": "nova-lite", + "raw_benchmark_id": "bfcl", + "source_urls_json": "[\"https://llm-stats.com/models/nova-lite\",\"https://llm-stats.com/benchmarks/bfcl\",\"https://api.llm-stats.com/leaderboard/benchmarks/bfcl\"]", + "raw_score_id": "bfcl::nova-lite", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "chartqa::chartqa-nova-lite", + "evaluation_name": "llm_stats.chartqa", + "source_data": { + "dataset_name": "ChartQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-lite", + "https://llm-stats.com/benchmarks/chartqa", + "https://api.llm-stats.com/leaderboard/benchmarks/chartqa" + ], + "additional_details": { + "raw_benchmark_id": "chartqa", + "raw_model_id": "nova-lite", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ChartQA is a large-scale benchmark comprising 9.6K human-written questions and 23.1K questions generated from human-written chart summaries, designed to evaluate models' abilities in visual and logical reasoning over charts.", + "metric_id": "llm_stats.chartqa.score", + "metric_name": "ChartQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "chartqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ChartQA", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "24" + } + }, + "score_details": { + "score": 0.868, + "details": { + "raw_score": "0.868", + "raw_score_field": "score", + "raw_model_id": "nova-lite", + "raw_benchmark_id": "chartqa", + "source_urls_json": "[\"https://llm-stats.com/models/nova-lite\",\"https://llm-stats.com/benchmarks/chartqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/chartqa\"]", + "raw_score_id": "chartqa::nova-lite", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "crag::crag-nova-lite", + "evaluation_name": "llm_stats.crag", + "source_data": { + "dataset_name": "CRAG", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-lite", + "https://llm-stats.com/benchmarks/crag", + "https://api.llm-stats.com/leaderboard/benchmarks/crag" + ], + "additional_details": { + "raw_benchmark_id": "crag", + "raw_model_id": "nova-lite", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CRAG (Comprehensive RAG Benchmark) is a factual question answering benchmark consisting of 4,409 question-answer pairs across 5 domains (finance, sports, music, movie, open domain) and 8 question categories. The benchmark includes mock APIs to simulate web and Knowledge Graph search, designed to represent the diverse and dynamic nature of real-world QA tasks with temporal dynamism ranging from years to seconds. It evaluates retrieval-augmented generation systems for trustworthy question answering.", + "metric_id": "llm_stats.crag.score", + "metric_name": "CRAG score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "crag", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CRAG", + "raw_categories": "[\"reasoning\",\"search\",\"finance\",\"economics\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.438, + "details": { + "raw_score": "0.438", + "raw_score_field": "score", + "raw_model_id": "nova-lite", + "raw_benchmark_id": "crag", + "source_urls_json": "[\"https://llm-stats.com/models/nova-lite\",\"https://llm-stats.com/benchmarks/crag\",\"https://api.llm-stats.com/leaderboard/benchmarks/crag\"]", + "raw_score_id": "crag::nova-lite", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "docvqa::docvqa-nova-lite", + "evaluation_name": "llm_stats.docvqa", + "source_data": { + "dataset_name": "DocVQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-lite", + "https://llm-stats.com/benchmarks/docvqa", + "https://api.llm-stats.com/leaderboard/benchmarks/docvqa" + ], + "additional_details": { + "raw_benchmark_id": "docvqa", + "raw_model_id": "nova-lite", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A dataset for Visual Question Answering on document images containing 50,000 questions defined on 12,000+ document images. The benchmark tests AI's ability to understand document structure and content, requiring models to comprehend document layout and perform information retrieval to answer questions about document images.", + "metric_id": "llm_stats.docvqa.score", + "metric_name": "DocVQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "docvqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DocVQA", + "raw_categories": "[\"image_to_text\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "26" + } + }, + "score_details": { + "score": 0.924, + "details": { + "raw_score": "0.924", + "raw_score_field": "score", + "raw_model_id": "nova-lite", + "raw_benchmark_id": "docvqa", + "source_urls_json": "[\"https://llm-stats.com/models/nova-lite\",\"https://llm-stats.com/benchmarks/docvqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/docvqa\"]", + "raw_score_id": "docvqa::nova-lite", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "drop::drop-nova-lite", + "evaluation_name": "llm_stats.drop", + "source_data": { + "dataset_name": "DROP", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-lite", + "https://llm-stats.com/benchmarks/drop", + "https://api.llm-stats.com/leaderboard/benchmarks/drop" + ], + "additional_details": { + "raw_benchmark_id": "drop", + "raw_model_id": "nova-lite", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "DROP (Discrete Reasoning Over Paragraphs) is a reading comprehension benchmark requiring discrete reasoning over paragraph content. It contains crowdsourced, adversarially-created questions that require resolving references and performing discrete operations like addition, counting, or sorting, demanding comprehensive paragraph understanding beyond paraphrase-and-entity-typing shortcuts.", + "metric_id": "llm_stats.drop.score", + "metric_name": "DROP score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "drop", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DROP", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "30" + } + }, + "score_details": { + "score": 0.802, + "details": { + "raw_score": "0.802", + "raw_score_field": "score", + "raw_model_id": "nova-lite", + "raw_benchmark_id": "drop", + "source_urls_json": "[\"https://llm-stats.com/models/nova-lite\",\"https://llm-stats.com/benchmarks/drop\",\"https://api.llm-stats.com/leaderboard/benchmarks/drop\"]", + "raw_score_id": "drop::nova-lite", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "egoschema::egoschema-nova-lite", + "evaluation_name": "llm_stats.egoschema", + "source_data": { + "dataset_name": "EgoSchema", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-lite", + "https://llm-stats.com/benchmarks/egoschema", + "https://api.llm-stats.com/leaderboard/benchmarks/egoschema" + ], + "additional_details": { + "raw_benchmark_id": "egoschema", + "raw_model_id": "nova-lite", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A diagnostic benchmark for very long-form video language understanding consisting of over 5000 human curated multiple choice questions based on 3-minute video clips from Ego4D, covering a broad range of natural human activities and behaviors", + "metric_id": "llm_stats.egoschema.score", + "metric_name": "EgoSchema score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "egoschema", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "EgoSchema", + "raw_categories": "[\"long_context\",\"reasoning\",\"vision\"]", + "raw_modality": "video", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.714, + "details": { + "raw_score": "0.714", + "raw_score_field": "score", + "raw_model_id": "nova-lite", + "raw_benchmark_id": "egoschema", + "source_urls_json": "[\"https://llm-stats.com/models/nova-lite\",\"https://llm-stats.com/benchmarks/egoschema\",\"https://api.llm-stats.com/leaderboard/benchmarks/egoschema\"]", + "raw_score_id": "egoschema::nova-lite", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "finqa::finqa-nova-lite", + "evaluation_name": "llm_stats.finqa", + "source_data": { + "dataset_name": "FinQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-lite", + "https://llm-stats.com/benchmarks/finqa", + "https://api.llm-stats.com/leaderboard/benchmarks/finqa" + ], + "additional_details": { + "raw_benchmark_id": "finqa", + "raw_model_id": "nova-lite", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A large-scale dataset for numerical reasoning over financial data with question-answering pairs written by financial experts, featuring complex numerical reasoning and understanding of heterogeneous representations with annotated gold reasoning programs for full explainability", + "metric_id": "llm_stats.finqa.score", + "metric_name": "FinQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "finqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FinQA", + "raw_categories": "[\"math\",\"reasoning\",\"finance\",\"economics\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.736, + "details": { + "raw_score": "0.736", + "raw_score_field": "score", + "raw_model_id": "nova-lite", + "raw_benchmark_id": "finqa", + "source_urls_json": "[\"https://llm-stats.com/models/nova-lite\",\"https://llm-stats.com/benchmarks/finqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/finqa\"]", + "raw_score_id": "finqa::nova-lite", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "groundui-1k::groundui-1k-nova-lite", + "evaluation_name": "llm_stats.groundui-1k", + "source_data": { + "dataset_name": "GroundUI-1K", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-lite", + "https://llm-stats.com/benchmarks/groundui-1k", + "https://api.llm-stats.com/leaderboard/benchmarks/groundui-1k" + ], + "additional_details": { + "raw_benchmark_id": "groundui-1k", + "raw_model_id": "nova-lite", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A subset of GroundUI-18K for UI grounding evaluation, where models must predict action coordinates on screenshots based on single-step instructions across web, desktop, and mobile platforms.", + "metric_id": "llm_stats.groundui-1k.score", + "metric_name": "GroundUI-1K score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "groundui-1k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GroundUI-1K", + "raw_categories": "[\"multimodal\",\"grounding\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.802, + "details": { + "raw_score": "0.802", + "raw_score_field": "score", + "raw_model_id": "nova-lite", + "raw_benchmark_id": "groundui-1k", + "source_urls_json": "[\"https://llm-stats.com/models/nova-lite\",\"https://llm-stats.com/benchmarks/groundui-1k\",\"https://api.llm-stats.com/leaderboard/benchmarks/groundui-1k\"]", + "raw_score_id": "groundui-1k::nova-lite", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gsm8k::gsm8k-nova-lite", + "evaluation_name": "llm_stats.gsm8k", + "source_data": { + "dataset_name": "GSM8k", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-lite", + "https://llm-stats.com/benchmarks/gsm8k", + "https://api.llm-stats.com/leaderboard/benchmarks/gsm8k" + ], + "additional_details": { + "raw_benchmark_id": "gsm8k", + "raw_model_id": "nova-lite", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Grade School Math 8K, a dataset of 8.5K high-quality linguistically diverse grade school math word problems requiring multi-step reasoning and elementary arithmetic operations.", + "metric_id": "llm_stats.gsm8k.score", + "metric_name": "GSM8k score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gsm8k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GSM8k", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "48" + } + }, + "score_details": { + "score": 0.945, + "details": { + "raw_score": "0.945", + "raw_score_field": "score", + "raw_model_id": "nova-lite", + "raw_benchmark_id": "gsm8k", + "source_urls_json": "[\"https://llm-stats.com/models/nova-lite\",\"https://llm-stats.com/benchmarks/gsm8k\",\"https://api.llm-stats.com/leaderboard/benchmarks/gsm8k\"]", + "raw_score_id": "gsm8k::nova-lite", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ifeval::ifeval-nova-lite", + "evaluation_name": "llm_stats.ifeval", + "source_data": { + "dataset_name": "IFEval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-lite", + "https://llm-stats.com/benchmarks/ifeval", + "https://api.llm-stats.com/leaderboard/benchmarks/ifeval" + ], + "additional_details": { + "raw_benchmark_id": "ifeval", + "raw_model_id": "nova-lite", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Instruction-Following Evaluation (IFEval) benchmark for large language models, focusing on verifiable instructions with 25 types of instructions and around 500 prompts containing one or more verifiable constraints", + "metric_id": "llm_stats.ifeval.score", + "metric_name": "IFEval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ifeval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IFEval", + "raw_categories": "[\"instruction_following\",\"structured_output\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "65" + } + }, + "score_details": { + "score": 0.897, + "details": { + "raw_score": "0.897", + "raw_score_field": "score", + "raw_model_id": "nova-lite", + "raw_benchmark_id": "ifeval", + "source_urls_json": "[\"https://llm-stats.com/models/nova-lite\",\"https://llm-stats.com/benchmarks/ifeval\",\"https://api.llm-stats.com/leaderboard/benchmarks/ifeval\"]", + "raw_score_id": "ifeval::nova-lite", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mm-mind2web::mm-mind2web-nova-lite", + "evaluation_name": "llm_stats.mm-mind2web", + "source_data": { + "dataset_name": "MM-Mind2Web", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-lite", + "https://llm-stats.com/benchmarks/mm-mind2web", + "https://api.llm-stats.com/leaderboard/benchmarks/mm-mind2web" + ], + "additional_details": { + "raw_benchmark_id": "mm-mind2web", + "raw_model_id": "nova-lite", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multimodal web navigation benchmark comprising 2,000 open-ended tasks spanning 137 websites across 31 domains. Each task includes HTML documents paired with webpage screenshots, action sequences, and complex web interactions.", + "metric_id": "llm_stats.mm-mind2web.score", + "metric_name": "MM-Mind2Web score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mm-mind2web", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MM-Mind2Web", + "raw_categories": "[\"multimodal\",\"reasoning\",\"frontend_development\",\"agents\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.607, + "details": { + "raw_score": "0.607", + "raw_score_field": "score", + "raw_model_id": "nova-lite", + "raw_benchmark_id": "mm-mind2web", + "source_urls_json": "[\"https://llm-stats.com/models/nova-lite\",\"https://llm-stats.com/benchmarks/mm-mind2web\",\"https://api.llm-stats.com/leaderboard/benchmarks/mm-mind2web\"]", + "raw_score_id": "mm-mind2web::nova-lite", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "squality::squality-nova-lite", + "evaluation_name": "llm_stats.squality", + "source_data": { + "dataset_name": "SQuALITY", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-lite", + "https://llm-stats.com/benchmarks/squality", + "https://api.llm-stats.com/leaderboard/benchmarks/squality" + ], + "additional_details": { + "raw_benchmark_id": "squality", + "raw_model_id": "nova-lite", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SQuALITY (Summarization-format QUestion Answering with Long Input Texts, Yes!) is a long-document summarization dataset built by hiring highly-qualified contractors to read public-domain short stories (3000-6000 words) and write original summaries from scratch. Each document has five summaries: one overview and four question-focused summaries. Designed to address limitations in existing summarization datasets by providing high-quality, faithful summaries.", + "metric_id": "llm_stats.squality.score", + "metric_name": "SQuALITY score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "squality", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SQuALITY", + "raw_categories": "[\"language\",\"long_context\",\"summarization\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.192, + "details": { + "raw_score": "0.192", + "raw_score_field": "score", + "raw_model_id": "nova-lite", + "raw_benchmark_id": "squality", + "source_urls_json": "[\"https://llm-stats.com/models/nova-lite\",\"https://llm-stats.com/benchmarks/squality\",\"https://api.llm-stats.com/leaderboard/benchmarks/squality\"]", + "raw_score_id": "squality::nova-lite", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "textvqa::textvqa-nova-lite", + "evaluation_name": "llm_stats.textvqa", + "source_data": { + "dataset_name": "TextVQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-lite", + "https://llm-stats.com/benchmarks/textvqa", + "https://api.llm-stats.com/leaderboard/benchmarks/textvqa" + ], + "additional_details": { + "raw_benchmark_id": "textvqa", + "raw_model_id": "nova-lite", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "TextVQA contains 45,336 questions on 28,408 images that require reasoning about text to answer. Introduced to benchmark VQA models' ability to read and reason about text within images, particularly for assistive technologies for visually impaired users. The dataset addresses the gap where existing VQA datasets had few text-based questions or were too small.", + "metric_id": "llm_stats.textvqa.score", + "metric_name": "TextVQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "textvqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TextVQA", + "raw_categories": "[\"image_to_text\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "15" + } + }, + "score_details": { + "score": 0.802, + "details": { + "raw_score": "0.802", + "raw_score_field": "score", + "raw_model_id": "nova-lite", + "raw_benchmark_id": "textvqa", + "source_urls_json": "[\"https://llm-stats.com/models/nova-lite\",\"https://llm-stats.com/benchmarks/textvqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/textvqa\"]", + "raw_score_id": "textvqa::nova-lite", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "translation-en-set1-comet22::translation-en-set1-comet22-nova-lite", + "evaluation_name": "llm_stats.translation-en-set1-comet22", + "source_data": { + "dataset_name": "Translation en→Set1 COMET22", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-lite", + "https://llm-stats.com/benchmarks/translation-en-set1-comet22", + "https://api.llm-stats.com/leaderboard/benchmarks/translation-en→set1-comet22" + ], + "additional_details": { + "raw_benchmark_id": "translation-en→set1-comet22", + "raw_model_id": "nova-lite", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "COMET-22 is an ensemble machine translation evaluation metric combining a COMET estimator model trained with Direct Assessments and a multitask model that predicts sentence-level scores and word-level OK/BAD tags. It demonstrates improved correlations compared to state-of-the-art metrics and increased robustness to critical errors.", + "metric_id": "llm_stats.translation-en-set1-comet22.score", + "metric_name": "Translation en→Set1 COMET22 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "translation-en→set1-comet22", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Translation en→Set1 COMET22", + "raw_categories": "[\"language\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.888, + "details": { + "raw_score": "0.888", + "raw_score_field": "score", + "raw_model_id": "nova-lite", + "raw_benchmark_id": "translation-en→set1-comet22", + "source_urls_json": "[\"https://llm-stats.com/models/nova-lite\",\"https://llm-stats.com/benchmarks/translation-en-set1-comet22\",\"https://api.llm-stats.com/leaderboard/benchmarks/translation-en\\u2192set1-comet22\"]", + "raw_score_id": "translation-en→set1-comet22::nova-lite", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "translation-en-set1-spbleu::translation-en-set1-spbleu-nova-lite", + "evaluation_name": "llm_stats.translation-en-set1-spbleu", + "source_data": { + "dataset_name": "Translation en→Set1 spBleu", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-lite", + "https://llm-stats.com/benchmarks/translation-en-set1-spbleu", + "https://api.llm-stats.com/leaderboard/benchmarks/translation-en→set1-spbleu" + ], + "additional_details": { + "raw_benchmark_id": "translation-en→set1-spbleu", + "raw_model_id": "nova-lite", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Translation evaluation using spBLEU (SentencePiece BLEU), a BLEU metric computed over text tokenized with a language-agnostic SentencePiece subword model. Introduced in the FLORES-101 evaluation benchmark for low-resource and multilingual machine translation.", + "metric_id": "llm_stats.translation-en-set1-spbleu.score", + "metric_name": "Translation en→Set1 spBleu score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "translation-en→set1-spbleu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Translation en→Set1 spBleu", + "raw_categories": "[\"language\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.415, + "details": { + "raw_score": "0.415", + "raw_score_field": "score", + "raw_model_id": "nova-lite", + "raw_benchmark_id": "translation-en→set1-spbleu", + "source_urls_json": "[\"https://llm-stats.com/models/nova-lite\",\"https://llm-stats.com/benchmarks/translation-en-set1-spbleu\",\"https://api.llm-stats.com/leaderboard/benchmarks/translation-en\\u2192set1-spbleu\"]", + "raw_score_id": "translation-en→set1-spbleu::nova-lite", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "translation-set1-en-comet22::translation-set1-en-comet22-nova-lite", + "evaluation_name": "llm_stats.translation-set1-en-comet22", + "source_data": { + "dataset_name": "Translation Set1→en COMET22", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-lite", + "https://llm-stats.com/benchmarks/translation-set1-en-comet22", + "https://api.llm-stats.com/leaderboard/benchmarks/translation-set1→en-comet22" + ], + "additional_details": { + "raw_benchmark_id": "translation-set1→en-comet22", + "raw_model_id": "nova-lite", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "COMET-22 is a neural machine translation evaluation metric that uses an ensemble of two models: a COMET estimator trained with Direct Assessments and a multitask model that predicts sentence-level scores and word-level OK/BAD tags. It provides improved correlations with human judgments and increased robustness to critical errors compared to previous metrics.", + "metric_id": "llm_stats.translation-set1-en-comet22.score", + "metric_name": "Translation Set1→en COMET22 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "translation-set1→en-comet22", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Translation Set1→en COMET22", + "raw_categories": "[\"language\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.888, + "details": { + "raw_score": "0.888", + "raw_score_field": "score", + "raw_model_id": "nova-lite", + "raw_benchmark_id": "translation-set1→en-comet22", + "source_urls_json": "[\"https://llm-stats.com/models/nova-lite\",\"https://llm-stats.com/benchmarks/translation-set1-en-comet22\",\"https://api.llm-stats.com/leaderboard/benchmarks/translation-set1\\u2192en-comet22\"]", + "raw_score_id": "translation-set1→en-comet22::nova-lite", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "translation-set1-en-spbleu::translation-set1-en-spbleu-nova-lite", + "evaluation_name": "llm_stats.translation-set1-en-spbleu", + "source_data": { + "dataset_name": "Translation Set1→en spBleu", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-lite", + "https://llm-stats.com/benchmarks/translation-set1-en-spbleu", + "https://api.llm-stats.com/leaderboard/benchmarks/translation-set1→en-spbleu" + ], + "additional_details": { + "raw_benchmark_id": "translation-set1→en-spbleu", + "raw_model_id": "nova-lite", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "spBLEU (SentencePiece BLEU) evaluation metric for machine translation quality assessment, using language-agnostic SentencePiece tokenization with BLEU scoring. Part of the FLORES-101 evaluation benchmark for low-resource and multilingual machine translation.", + "metric_id": "llm_stats.translation-set1-en-spbleu.score", + "metric_name": "Translation Set1→en spBleu score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "translation-set1→en-spbleu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Translation Set1→en spBleu", + "raw_categories": "[\"language\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.431, + "details": { + "raw_score": "0.431", + "raw_score_field": "score", + "raw_model_id": "nova-lite", + "raw_benchmark_id": "translation-set1→en-spbleu", + "source_urls_json": "[\"https://llm-stats.com/models/nova-lite\",\"https://llm-stats.com/benchmarks/translation-set1-en-spbleu\",\"https://api.llm-stats.com/leaderboard/benchmarks/translation-set1\\u2192en-spbleu\"]", + "raw_score_id": "translation-set1→en-spbleu::nova-lite", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "vatex::vatex-nova-lite", + "evaluation_name": "llm_stats.vatex", + "source_data": { + "dataset_name": "VATEX", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-lite", + "https://llm-stats.com/benchmarks/vatex", + "https://api.llm-stats.com/leaderboard/benchmarks/vatex" + ], + "additional_details": { + "raw_benchmark_id": "vatex", + "raw_model_id": "nova-lite", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VaTeX: A Large-Scale, High-Quality Multilingual Dataset for Video-and-Language Research. Contains over 41,250 videos and 825,000 captions in both English and Chinese, with over 206,000 English-Chinese parallel translation pairs. Supports multilingual video captioning and video-guided machine translation tasks.", + "metric_id": "llm_stats.vatex.score", + "metric_name": "VATEX score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "vatex", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VATEX", + "raw_categories": "[\"language\",\"multimodal\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.778, + "details": { + "raw_score": "0.778", + "raw_score_field": "score", + "raw_model_id": "nova-lite", + "raw_benchmark_id": "vatex", + "source_urls_json": "[\"https://llm-stats.com/models/nova-lite\",\"https://llm-stats.com/benchmarks/vatex\",\"https://api.llm-stats.com/leaderboard/benchmarks/vatex\"]", + "raw_score_id": "vatex::nova-lite", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "visualwebbench::visualwebbench-nova-lite", + "evaluation_name": "llm_stats.visualwebbench", + "source_data": { + "dataset_name": "VisualWebBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-lite", + "https://llm-stats.com/benchmarks/visualwebbench", + "https://api.llm-stats.com/leaderboard/benchmarks/visualwebbench" + ], + "additional_details": { + "raw_benchmark_id": "visualwebbench", + "raw_model_id": "nova-lite", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multimodal benchmark designed to assess the capabilities of multimodal large language models (MLLMs) across web page understanding and grounding tasks. Comprises 7 tasks (captioning, webpage QA, heading OCR, element OCR, element grounding, action prediction, and action grounding) with 1.5K human-curated instances from 139 real websites across 87 sub-domains.", + "metric_id": "llm_stats.visualwebbench.score", + "metric_name": "VisualWebBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "visualwebbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VisualWebBench", + "raw_categories": "[\"multimodal\",\"frontend_development\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.777, + "details": { + "raw_score": "0.777", + "raw_score_field": "score", + "raw_model_id": "nova-lite", + "raw_benchmark_id": "visualwebbench", + "source_urls_json": "[\"https://llm-stats.com/models/nova-lite\",\"https://llm-stats.com/benchmarks/visualwebbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/visualwebbench\"]", + "raw_score_id": "visualwebbench::nova-lite", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/amazon/nova-micro/f2c59c6b-f509-425a-8117-144bf8364fef.json b/data/llm_stats/amazon/nova-micro/f2c59c6b-f509-425a-8117-144bf8364fef.json new file mode 100644 index 0000000000000000000000000000000000000000..61c5b87b557be9864ab06e28e4b59c897ffdabfa --- /dev/null +++ b/data/llm_stats/amazon/nova-micro/f2c59c6b-f509-425a-8117-144bf8364fef.json @@ -0,0 +1,670 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/amazon_nova-micro/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Nova Micro", + "id": "amazon/nova-micro", + "developer": "amazon", + "additional_details": { + "raw_id": "nova-micro", + "raw_name": "Nova Micro", + "raw_model_id": "nova-micro", + "raw_model_name": "Nova Micro", + "raw_organization_id": "amazon", + "raw_organization_name": "Amazon", + "raw_release_date": "2024-11-20", + "raw_announcement_date": "2024-11-20", + "raw_multimodal": "false", + "raw_provider_slug": "amazon", + "raw_provider_name": "Amazon" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "arc-c::arc-c-nova-micro", + "evaluation_name": "llm_stats.arc-c", + "source_data": { + "dataset_name": "ARC-C", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-micro", + "https://llm-stats.com/benchmarks/arc-c", + "https://api.llm-stats.com/leaderboard/benchmarks/arc-c" + ], + "additional_details": { + "raw_benchmark_id": "arc-c", + "raw_model_id": "nova-micro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The AI2 Reasoning Challenge (ARC) Challenge Set is a multiple-choice question-answering benchmark containing grade-school level science questions that require advanced reasoning capabilities. ARC-C specifically contains questions that were answered incorrectly by both retrieval-based and word co-occurrence algorithms, making it a particularly challenging subset designed to test commonsense reasoning abilities in AI systems.", + "metric_id": "llm_stats.arc-c.score", + "metric_name": "ARC-C score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arc-c", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ARC-C", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "34" + } + }, + "score_details": { + "score": 0.902, + "details": { + "raw_score": "0.902", + "raw_score_field": "score", + "raw_model_id": "nova-micro", + "raw_benchmark_id": "arc-c", + "source_urls_json": "[\"https://llm-stats.com/models/nova-micro\",\"https://llm-stats.com/benchmarks/arc-c\",\"https://api.llm-stats.com/leaderboard/benchmarks/arc-c\"]", + "raw_score_id": "arc-c::nova-micro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bbh::bbh-nova-micro", + "evaluation_name": "llm_stats.bbh", + "source_data": { + "dataset_name": "BBH", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-micro", + "https://llm-stats.com/benchmarks/bbh", + "https://api.llm-stats.com/leaderboard/benchmarks/bbh" + ], + "additional_details": { + "raw_benchmark_id": "bbh", + "raw_model_id": "nova-micro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Big-Bench Hard (BBH) is a suite of 23 challenging tasks selected from BIG-Bench for which prior language model evaluations did not outperform the average human-rater. These tasks require multi-step reasoning across diverse domains including arithmetic, logical reasoning, reading comprehension, and commonsense reasoning. The benchmark was designed to test capabilities believed to be beyond current language models and focuses on evaluating complex reasoning skills including temporal understanding, spatial reasoning, causal understanding, and deductive logical reasoning.", + "metric_id": "llm_stats.bbh.score", + "metric_name": "BBH score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bbh", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BBH", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.795, + "details": { + "raw_score": "0.795", + "raw_score_field": "score", + "raw_model_id": "nova-micro", + "raw_benchmark_id": "bbh", + "source_urls_json": "[\"https://llm-stats.com/models/nova-micro\",\"https://llm-stats.com/benchmarks/bbh\",\"https://api.llm-stats.com/leaderboard/benchmarks/bbh\"]", + "raw_score_id": "bbh::nova-micro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bfcl::bfcl-nova-micro", + "evaluation_name": "llm_stats.bfcl", + "source_data": { + "dataset_name": "BFCL", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-micro", + "https://llm-stats.com/benchmarks/bfcl", + "https://api.llm-stats.com/leaderboard/benchmarks/bfcl" + ], + "additional_details": { + "raw_benchmark_id": "bfcl", + "raw_model_id": "nova-micro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The Berkeley Function Calling Leaderboard (BFCL) is the first comprehensive and executable function call evaluation dedicated to assessing Large Language Models' ability to invoke functions. It evaluates serial and parallel function calls across multiple programming languages (Python, Java, JavaScript, REST API) using a novel Abstract Syntax Tree (AST) evaluation method. The benchmark consists of over 2,000 question-function-answer pairs covering diverse application domains and complex use cases including multiple function calls, parallel function calls, and multi-turn interactions.", + "metric_id": "llm_stats.bfcl.score", + "metric_name": "BFCL score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bfcl", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BFCL", + "raw_categories": "[\"reasoning\",\"general\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.562, + "details": { + "raw_score": "0.562", + "raw_score_field": "score", + "raw_model_id": "nova-micro", + "raw_benchmark_id": "bfcl", + "source_urls_json": "[\"https://llm-stats.com/models/nova-micro\",\"https://llm-stats.com/benchmarks/bfcl\",\"https://api.llm-stats.com/leaderboard/benchmarks/bfcl\"]", + "raw_score_id": "bfcl::nova-micro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "crag::crag-nova-micro", + "evaluation_name": "llm_stats.crag", + "source_data": { + "dataset_name": "CRAG", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-micro", + "https://llm-stats.com/benchmarks/crag", + "https://api.llm-stats.com/leaderboard/benchmarks/crag" + ], + "additional_details": { + "raw_benchmark_id": "crag", + "raw_model_id": "nova-micro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CRAG (Comprehensive RAG Benchmark) is a factual question answering benchmark consisting of 4,409 question-answer pairs across 5 domains (finance, sports, music, movie, open domain) and 8 question categories. The benchmark includes mock APIs to simulate web and Knowledge Graph search, designed to represent the diverse and dynamic nature of real-world QA tasks with temporal dynamism ranging from years to seconds. It evaluates retrieval-augmented generation systems for trustworthy question answering.", + "metric_id": "llm_stats.crag.score", + "metric_name": "CRAG score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "crag", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CRAG", + "raw_categories": "[\"reasoning\",\"search\",\"finance\",\"economics\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.431, + "details": { + "raw_score": "0.431", + "raw_score_field": "score", + "raw_model_id": "nova-micro", + "raw_benchmark_id": "crag", + "source_urls_json": "[\"https://llm-stats.com/models/nova-micro\",\"https://llm-stats.com/benchmarks/crag\",\"https://api.llm-stats.com/leaderboard/benchmarks/crag\"]", + "raw_score_id": "crag::nova-micro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "drop::drop-nova-micro", + "evaluation_name": "llm_stats.drop", + "source_data": { + "dataset_name": "DROP", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-micro", + "https://llm-stats.com/benchmarks/drop", + "https://api.llm-stats.com/leaderboard/benchmarks/drop" + ], + "additional_details": { + "raw_benchmark_id": "drop", + "raw_model_id": "nova-micro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "DROP (Discrete Reasoning Over Paragraphs) is a reading comprehension benchmark requiring discrete reasoning over paragraph content. It contains crowdsourced, adversarially-created questions that require resolving references and performing discrete operations like addition, counting, or sorting, demanding comprehensive paragraph understanding beyond paraphrase-and-entity-typing shortcuts.", + "metric_id": "llm_stats.drop.score", + "metric_name": "DROP score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "drop", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DROP", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "30" + } + }, + "score_details": { + "score": 0.793, + "details": { + "raw_score": "0.793", + "raw_score_field": "score", + "raw_model_id": "nova-micro", + "raw_benchmark_id": "drop", + "source_urls_json": "[\"https://llm-stats.com/models/nova-micro\",\"https://llm-stats.com/benchmarks/drop\",\"https://api.llm-stats.com/leaderboard/benchmarks/drop\"]", + "raw_score_id": "drop::nova-micro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "finqa::finqa-nova-micro", + "evaluation_name": "llm_stats.finqa", + "source_data": { + "dataset_name": "FinQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-micro", + "https://llm-stats.com/benchmarks/finqa", + "https://api.llm-stats.com/leaderboard/benchmarks/finqa" + ], + "additional_details": { + "raw_benchmark_id": "finqa", + "raw_model_id": "nova-micro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A large-scale dataset for numerical reasoning over financial data with question-answering pairs written by financial experts, featuring complex numerical reasoning and understanding of heterogeneous representations with annotated gold reasoning programs for full explainability", + "metric_id": "llm_stats.finqa.score", + "metric_name": "FinQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "finqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FinQA", + "raw_categories": "[\"math\",\"reasoning\",\"finance\",\"economics\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.652, + "details": { + "raw_score": "0.652", + "raw_score_field": "score", + "raw_model_id": "nova-micro", + "raw_benchmark_id": "finqa", + "source_urls_json": "[\"https://llm-stats.com/models/nova-micro\",\"https://llm-stats.com/benchmarks/finqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/finqa\"]", + "raw_score_id": "finqa::nova-micro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gsm8k::gsm8k-nova-micro", + "evaluation_name": "llm_stats.gsm8k", + "source_data": { + "dataset_name": "GSM8k", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-micro", + "https://llm-stats.com/benchmarks/gsm8k", + "https://api.llm-stats.com/leaderboard/benchmarks/gsm8k" + ], + "additional_details": { + "raw_benchmark_id": "gsm8k", + "raw_model_id": "nova-micro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Grade School Math 8K, a dataset of 8.5K high-quality linguistically diverse grade school math word problems requiring multi-step reasoning and elementary arithmetic operations.", + "metric_id": "llm_stats.gsm8k.score", + "metric_name": "GSM8k score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gsm8k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GSM8k", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "48" + } + }, + "score_details": { + "score": 0.923, + "details": { + "raw_score": "0.923", + "raw_score_field": "score", + "raw_model_id": "nova-micro", + "raw_benchmark_id": "gsm8k", + "source_urls_json": "[\"https://llm-stats.com/models/nova-micro\",\"https://llm-stats.com/benchmarks/gsm8k\",\"https://api.llm-stats.com/leaderboard/benchmarks/gsm8k\"]", + "raw_score_id": "gsm8k::nova-micro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "squality::squality-nova-micro", + "evaluation_name": "llm_stats.squality", + "source_data": { + "dataset_name": "SQuALITY", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-micro", + "https://llm-stats.com/benchmarks/squality", + "https://api.llm-stats.com/leaderboard/benchmarks/squality" + ], + "additional_details": { + "raw_benchmark_id": "squality", + "raw_model_id": "nova-micro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SQuALITY (Summarization-format QUestion Answering with Long Input Texts, Yes!) is a long-document summarization dataset built by hiring highly-qualified contractors to read public-domain short stories (3000-6000 words) and write original summaries from scratch. Each document has five summaries: one overview and four question-focused summaries. Designed to address limitations in existing summarization datasets by providing high-quality, faithful summaries.", + "metric_id": "llm_stats.squality.score", + "metric_name": "SQuALITY score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "squality", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SQuALITY", + "raw_categories": "[\"language\",\"long_context\",\"summarization\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.188, + "details": { + "raw_score": "0.188", + "raw_score_field": "score", + "raw_model_id": "nova-micro", + "raw_benchmark_id": "squality", + "source_urls_json": "[\"https://llm-stats.com/models/nova-micro\",\"https://llm-stats.com/benchmarks/squality\",\"https://api.llm-stats.com/leaderboard/benchmarks/squality\"]", + "raw_score_id": "squality::nova-micro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "translation-en-set1-comet22::translation-en-set1-comet22-nova-micro", + "evaluation_name": "llm_stats.translation-en-set1-comet22", + "source_data": { + "dataset_name": "Translation en→Set1 COMET22", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-micro", + "https://llm-stats.com/benchmarks/translation-en-set1-comet22", + "https://api.llm-stats.com/leaderboard/benchmarks/translation-en→set1-comet22" + ], + "additional_details": { + "raw_benchmark_id": "translation-en→set1-comet22", + "raw_model_id": "nova-micro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "COMET-22 is an ensemble machine translation evaluation metric combining a COMET estimator model trained with Direct Assessments and a multitask model that predicts sentence-level scores and word-level OK/BAD tags. It demonstrates improved correlations compared to state-of-the-art metrics and increased robustness to critical errors.", + "metric_id": "llm_stats.translation-en-set1-comet22.score", + "metric_name": "Translation en→Set1 COMET22 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "translation-en→set1-comet22", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Translation en→Set1 COMET22", + "raw_categories": "[\"language\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.885, + "details": { + "raw_score": "0.885", + "raw_score_field": "score", + "raw_model_id": "nova-micro", + "raw_benchmark_id": "translation-en→set1-comet22", + "source_urls_json": "[\"https://llm-stats.com/models/nova-micro\",\"https://llm-stats.com/benchmarks/translation-en-set1-comet22\",\"https://api.llm-stats.com/leaderboard/benchmarks/translation-en\\u2192set1-comet22\"]", + "raw_score_id": "translation-en→set1-comet22::nova-micro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "translation-en-set1-spbleu::translation-en-set1-spbleu-nova-micro", + "evaluation_name": "llm_stats.translation-en-set1-spbleu", + "source_data": { + "dataset_name": "Translation en→Set1 spBleu", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-micro", + "https://llm-stats.com/benchmarks/translation-en-set1-spbleu", + "https://api.llm-stats.com/leaderboard/benchmarks/translation-en→set1-spbleu" + ], + "additional_details": { + "raw_benchmark_id": "translation-en→set1-spbleu", + "raw_model_id": "nova-micro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Translation evaluation using spBLEU (SentencePiece BLEU), a BLEU metric computed over text tokenized with a language-agnostic SentencePiece subword model. Introduced in the FLORES-101 evaluation benchmark for low-resource and multilingual machine translation.", + "metric_id": "llm_stats.translation-en-set1-spbleu.score", + "metric_name": "Translation en→Set1 spBleu score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "translation-en→set1-spbleu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Translation en→Set1 spBleu", + "raw_categories": "[\"language\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.402, + "details": { + "raw_score": "0.402", + "raw_score_field": "score", + "raw_model_id": "nova-micro", + "raw_benchmark_id": "translation-en→set1-spbleu", + "source_urls_json": "[\"https://llm-stats.com/models/nova-micro\",\"https://llm-stats.com/benchmarks/translation-en-set1-spbleu\",\"https://api.llm-stats.com/leaderboard/benchmarks/translation-en\\u2192set1-spbleu\"]", + "raw_score_id": "translation-en→set1-spbleu::nova-micro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "translation-set1-en-comet22::translation-set1-en-comet22-nova-micro", + "evaluation_name": "llm_stats.translation-set1-en-comet22", + "source_data": { + "dataset_name": "Translation Set1→en COMET22", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-micro", + "https://llm-stats.com/benchmarks/translation-set1-en-comet22", + "https://api.llm-stats.com/leaderboard/benchmarks/translation-set1→en-comet22" + ], + "additional_details": { + "raw_benchmark_id": "translation-set1→en-comet22", + "raw_model_id": "nova-micro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "COMET-22 is a neural machine translation evaluation metric that uses an ensemble of two models: a COMET estimator trained with Direct Assessments and a multitask model that predicts sentence-level scores and word-level OK/BAD tags. It provides improved correlations with human judgments and increased robustness to critical errors compared to previous metrics.", + "metric_id": "llm_stats.translation-set1-en-comet22.score", + "metric_name": "Translation Set1→en COMET22 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "translation-set1→en-comet22", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Translation Set1→en COMET22", + "raw_categories": "[\"language\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.887, + "details": { + "raw_score": "0.887", + "raw_score_field": "score", + "raw_model_id": "nova-micro", + "raw_benchmark_id": "translation-set1→en-comet22", + "source_urls_json": "[\"https://llm-stats.com/models/nova-micro\",\"https://llm-stats.com/benchmarks/translation-set1-en-comet22\",\"https://api.llm-stats.com/leaderboard/benchmarks/translation-set1\\u2192en-comet22\"]", + "raw_score_id": "translation-set1→en-comet22::nova-micro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "translation-set1-en-spbleu::translation-set1-en-spbleu-nova-micro", + "evaluation_name": "llm_stats.translation-set1-en-spbleu", + "source_data": { + "dataset_name": "Translation Set1→en spBleu", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-micro", + "https://llm-stats.com/benchmarks/translation-set1-en-spbleu", + "https://api.llm-stats.com/leaderboard/benchmarks/translation-set1→en-spbleu" + ], + "additional_details": { + "raw_benchmark_id": "translation-set1→en-spbleu", + "raw_model_id": "nova-micro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "spBLEU (SentencePiece BLEU) evaluation metric for machine translation quality assessment, using language-agnostic SentencePiece tokenization with BLEU scoring. Part of the FLORES-101 evaluation benchmark for low-resource and multilingual machine translation.", + "metric_id": "llm_stats.translation-set1-en-spbleu.score", + "metric_name": "Translation Set1→en spBleu score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "translation-set1→en-spbleu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Translation Set1→en spBleu", + "raw_categories": "[\"language\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.426, + "details": { + "raw_score": "0.426", + "raw_score_field": "score", + "raw_model_id": "nova-micro", + "raw_benchmark_id": "translation-set1→en-spbleu", + "source_urls_json": "[\"https://llm-stats.com/models/nova-micro\",\"https://llm-stats.com/benchmarks/translation-set1-en-spbleu\",\"https://api.llm-stats.com/leaderboard/benchmarks/translation-set1\\u2192en-spbleu\"]", + "raw_score_id": "translation-set1→en-spbleu::nova-micro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/amazon/nova-pro/3530a34e-e2ef-475d-bb97-56b190a48cf5.json b/data/llm_stats/amazon/nova-pro/3530a34e-e2ef-475d-bb97-56b190a48cf5.json new file mode 100644 index 0000000000000000000000000000000000000000..24ab8e9b4f894bb922c571bd42cf2313d53e2b40 --- /dev/null +++ b/data/llm_stats/amazon/nova-pro/3530a34e-e2ef-475d-bb97-56b190a48cf5.json @@ -0,0 +1,1190 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/amazon_nova-pro/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Nova Pro", + "id": "amazon/nova-pro", + "developer": "amazon", + "additional_details": { + "raw_id": "nova-pro", + "raw_name": "Nova Pro", + "raw_model_id": "nova-pro", + "raw_model_name": "Nova Pro", + "raw_organization_id": "amazon", + "raw_organization_name": "Amazon", + "raw_release_date": "2024-11-20", + "raw_announcement_date": "2024-11-20", + "raw_multimodal": "true", + "raw_provider_slug": "amazon", + "raw_provider_name": "Amazon" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "arc-c::arc-c-nova-pro", + "evaluation_name": "llm_stats.arc-c", + "source_data": { + "dataset_name": "ARC-C", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-pro", + "https://llm-stats.com/benchmarks/arc-c", + "https://api.llm-stats.com/leaderboard/benchmarks/arc-c" + ], + "additional_details": { + "raw_benchmark_id": "arc-c", + "raw_model_id": "nova-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The AI2 Reasoning Challenge (ARC) Challenge Set is a multiple-choice question-answering benchmark containing grade-school level science questions that require advanced reasoning capabilities. ARC-C specifically contains questions that were answered incorrectly by both retrieval-based and word co-occurrence algorithms, making it a particularly challenging subset designed to test commonsense reasoning abilities in AI systems.", + "metric_id": "llm_stats.arc-c.score", + "metric_name": "ARC-C score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arc-c", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ARC-C", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "34" + } + }, + "score_details": { + "score": 0.948, + "details": { + "raw_score": "0.948", + "raw_score_field": "score", + "raw_model_id": "nova-pro", + "raw_benchmark_id": "arc-c", + "source_urls_json": "[\"https://llm-stats.com/models/nova-pro\",\"https://llm-stats.com/benchmarks/arc-c\",\"https://api.llm-stats.com/leaderboard/benchmarks/arc-c\"]", + "raw_score_id": "arc-c::nova-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bbh::bbh-nova-pro", + "evaluation_name": "llm_stats.bbh", + "source_data": { + "dataset_name": "BBH", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-pro", + "https://llm-stats.com/benchmarks/bbh", + "https://api.llm-stats.com/leaderboard/benchmarks/bbh" + ], + "additional_details": { + "raw_benchmark_id": "bbh", + "raw_model_id": "nova-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Big-Bench Hard (BBH) is a suite of 23 challenging tasks selected from BIG-Bench for which prior language model evaluations did not outperform the average human-rater. These tasks require multi-step reasoning across diverse domains including arithmetic, logical reasoning, reading comprehension, and commonsense reasoning. The benchmark was designed to test capabilities believed to be beyond current language models and focuses on evaluating complex reasoning skills including temporal understanding, spatial reasoning, causal understanding, and deductive logical reasoning.", + "metric_id": "llm_stats.bbh.score", + "metric_name": "BBH score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bbh", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BBH", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.869, + "details": { + "raw_score": "0.869", + "raw_score_field": "score", + "raw_model_id": "nova-pro", + "raw_benchmark_id": "bbh", + "source_urls_json": "[\"https://llm-stats.com/models/nova-pro\",\"https://llm-stats.com/benchmarks/bbh\",\"https://api.llm-stats.com/leaderboard/benchmarks/bbh\"]", + "raw_score_id": "bbh::nova-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bfcl::bfcl-nova-pro", + "evaluation_name": "llm_stats.bfcl", + "source_data": { + "dataset_name": "BFCL", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-pro", + "https://llm-stats.com/benchmarks/bfcl", + "https://api.llm-stats.com/leaderboard/benchmarks/bfcl" + ], + "additional_details": { + "raw_benchmark_id": "bfcl", + "raw_model_id": "nova-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The Berkeley Function Calling Leaderboard (BFCL) is the first comprehensive and executable function call evaluation dedicated to assessing Large Language Models' ability to invoke functions. It evaluates serial and parallel function calls across multiple programming languages (Python, Java, JavaScript, REST API) using a novel Abstract Syntax Tree (AST) evaluation method. The benchmark consists of over 2,000 question-function-answer pairs covering diverse application domains and complex use cases including multiple function calls, parallel function calls, and multi-turn interactions.", + "metric_id": "llm_stats.bfcl.score", + "metric_name": "BFCL score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bfcl", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BFCL", + "raw_categories": "[\"reasoning\",\"general\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.684, + "details": { + "raw_score": "0.684", + "raw_score_field": "score", + "raw_model_id": "nova-pro", + "raw_benchmark_id": "bfcl", + "source_urls_json": "[\"https://llm-stats.com/models/nova-pro\",\"https://llm-stats.com/benchmarks/bfcl\",\"https://api.llm-stats.com/leaderboard/benchmarks/bfcl\"]", + "raw_score_id": "bfcl::nova-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "chartqa::chartqa-nova-pro", + "evaluation_name": "llm_stats.chartqa", + "source_data": { + "dataset_name": "ChartQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-pro", + "https://llm-stats.com/benchmarks/chartqa", + "https://api.llm-stats.com/leaderboard/benchmarks/chartqa" + ], + "additional_details": { + "raw_benchmark_id": "chartqa", + "raw_model_id": "nova-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ChartQA is a large-scale benchmark comprising 9.6K human-written questions and 23.1K questions generated from human-written chart summaries, designed to evaluate models' abilities in visual and logical reasoning over charts.", + "metric_id": "llm_stats.chartqa.score", + "metric_name": "ChartQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "chartqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ChartQA", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "24" + } + }, + "score_details": { + "score": 0.892, + "details": { + "raw_score": "0.892", + "raw_score_field": "score", + "raw_model_id": "nova-pro", + "raw_benchmark_id": "chartqa", + "source_urls_json": "[\"https://llm-stats.com/models/nova-pro\",\"https://llm-stats.com/benchmarks/chartqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/chartqa\"]", + "raw_score_id": "chartqa::nova-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "crag::crag-nova-pro", + "evaluation_name": "llm_stats.crag", + "source_data": { + "dataset_name": "CRAG", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-pro", + "https://llm-stats.com/benchmarks/crag", + "https://api.llm-stats.com/leaderboard/benchmarks/crag" + ], + "additional_details": { + "raw_benchmark_id": "crag", + "raw_model_id": "nova-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CRAG (Comprehensive RAG Benchmark) is a factual question answering benchmark consisting of 4,409 question-answer pairs across 5 domains (finance, sports, music, movie, open domain) and 8 question categories. The benchmark includes mock APIs to simulate web and Knowledge Graph search, designed to represent the diverse and dynamic nature of real-world QA tasks with temporal dynamism ranging from years to seconds. It evaluates retrieval-augmented generation systems for trustworthy question answering.", + "metric_id": "llm_stats.crag.score", + "metric_name": "CRAG score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "crag", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CRAG", + "raw_categories": "[\"reasoning\",\"search\",\"finance\",\"economics\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.503, + "details": { + "raw_score": "0.503", + "raw_score_field": "score", + "raw_model_id": "nova-pro", + "raw_benchmark_id": "crag", + "source_urls_json": "[\"https://llm-stats.com/models/nova-pro\",\"https://llm-stats.com/benchmarks/crag\",\"https://api.llm-stats.com/leaderboard/benchmarks/crag\"]", + "raw_score_id": "crag::nova-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "docvqa::docvqa-nova-pro", + "evaluation_name": "llm_stats.docvqa", + "source_data": { + "dataset_name": "DocVQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-pro", + "https://llm-stats.com/benchmarks/docvqa", + "https://api.llm-stats.com/leaderboard/benchmarks/docvqa" + ], + "additional_details": { + "raw_benchmark_id": "docvqa", + "raw_model_id": "nova-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A dataset for Visual Question Answering on document images containing 50,000 questions defined on 12,000+ document images. The benchmark tests AI's ability to understand document structure and content, requiring models to comprehend document layout and perform information retrieval to answer questions about document images.", + "metric_id": "llm_stats.docvqa.score", + "metric_name": "DocVQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "docvqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DocVQA", + "raw_categories": "[\"image_to_text\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "26" + } + }, + "score_details": { + "score": 0.935, + "details": { + "raw_score": "0.935", + "raw_score_field": "score", + "raw_model_id": "nova-pro", + "raw_benchmark_id": "docvqa", + "source_urls_json": "[\"https://llm-stats.com/models/nova-pro\",\"https://llm-stats.com/benchmarks/docvqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/docvqa\"]", + "raw_score_id": "docvqa::nova-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "drop::drop-nova-pro", + "evaluation_name": "llm_stats.drop", + "source_data": { + "dataset_name": "DROP", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-pro", + "https://llm-stats.com/benchmarks/drop", + "https://api.llm-stats.com/leaderboard/benchmarks/drop" + ], + "additional_details": { + "raw_benchmark_id": "drop", + "raw_model_id": "nova-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "DROP (Discrete Reasoning Over Paragraphs) is a reading comprehension benchmark requiring discrete reasoning over paragraph content. It contains crowdsourced, adversarially-created questions that require resolving references and performing discrete operations like addition, counting, or sorting, demanding comprehensive paragraph understanding beyond paraphrase-and-entity-typing shortcuts.", + "metric_id": "llm_stats.drop.score", + "metric_name": "DROP score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "drop", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DROP", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "30" + } + }, + "score_details": { + "score": 0.854, + "details": { + "raw_score": "0.854", + "raw_score_field": "score", + "raw_model_id": "nova-pro", + "raw_benchmark_id": "drop", + "source_urls_json": "[\"https://llm-stats.com/models/nova-pro\",\"https://llm-stats.com/benchmarks/drop\",\"https://api.llm-stats.com/leaderboard/benchmarks/drop\"]", + "raw_score_id": "drop::nova-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "egoschema::egoschema-nova-pro", + "evaluation_name": "llm_stats.egoschema", + "source_data": { + "dataset_name": "EgoSchema", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-pro", + "https://llm-stats.com/benchmarks/egoschema", + "https://api.llm-stats.com/leaderboard/benchmarks/egoschema" + ], + "additional_details": { + "raw_benchmark_id": "egoschema", + "raw_model_id": "nova-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A diagnostic benchmark for very long-form video language understanding consisting of over 5000 human curated multiple choice questions based on 3-minute video clips from Ego4D, covering a broad range of natural human activities and behaviors", + "metric_id": "llm_stats.egoschema.score", + "metric_name": "EgoSchema score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "egoschema", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "EgoSchema", + "raw_categories": "[\"long_context\",\"reasoning\",\"vision\"]", + "raw_modality": "video", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.721, + "details": { + "raw_score": "0.721", + "raw_score_field": "score", + "raw_model_id": "nova-pro", + "raw_benchmark_id": "egoschema", + "source_urls_json": "[\"https://llm-stats.com/models/nova-pro\",\"https://llm-stats.com/benchmarks/egoschema\",\"https://api.llm-stats.com/leaderboard/benchmarks/egoschema\"]", + "raw_score_id": "egoschema::nova-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "finqa::finqa-nova-pro", + "evaluation_name": "llm_stats.finqa", + "source_data": { + "dataset_name": "FinQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-pro", + "https://llm-stats.com/benchmarks/finqa", + "https://api.llm-stats.com/leaderboard/benchmarks/finqa" + ], + "additional_details": { + "raw_benchmark_id": "finqa", + "raw_model_id": "nova-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A large-scale dataset for numerical reasoning over financial data with question-answering pairs written by financial experts, featuring complex numerical reasoning and understanding of heterogeneous representations with annotated gold reasoning programs for full explainability", + "metric_id": "llm_stats.finqa.score", + "metric_name": "FinQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "finqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FinQA", + "raw_categories": "[\"math\",\"reasoning\",\"finance\",\"economics\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.772, + "details": { + "raw_score": "0.772", + "raw_score_field": "score", + "raw_model_id": "nova-pro", + "raw_benchmark_id": "finqa", + "source_urls_json": "[\"https://llm-stats.com/models/nova-pro\",\"https://llm-stats.com/benchmarks/finqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/finqa\"]", + "raw_score_id": "finqa::nova-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "groundui-1k::groundui-1k-nova-pro", + "evaluation_name": "llm_stats.groundui-1k", + "source_data": { + "dataset_name": "GroundUI-1K", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-pro", + "https://llm-stats.com/benchmarks/groundui-1k", + "https://api.llm-stats.com/leaderboard/benchmarks/groundui-1k" + ], + "additional_details": { + "raw_benchmark_id": "groundui-1k", + "raw_model_id": "nova-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A subset of GroundUI-18K for UI grounding evaluation, where models must predict action coordinates on screenshots based on single-step instructions across web, desktop, and mobile platforms.", + "metric_id": "llm_stats.groundui-1k.score", + "metric_name": "GroundUI-1K score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "groundui-1k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GroundUI-1K", + "raw_categories": "[\"multimodal\",\"grounding\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.814, + "details": { + "raw_score": "0.814", + "raw_score_field": "score", + "raw_model_id": "nova-pro", + "raw_benchmark_id": "groundui-1k", + "source_urls_json": "[\"https://llm-stats.com/models/nova-pro\",\"https://llm-stats.com/benchmarks/groundui-1k\",\"https://api.llm-stats.com/leaderboard/benchmarks/groundui-1k\"]", + "raw_score_id": "groundui-1k::nova-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gsm8k::gsm8k-nova-pro", + "evaluation_name": "llm_stats.gsm8k", + "source_data": { + "dataset_name": "GSM8k", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-pro", + "https://llm-stats.com/benchmarks/gsm8k", + "https://api.llm-stats.com/leaderboard/benchmarks/gsm8k" + ], + "additional_details": { + "raw_benchmark_id": "gsm8k", + "raw_model_id": "nova-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Grade School Math 8K, a dataset of 8.5K high-quality linguistically diverse grade school math word problems requiring multi-step reasoning and elementary arithmetic operations.", + "metric_id": "llm_stats.gsm8k.score", + "metric_name": "GSM8k score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gsm8k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GSM8k", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "48" + } + }, + "score_details": { + "score": 0.948, + "details": { + "raw_score": "0.948", + "raw_score_field": "score", + "raw_model_id": "nova-pro", + "raw_benchmark_id": "gsm8k", + "source_urls_json": "[\"https://llm-stats.com/models/nova-pro\",\"https://llm-stats.com/benchmarks/gsm8k\",\"https://api.llm-stats.com/leaderboard/benchmarks/gsm8k\"]", + "raw_score_id": "gsm8k::nova-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "humaneval::humaneval-nova-pro", + "evaluation_name": "llm_stats.humaneval", + "source_data": { + "dataset_name": "HumanEval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-pro", + "https://llm-stats.com/benchmarks/humaneval", + "https://api.llm-stats.com/leaderboard/benchmarks/humaneval" + ], + "additional_details": { + "raw_benchmark_id": "humaneval", + "raw_model_id": "nova-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A benchmark that measures functional correctness for synthesizing programs from docstrings, consisting of 164 original programming problems assessing language comprehension, algorithms, and simple mathematics", + "metric_id": "llm_stats.humaneval.score", + "metric_name": "HumanEval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "humaneval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HumanEval", + "raw_categories": "[\"reasoning\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "66" + } + }, + "score_details": { + "score": 0.89, + "details": { + "raw_score": "0.89", + "raw_score_field": "score", + "raw_model_id": "nova-pro", + "raw_benchmark_id": "humaneval", + "source_urls_json": "[\"https://llm-stats.com/models/nova-pro\",\"https://llm-stats.com/benchmarks/humaneval\",\"https://api.llm-stats.com/leaderboard/benchmarks/humaneval\"]", + "raw_score_id": "humaneval::nova-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ifeval::ifeval-nova-pro", + "evaluation_name": "llm_stats.ifeval", + "source_data": { + "dataset_name": "IFEval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-pro", + "https://llm-stats.com/benchmarks/ifeval", + "https://api.llm-stats.com/leaderboard/benchmarks/ifeval" + ], + "additional_details": { + "raw_benchmark_id": "ifeval", + "raw_model_id": "nova-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Instruction-Following Evaluation (IFEval) benchmark for large language models, focusing on verifiable instructions with 25 types of instructions and around 500 prompts containing one or more verifiable constraints", + "metric_id": "llm_stats.ifeval.score", + "metric_name": "IFEval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ifeval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IFEval", + "raw_categories": "[\"instruction_following\",\"structured_output\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "65" + } + }, + "score_details": { + "score": 0.921, + "details": { + "raw_score": "0.921", + "raw_score_field": "score", + "raw_model_id": "nova-pro", + "raw_benchmark_id": "ifeval", + "source_urls_json": "[\"https://llm-stats.com/models/nova-pro\",\"https://llm-stats.com/benchmarks/ifeval\",\"https://api.llm-stats.com/leaderboard/benchmarks/ifeval\"]", + "raw_score_id": "ifeval::nova-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mm-mind2web::mm-mind2web-nova-pro", + "evaluation_name": "llm_stats.mm-mind2web", + "source_data": { + "dataset_name": "MM-Mind2Web", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-pro", + "https://llm-stats.com/benchmarks/mm-mind2web", + "https://api.llm-stats.com/leaderboard/benchmarks/mm-mind2web" + ], + "additional_details": { + "raw_benchmark_id": "mm-mind2web", + "raw_model_id": "nova-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multimodal web navigation benchmark comprising 2,000 open-ended tasks spanning 137 websites across 31 domains. Each task includes HTML documents paired with webpage screenshots, action sequences, and complex web interactions.", + "metric_id": "llm_stats.mm-mind2web.score", + "metric_name": "MM-Mind2Web score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mm-mind2web", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MM-Mind2Web", + "raw_categories": "[\"multimodal\",\"reasoning\",\"frontend_development\",\"agents\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.637, + "details": { + "raw_score": "0.637", + "raw_score_field": "score", + "raw_model_id": "nova-pro", + "raw_benchmark_id": "mm-mind2web", + "source_urls_json": "[\"https://llm-stats.com/models/nova-pro\",\"https://llm-stats.com/benchmarks/mm-mind2web\",\"https://api.llm-stats.com/leaderboard/benchmarks/mm-mind2web\"]", + "raw_score_id": "mm-mind2web::nova-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "squality::squality-nova-pro", + "evaluation_name": "llm_stats.squality", + "source_data": { + "dataset_name": "SQuALITY", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-pro", + "https://llm-stats.com/benchmarks/squality", + "https://api.llm-stats.com/leaderboard/benchmarks/squality" + ], + "additional_details": { + "raw_benchmark_id": "squality", + "raw_model_id": "nova-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SQuALITY (Summarization-format QUestion Answering with Long Input Texts, Yes!) is a long-document summarization dataset built by hiring highly-qualified contractors to read public-domain short stories (3000-6000 words) and write original summaries from scratch. Each document has five summaries: one overview and four question-focused summaries. Designed to address limitations in existing summarization datasets by providing high-quality, faithful summaries.", + "metric_id": "llm_stats.squality.score", + "metric_name": "SQuALITY score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "squality", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SQuALITY", + "raw_categories": "[\"language\",\"long_context\",\"summarization\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.198, + "details": { + "raw_score": "0.198", + "raw_score_field": "score", + "raw_model_id": "nova-pro", + "raw_benchmark_id": "squality", + "source_urls_json": "[\"https://llm-stats.com/models/nova-pro\",\"https://llm-stats.com/benchmarks/squality\",\"https://api.llm-stats.com/leaderboard/benchmarks/squality\"]", + "raw_score_id": "squality::nova-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "textvqa::textvqa-nova-pro", + "evaluation_name": "llm_stats.textvqa", + "source_data": { + "dataset_name": "TextVQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-pro", + "https://llm-stats.com/benchmarks/textvqa", + "https://api.llm-stats.com/leaderboard/benchmarks/textvqa" + ], + "additional_details": { + "raw_benchmark_id": "textvqa", + "raw_model_id": "nova-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "TextVQA contains 45,336 questions on 28,408 images that require reasoning about text to answer. Introduced to benchmark VQA models' ability to read and reason about text within images, particularly for assistive technologies for visually impaired users. The dataset addresses the gap where existing VQA datasets had few text-based questions or were too small.", + "metric_id": "llm_stats.textvqa.score", + "metric_name": "TextVQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "textvqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TextVQA", + "raw_categories": "[\"image_to_text\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "15" + } + }, + "score_details": { + "score": 0.815, + "details": { + "raw_score": "0.815", + "raw_score_field": "score", + "raw_model_id": "nova-pro", + "raw_benchmark_id": "textvqa", + "source_urls_json": "[\"https://llm-stats.com/models/nova-pro\",\"https://llm-stats.com/benchmarks/textvqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/textvqa\"]", + "raw_score_id": "textvqa::nova-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "translation-en-set1-comet22::translation-en-set1-comet22-nova-pro", + "evaluation_name": "llm_stats.translation-en-set1-comet22", + "source_data": { + "dataset_name": "Translation en→Set1 COMET22", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-pro", + "https://llm-stats.com/benchmarks/translation-en-set1-comet22", + "https://api.llm-stats.com/leaderboard/benchmarks/translation-en→set1-comet22" + ], + "additional_details": { + "raw_benchmark_id": "translation-en→set1-comet22", + "raw_model_id": "nova-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "COMET-22 is an ensemble machine translation evaluation metric combining a COMET estimator model trained with Direct Assessments and a multitask model that predicts sentence-level scores and word-level OK/BAD tags. It demonstrates improved correlations compared to state-of-the-art metrics and increased robustness to critical errors.", + "metric_id": "llm_stats.translation-en-set1-comet22.score", + "metric_name": "Translation en→Set1 COMET22 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "translation-en→set1-comet22", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Translation en→Set1 COMET22", + "raw_categories": "[\"language\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.891, + "details": { + "raw_score": "0.891", + "raw_score_field": "score", + "raw_model_id": "nova-pro", + "raw_benchmark_id": "translation-en→set1-comet22", + "source_urls_json": "[\"https://llm-stats.com/models/nova-pro\",\"https://llm-stats.com/benchmarks/translation-en-set1-comet22\",\"https://api.llm-stats.com/leaderboard/benchmarks/translation-en\\u2192set1-comet22\"]", + "raw_score_id": "translation-en→set1-comet22::nova-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "translation-en-set1-spbleu::translation-en-set1-spbleu-nova-pro", + "evaluation_name": "llm_stats.translation-en-set1-spbleu", + "source_data": { + "dataset_name": "Translation en→Set1 spBleu", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-pro", + "https://llm-stats.com/benchmarks/translation-en-set1-spbleu", + "https://api.llm-stats.com/leaderboard/benchmarks/translation-en→set1-spbleu" + ], + "additional_details": { + "raw_benchmark_id": "translation-en→set1-spbleu", + "raw_model_id": "nova-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Translation evaluation using spBLEU (SentencePiece BLEU), a BLEU metric computed over text tokenized with a language-agnostic SentencePiece subword model. Introduced in the FLORES-101 evaluation benchmark for low-resource and multilingual machine translation.", + "metric_id": "llm_stats.translation-en-set1-spbleu.score", + "metric_name": "Translation en→Set1 spBleu score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "translation-en→set1-spbleu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Translation en→Set1 spBleu", + "raw_categories": "[\"language\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.434, + "details": { + "raw_score": "0.434", + "raw_score_field": "score", + "raw_model_id": "nova-pro", + "raw_benchmark_id": "translation-en→set1-spbleu", + "source_urls_json": "[\"https://llm-stats.com/models/nova-pro\",\"https://llm-stats.com/benchmarks/translation-en-set1-spbleu\",\"https://api.llm-stats.com/leaderboard/benchmarks/translation-en\\u2192set1-spbleu\"]", + "raw_score_id": "translation-en→set1-spbleu::nova-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "translation-set1-en-comet22::translation-set1-en-comet22-nova-pro", + "evaluation_name": "llm_stats.translation-set1-en-comet22", + "source_data": { + "dataset_name": "Translation Set1→en COMET22", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-pro", + "https://llm-stats.com/benchmarks/translation-set1-en-comet22", + "https://api.llm-stats.com/leaderboard/benchmarks/translation-set1→en-comet22" + ], + "additional_details": { + "raw_benchmark_id": "translation-set1→en-comet22", + "raw_model_id": "nova-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "COMET-22 is a neural machine translation evaluation metric that uses an ensemble of two models: a COMET estimator trained with Direct Assessments and a multitask model that predicts sentence-level scores and word-level OK/BAD tags. It provides improved correlations with human judgments and increased robustness to critical errors compared to previous metrics.", + "metric_id": "llm_stats.translation-set1-en-comet22.score", + "metric_name": "Translation Set1→en COMET22 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "translation-set1→en-comet22", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Translation Set1→en COMET22", + "raw_categories": "[\"language\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.89, + "details": { + "raw_score": "0.89", + "raw_score_field": "score", + "raw_model_id": "nova-pro", + "raw_benchmark_id": "translation-set1→en-comet22", + "source_urls_json": "[\"https://llm-stats.com/models/nova-pro\",\"https://llm-stats.com/benchmarks/translation-set1-en-comet22\",\"https://api.llm-stats.com/leaderboard/benchmarks/translation-set1\\u2192en-comet22\"]", + "raw_score_id": "translation-set1→en-comet22::nova-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "translation-set1-en-spbleu::translation-set1-en-spbleu-nova-pro", + "evaluation_name": "llm_stats.translation-set1-en-spbleu", + "source_data": { + "dataset_name": "Translation Set1→en spBleu", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-pro", + "https://llm-stats.com/benchmarks/translation-set1-en-spbleu", + "https://api.llm-stats.com/leaderboard/benchmarks/translation-set1→en-spbleu" + ], + "additional_details": { + "raw_benchmark_id": "translation-set1→en-spbleu", + "raw_model_id": "nova-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "spBLEU (SentencePiece BLEU) evaluation metric for machine translation quality assessment, using language-agnostic SentencePiece tokenization with BLEU scoring. Part of the FLORES-101 evaluation benchmark for low-resource and multilingual machine translation.", + "metric_id": "llm_stats.translation-set1-en-spbleu.score", + "metric_name": "Translation Set1→en spBleu score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "translation-set1→en-spbleu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Translation Set1→en spBleu", + "raw_categories": "[\"language\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.444, + "details": { + "raw_score": "0.444", + "raw_score_field": "score", + "raw_model_id": "nova-pro", + "raw_benchmark_id": "translation-set1→en-spbleu", + "source_urls_json": "[\"https://llm-stats.com/models/nova-pro\",\"https://llm-stats.com/benchmarks/translation-set1-en-spbleu\",\"https://api.llm-stats.com/leaderboard/benchmarks/translation-set1\\u2192en-spbleu\"]", + "raw_score_id": "translation-set1→en-spbleu::nova-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "vatex::vatex-nova-pro", + "evaluation_name": "llm_stats.vatex", + "source_data": { + "dataset_name": "VATEX", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-pro", + "https://llm-stats.com/benchmarks/vatex", + "https://api.llm-stats.com/leaderboard/benchmarks/vatex" + ], + "additional_details": { + "raw_benchmark_id": "vatex", + "raw_model_id": "nova-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VaTeX: A Large-Scale, High-Quality Multilingual Dataset for Video-and-Language Research. Contains over 41,250 videos and 825,000 captions in both English and Chinese, with over 206,000 English-Chinese parallel translation pairs. Supports multilingual video captioning and video-guided machine translation tasks.", + "metric_id": "llm_stats.vatex.score", + "metric_name": "VATEX score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "vatex", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VATEX", + "raw_categories": "[\"language\",\"multimodal\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.778, + "details": { + "raw_score": "0.778", + "raw_score_field": "score", + "raw_model_id": "nova-pro", + "raw_benchmark_id": "vatex", + "source_urls_json": "[\"https://llm-stats.com/models/nova-pro\",\"https://llm-stats.com/benchmarks/vatex\",\"https://api.llm-stats.com/leaderboard/benchmarks/vatex\"]", + "raw_score_id": "vatex::nova-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "visualwebbench::visualwebbench-nova-pro", + "evaluation_name": "llm_stats.visualwebbench", + "source_data": { + "dataset_name": "VisualWebBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nova-pro", + "https://llm-stats.com/benchmarks/visualwebbench", + "https://api.llm-stats.com/leaderboard/benchmarks/visualwebbench" + ], + "additional_details": { + "raw_benchmark_id": "visualwebbench", + "raw_model_id": "nova-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multimodal benchmark designed to assess the capabilities of multimodal large language models (MLLMs) across web page understanding and grounding tasks. Comprises 7 tasks (captioning, webpage QA, heading OCR, element OCR, element grounding, action prediction, and action grounding) with 1.5K human-curated instances from 139 real websites across 87 sub-domains.", + "metric_id": "llm_stats.visualwebbench.score", + "metric_name": "VisualWebBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "visualwebbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VisualWebBench", + "raw_categories": "[\"multimodal\",\"frontend_development\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.797, + "details": { + "raw_score": "0.797", + "raw_score_field": "score", + "raw_model_id": "nova-pro", + "raw_benchmark_id": "visualwebbench", + "source_urls_json": "[\"https://llm-stats.com/models/nova-pro\",\"https://llm-stats.com/benchmarks/visualwebbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/visualwebbench\"]", + "raw_score_id": "visualwebbench::nova-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/anthropic/claude-3-5-sonnet-20240620/b88e95e1-eda0-4bac-a9b5-2080e2754b4f.json b/data/llm_stats/anthropic/claude-3-5-sonnet-20240620/b88e95e1-eda0-4bac-a9b5-2080e2754b4f.json new file mode 100644 index 0000000000000000000000000000000000000000..022bf47b917a8b29e9820ffdca6fa3b69827a68a --- /dev/null +++ b/data/llm_stats/anthropic/claude-3-5-sonnet-20240620/b88e95e1-eda0-4bac-a9b5-2080e2754b4f.json @@ -0,0 +1,358 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/anthropic_claude-3-5-sonnet-20240620/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Claude 3.5 Sonnet", + "id": "anthropic/claude-3-5-sonnet-20240620", + "developer": "anthropic", + "additional_details": { + "raw_id": "claude-3-5-sonnet-20240620", + "raw_name": "Claude 3.5 Sonnet", + "raw_model_id": "claude-3-5-sonnet-20240620", + "raw_model_name": "Claude 3.5 Sonnet", + "raw_organization_id": "anthropic", + "raw_organization_name": "Anthropic", + "raw_release_date": "2024-06-21", + "raw_announcement_date": "2024-06-21", + "raw_multimodal": "true", + "raw_provider_slug": "anthropic", + "raw_provider_name": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "big-bench-hard::big-bench-hard-claude-3-5-sonnet-20240620", + "evaluation_name": "llm_stats.big-bench-hard", + "source_data": { + "dataset_name": "BIG-Bench Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-5-sonnet-20240620", + "https://llm-stats.com/benchmarks/big-bench-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard" + ], + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_model_id": "claude-3-5-sonnet-20240620", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIG-Bench Hard (BBH) is a subset of 23 challenging BIG-Bench tasks selected because prior language model evaluations did not outperform average human-rater performance. The benchmark contains 6,511 evaluation examples testing various forms of multi-step reasoning including arithmetic, logical reasoning (Boolean expressions, logical deduction), geometric reasoning, temporal reasoning, and language understanding. Tasks require capabilities such as causal judgment, object counting, navigation, pattern recognition, and complex problem solving.", + "metric_id": "llm_stats.big-bench-hard.score", + "metric_name": "BIG-Bench Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench Hard", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "21" + } + }, + "score_details": { + "score": 0.931, + "details": { + "raw_score": "0.931", + "raw_score_field": "score", + "raw_model_id": "claude-3-5-sonnet-20240620", + "raw_benchmark_id": "big-bench-hard", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-5-sonnet-20240620\",\"https://llm-stats.com/benchmarks/big-bench-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard\"]", + "raw_score_id": "big-bench-hard::claude-3-5-sonnet-20240620", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "drop::drop-claude-3-5-sonnet-20240620", + "evaluation_name": "llm_stats.drop", + "source_data": { + "dataset_name": "DROP", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-5-sonnet-20240620", + "https://llm-stats.com/benchmarks/drop", + "https://api.llm-stats.com/leaderboard/benchmarks/drop" + ], + "additional_details": { + "raw_benchmark_id": "drop", + "raw_model_id": "claude-3-5-sonnet-20240620", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "DROP (Discrete Reasoning Over Paragraphs) is a reading comprehension benchmark requiring discrete reasoning over paragraph content. It contains crowdsourced, adversarially-created questions that require resolving references and performing discrete operations like addition, counting, or sorting, demanding comprehensive paragraph understanding beyond paraphrase-and-entity-typing shortcuts.", + "metric_id": "llm_stats.drop.score", + "metric_name": "DROP score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "drop", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DROP", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "30" + } + }, + "score_details": { + "score": 0.871, + "details": { + "raw_score": "0.871", + "raw_score_field": "score", + "raw_model_id": "claude-3-5-sonnet-20240620", + "raw_benchmark_id": "drop", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-5-sonnet-20240620\",\"https://llm-stats.com/benchmarks/drop\",\"https://api.llm-stats.com/leaderboard/benchmarks/drop\"]", + "raw_score_id": "drop::claude-3-5-sonnet-20240620", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gsm8k::gsm8k-claude-3-5-sonnet-20240620", + "evaluation_name": "llm_stats.gsm8k", + "source_data": { + "dataset_name": "GSM8k", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-5-sonnet-20240620", + "https://llm-stats.com/benchmarks/gsm8k", + "https://api.llm-stats.com/leaderboard/benchmarks/gsm8k" + ], + "additional_details": { + "raw_benchmark_id": "gsm8k", + "raw_model_id": "claude-3-5-sonnet-20240620", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Grade School Math 8K, a dataset of 8.5K high-quality linguistically diverse grade school math word problems requiring multi-step reasoning and elementary arithmetic operations.", + "metric_id": "llm_stats.gsm8k.score", + "metric_name": "GSM8k score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gsm8k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GSM8k", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "48" + } + }, + "score_details": { + "score": 0.964, + "details": { + "raw_score": "0.964", + "raw_score_field": "score", + "raw_model_id": "claude-3-5-sonnet-20240620", + "raw_benchmark_id": "gsm8k", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-5-sonnet-20240620\",\"https://llm-stats.com/benchmarks/gsm8k\",\"https://api.llm-stats.com/leaderboard/benchmarks/gsm8k\"]", + "raw_score_id": "gsm8k::claude-3-5-sonnet-20240620", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "humaneval::humaneval-claude-3-5-sonnet-20240620", + "evaluation_name": "llm_stats.humaneval", + "source_data": { + "dataset_name": "HumanEval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-5-sonnet-20240620", + "https://llm-stats.com/benchmarks/humaneval", + "https://api.llm-stats.com/leaderboard/benchmarks/humaneval" + ], + "additional_details": { + "raw_benchmark_id": "humaneval", + "raw_model_id": "claude-3-5-sonnet-20240620", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A benchmark that measures functional correctness for synthesizing programs from docstrings, consisting of 164 original programming problems assessing language comprehension, algorithms, and simple mathematics", + "metric_id": "llm_stats.humaneval.score", + "metric_name": "HumanEval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "humaneval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HumanEval", + "raw_categories": "[\"reasoning\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "66" + } + }, + "score_details": { + "score": 0.92, + "details": { + "raw_score": "0.92", + "raw_score_field": "score", + "raw_model_id": "claude-3-5-sonnet-20240620", + "raw_benchmark_id": "humaneval", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-5-sonnet-20240620\",\"https://llm-stats.com/benchmarks/humaneval\",\"https://api.llm-stats.com/leaderboard/benchmarks/humaneval\"]", + "raw_score_id": "humaneval::claude-3-5-sonnet-20240620", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mgsm::mgsm-claude-3-5-sonnet-20240620", + "evaluation_name": "llm_stats.mgsm", + "source_data": { + "dataset_name": "MGSM", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-5-sonnet-20240620", + "https://llm-stats.com/benchmarks/mgsm", + "https://api.llm-stats.com/leaderboard/benchmarks/mgsm" + ], + "additional_details": { + "raw_benchmark_id": "mgsm", + "raw_model_id": "claude-3-5-sonnet-20240620", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MGSM (Multilingual Grade School Math) is a benchmark of grade-school math problems. Contains 250 grade-school math problems manually translated from the GSM8K dataset into ten typologically diverse languages: Spanish, French, German, Russian, Chinese, Japanese, Thai, Swahili, Bengali, and Telugu. Evaluates multilingual mathematical reasoning capabilities.", + "metric_id": "llm_stats.mgsm.score", + "metric_name": "MGSM score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mgsm", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MGSM", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "31" + } + }, + "score_details": { + "score": 0.916, + "details": { + "raw_score": "0.916", + "raw_score_field": "score", + "raw_model_id": "claude-3-5-sonnet-20240620", + "raw_benchmark_id": "mgsm", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-5-sonnet-20240620\",\"https://llm-stats.com/benchmarks/mgsm\",\"https://api.llm-stats.com/leaderboard/benchmarks/mgsm\"]", + "raw_score_id": "mgsm::claude-3-5-sonnet-20240620", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu::mmlu-claude-3-5-sonnet-20240620", + "evaluation_name": "llm_stats.mmlu", + "source_data": { + "dataset_name": "MMLU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-5-sonnet-20240620", + "https://llm-stats.com/benchmarks/mmlu", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu" + ], + "additional_details": { + "raw_benchmark_id": "mmlu", + "raw_model_id": "claude-3-5-sonnet-20240620", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Massive Multitask Language Understanding benchmark testing knowledge across 57 diverse subjects including STEM, humanities, social sciences, and professional domains", + "metric_id": "llm_stats.mmlu.score", + "metric_name": "MMLU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU", + "raw_categories": "[\"language\",\"legal\",\"math\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "100" + } + }, + "score_details": { + "score": 0.904, + "details": { + "raw_score": "0.904", + "raw_score_field": "score", + "raw_model_id": "claude-3-5-sonnet-20240620", + "raw_benchmark_id": "mmlu", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-5-sonnet-20240620\",\"https://llm-stats.com/benchmarks/mmlu\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu\"]", + "raw_score_id": "mmlu::claude-3-5-sonnet-20240620", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/anthropic/claude-3-5-sonnet-20241022/f3e26456-131f-4ceb-b69c-e554c8e5f757.json b/data/llm_stats/anthropic/claude-3-5-sonnet-20241022/f3e26456-131f-4ceb-b69c-e554c8e5f757.json new file mode 100644 index 0000000000000000000000000000000000000000..dd0a75d6b62c161c70444c213c3fb66d3e0ad2d8 --- /dev/null +++ b/data/llm_stats/anthropic/claude-3-5-sonnet-20241022/f3e26456-131f-4ceb-b69c-e554c8e5f757.json @@ -0,0 +1,774 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/anthropic_claude-3-5-sonnet-20241022/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Claude 3.5 Sonnet", + "id": "anthropic/claude-3-5-sonnet-20241022", + "developer": "anthropic", + "additional_details": { + "raw_id": "claude-3-5-sonnet-20241022", + "raw_name": "Claude 3.5 Sonnet", + "raw_model_id": "claude-3-5-sonnet-20241022", + "raw_model_name": "Claude 3.5 Sonnet", + "raw_organization_id": "anthropic", + "raw_organization_name": "Anthropic", + "raw_release_date": "2024-10-22", + "raw_announcement_date": "2024-10-22", + "raw_multimodal": "true", + "raw_provider_slug": "anthropic", + "raw_provider_name": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ai2d::ai2d-claude-3-5-sonnet-20241022", + "evaluation_name": "llm_stats.ai2d", + "source_data": { + "dataset_name": "AI2D", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-5-sonnet-20241022", + "https://llm-stats.com/benchmarks/ai2d", + "https://api.llm-stats.com/leaderboard/benchmarks/ai2d" + ], + "additional_details": { + "raw_benchmark_id": "ai2d", + "raw_model_id": "claude-3-5-sonnet-20241022", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "AI2D is a dataset of 4,903 illustrative diagrams from grade school natural sciences (such as food webs, human physiology, and life cycles) with over 15,000 multiple choice questions and answers. The benchmark evaluates diagram understanding and visual reasoning capabilities, requiring models to interpret diagrammatic elements, relationships, and structure to answer questions about scientific concepts represented in visual form.", + "metric_id": "llm_stats.ai2d.score", + "metric_name": "AI2D score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ai2d", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AI2D", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "32" + } + }, + "score_details": { + "score": 0.947, + "details": { + "raw_score": "0.947", + "raw_score_field": "score", + "raw_model_id": "claude-3-5-sonnet-20241022", + "raw_benchmark_id": "ai2d", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-5-sonnet-20241022\",\"https://llm-stats.com/benchmarks/ai2d\",\"https://api.llm-stats.com/leaderboard/benchmarks/ai2d\"]", + "raw_score_id": "ai2d::claude-3-5-sonnet-20241022", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "big-bench-hard::big-bench-hard-claude-3-5-sonnet-20241022", + "evaluation_name": "llm_stats.big-bench-hard", + "source_data": { + "dataset_name": "BIG-Bench Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-5-sonnet-20241022", + "https://llm-stats.com/benchmarks/big-bench-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard" + ], + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_model_id": "claude-3-5-sonnet-20241022", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIG-Bench Hard (BBH) is a subset of 23 challenging BIG-Bench tasks selected because prior language model evaluations did not outperform average human-rater performance. The benchmark contains 6,511 evaluation examples testing various forms of multi-step reasoning including arithmetic, logical reasoning (Boolean expressions, logical deduction), geometric reasoning, temporal reasoning, and language understanding. Tasks require capabilities such as causal judgment, object counting, navigation, pattern recognition, and complex problem solving.", + "metric_id": "llm_stats.big-bench-hard.score", + "metric_name": "BIG-Bench Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench Hard", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "21" + } + }, + "score_details": { + "score": 0.931, + "details": { + "raw_score": "0.931", + "raw_score_field": "score", + "raw_model_id": "claude-3-5-sonnet-20241022", + "raw_benchmark_id": "big-bench-hard", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-5-sonnet-20241022\",\"https://llm-stats.com/benchmarks/big-bench-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard\"]", + "raw_score_id": "big-bench-hard::claude-3-5-sonnet-20241022", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "chartqa::chartqa-claude-3-5-sonnet-20241022", + "evaluation_name": "llm_stats.chartqa", + "source_data": { + "dataset_name": "ChartQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-5-sonnet-20241022", + "https://llm-stats.com/benchmarks/chartqa", + "https://api.llm-stats.com/leaderboard/benchmarks/chartqa" + ], + "additional_details": { + "raw_benchmark_id": "chartqa", + "raw_model_id": "claude-3-5-sonnet-20241022", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ChartQA is a large-scale benchmark comprising 9.6K human-written questions and 23.1K questions generated from human-written chart summaries, designed to evaluate models' abilities in visual and logical reasoning over charts.", + "metric_id": "llm_stats.chartqa.score", + "metric_name": "ChartQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "chartqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ChartQA", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "24" + } + }, + "score_details": { + "score": 0.908, + "details": { + "raw_score": "0.908", + "raw_score_field": "score", + "raw_model_id": "claude-3-5-sonnet-20241022", + "raw_benchmark_id": "chartqa", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-5-sonnet-20241022\",\"https://llm-stats.com/benchmarks/chartqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/chartqa\"]", + "raw_score_id": "chartqa::claude-3-5-sonnet-20241022", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "docvqa::docvqa-claude-3-5-sonnet-20241022", + "evaluation_name": "llm_stats.docvqa", + "source_data": { + "dataset_name": "DocVQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-5-sonnet-20241022", + "https://llm-stats.com/benchmarks/docvqa", + "https://api.llm-stats.com/leaderboard/benchmarks/docvqa" + ], + "additional_details": { + "raw_benchmark_id": "docvqa", + "raw_model_id": "claude-3-5-sonnet-20241022", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A dataset for Visual Question Answering on document images containing 50,000 questions defined on 12,000+ document images. The benchmark tests AI's ability to understand document structure and content, requiring models to comprehend document layout and perform information retrieval to answer questions about document images.", + "metric_id": "llm_stats.docvqa.score", + "metric_name": "DocVQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "docvqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DocVQA", + "raw_categories": "[\"image_to_text\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "26" + } + }, + "score_details": { + "score": 0.952, + "details": { + "raw_score": "0.952", + "raw_score_field": "score", + "raw_model_id": "claude-3-5-sonnet-20241022", + "raw_benchmark_id": "docvqa", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-5-sonnet-20241022\",\"https://llm-stats.com/benchmarks/docvqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/docvqa\"]", + "raw_score_id": "docvqa::claude-3-5-sonnet-20241022", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "drop::drop-claude-3-5-sonnet-20241022", + "evaluation_name": "llm_stats.drop", + "source_data": { + "dataset_name": "DROP", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-5-sonnet-20241022", + "https://llm-stats.com/benchmarks/drop", + "https://api.llm-stats.com/leaderboard/benchmarks/drop" + ], + "additional_details": { + "raw_benchmark_id": "drop", + "raw_model_id": "claude-3-5-sonnet-20241022", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "DROP (Discrete Reasoning Over Paragraphs) is a reading comprehension benchmark requiring discrete reasoning over paragraph content. It contains crowdsourced, adversarially-created questions that require resolving references and performing discrete operations like addition, counting, or sorting, demanding comprehensive paragraph understanding beyond paraphrase-and-entity-typing shortcuts.", + "metric_id": "llm_stats.drop.score", + "metric_name": "DROP score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "drop", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DROP", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "30" + } + }, + "score_details": { + "score": 0.871, + "details": { + "raw_score": "0.871", + "raw_score_field": "score", + "raw_model_id": "claude-3-5-sonnet-20241022", + "raw_benchmark_id": "drop", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-5-sonnet-20241022\",\"https://llm-stats.com/benchmarks/drop\",\"https://api.llm-stats.com/leaderboard/benchmarks/drop\"]", + "raw_score_id": "drop::claude-3-5-sonnet-20241022", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gsm8k::gsm8k-claude-3-5-sonnet-20241022", + "evaluation_name": "llm_stats.gsm8k", + "source_data": { + "dataset_name": "GSM8k", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-5-sonnet-20241022", + "https://llm-stats.com/benchmarks/gsm8k", + "https://api.llm-stats.com/leaderboard/benchmarks/gsm8k" + ], + "additional_details": { + "raw_benchmark_id": "gsm8k", + "raw_model_id": "claude-3-5-sonnet-20241022", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Grade School Math 8K, a dataset of 8.5K high-quality linguistically diverse grade school math word problems requiring multi-step reasoning and elementary arithmetic operations.", + "metric_id": "llm_stats.gsm8k.score", + "metric_name": "GSM8k score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gsm8k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GSM8k", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "48" + } + }, + "score_details": { + "score": 0.964, + "details": { + "raw_score": "0.964", + "raw_score_field": "score", + "raw_model_id": "claude-3-5-sonnet-20241022", + "raw_benchmark_id": "gsm8k", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-5-sonnet-20241022\",\"https://llm-stats.com/benchmarks/gsm8k\",\"https://api.llm-stats.com/leaderboard/benchmarks/gsm8k\"]", + "raw_score_id": "gsm8k::claude-3-5-sonnet-20241022", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "humaneval::humaneval-claude-3-5-sonnet-20241022", + "evaluation_name": "llm_stats.humaneval", + "source_data": { + "dataset_name": "HumanEval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-5-sonnet-20241022", + "https://llm-stats.com/benchmarks/humaneval", + "https://api.llm-stats.com/leaderboard/benchmarks/humaneval" + ], + "additional_details": { + "raw_benchmark_id": "humaneval", + "raw_model_id": "claude-3-5-sonnet-20241022", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A benchmark that measures functional correctness for synthesizing programs from docstrings, consisting of 164 original programming problems assessing language comprehension, algorithms, and simple mathematics", + "metric_id": "llm_stats.humaneval.score", + "metric_name": "HumanEval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "humaneval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HumanEval", + "raw_categories": "[\"reasoning\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "66" + } + }, + "score_details": { + "score": 0.937, + "details": { + "raw_score": "0.937", + "raw_score_field": "score", + "raw_model_id": "claude-3-5-sonnet-20241022", + "raw_benchmark_id": "humaneval", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-5-sonnet-20241022\",\"https://llm-stats.com/benchmarks/humaneval\",\"https://api.llm-stats.com/leaderboard/benchmarks/humaneval\"]", + "raw_score_id": "humaneval::claude-3-5-sonnet-20241022", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mathvista::mathvista-claude-3-5-sonnet-20241022", + "evaluation_name": "llm_stats.mathvista", + "source_data": { + "dataset_name": "MathVista", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-5-sonnet-20241022", + "https://llm-stats.com/benchmarks/mathvista", + "https://api.llm-stats.com/leaderboard/benchmarks/mathvista" + ], + "additional_details": { + "raw_benchmark_id": "mathvista", + "raw_model_id": "claude-3-5-sonnet-20241022", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MathVista evaluates mathematical reasoning of foundation models in visual contexts. It consists of 6,141 examples derived from 28 existing multimodal datasets and 3 newly created datasets (IQTest, FunctionQA, and PaperQA), combining challenges from diverse mathematical and visual tasks to assess models' ability to understand complex figures and perform rigorous reasoning.", + "metric_id": "llm_stats.mathvista.score", + "metric_name": "MathVista score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mathvista", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MathVista", + "raw_categories": "[\"math\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "39" + } + }, + "score_details": { + "score": 0.677, + "details": { + "raw_score": "0.677", + "raw_score_field": "score", + "raw_model_id": "claude-3-5-sonnet-20241022", + "raw_benchmark_id": "mathvista", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-5-sonnet-20241022\",\"https://llm-stats.com/benchmarks/mathvista\",\"https://api.llm-stats.com/leaderboard/benchmarks/mathvista\"]", + "raw_score_id": "mathvista::claude-3-5-sonnet-20241022", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mgsm::mgsm-claude-3-5-sonnet-20241022", + "evaluation_name": "llm_stats.mgsm", + "source_data": { + "dataset_name": "MGSM", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-5-sonnet-20241022", + "https://llm-stats.com/benchmarks/mgsm", + "https://api.llm-stats.com/leaderboard/benchmarks/mgsm" + ], + "additional_details": { + "raw_benchmark_id": "mgsm", + "raw_model_id": "claude-3-5-sonnet-20241022", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MGSM (Multilingual Grade School Math) is a benchmark of grade-school math problems. Contains 250 grade-school math problems manually translated from the GSM8K dataset into ten typologically diverse languages: Spanish, French, German, Russian, Chinese, Japanese, Thai, Swahili, Bengali, and Telugu. Evaluates multilingual mathematical reasoning capabilities.", + "metric_id": "llm_stats.mgsm.score", + "metric_name": "MGSM score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mgsm", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MGSM", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "31" + } + }, + "score_details": { + "score": 0.916, + "details": { + "raw_score": "0.916", + "raw_score_field": "score", + "raw_model_id": "claude-3-5-sonnet-20241022", + "raw_benchmark_id": "mgsm", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-5-sonnet-20241022\",\"https://llm-stats.com/benchmarks/mgsm\",\"https://api.llm-stats.com/leaderboard/benchmarks/mgsm\"]", + "raw_score_id": "mgsm::claude-3-5-sonnet-20241022", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu::mmlu-claude-3-5-sonnet-20241022", + "evaluation_name": "llm_stats.mmlu", + "source_data": { + "dataset_name": "MMLU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-5-sonnet-20241022", + "https://llm-stats.com/benchmarks/mmlu", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu" + ], + "additional_details": { + "raw_benchmark_id": "mmlu", + "raw_model_id": "claude-3-5-sonnet-20241022", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Massive Multitask Language Understanding benchmark testing knowledge across 57 diverse subjects including STEM, humanities, social sciences, and professional domains", + "metric_id": "llm_stats.mmlu.score", + "metric_name": "MMLU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU", + "raw_categories": "[\"language\",\"legal\",\"math\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "100" + } + }, + "score_details": { + "score": 0.904, + "details": { + "raw_score": "0.904", + "raw_score_field": "score", + "raw_model_id": "claude-3-5-sonnet-20241022", + "raw_benchmark_id": "mmlu", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-5-sonnet-20241022\",\"https://llm-stats.com/benchmarks/mmlu\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu\"]", + "raw_score_id": "mmlu::claude-3-5-sonnet-20241022", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "osworld-extended::osworld-extended-claude-3-5-sonnet-20241022", + "evaluation_name": "llm_stats.osworld-extended", + "source_data": { + "dataset_name": "OSWorld Extended", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-5-sonnet-20241022", + "https://llm-stats.com/benchmarks/osworld-extended", + "https://api.llm-stats.com/leaderboard/benchmarks/osworld-extended" + ], + "additional_details": { + "raw_benchmark_id": "osworld-extended", + "raw_model_id": "claude-3-5-sonnet-20241022", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OSWorld is a scalable, real computer environment benchmark for evaluating multimodal agents on open-ended tasks across Ubuntu, Windows, and macOS. It comprises 369 computer tasks involving real web and desktop applications, OS file I/O, and multi-application workflows. The benchmark evaluates agents' ability to interact with computer interfaces using screenshots and actions in realistic computing environments.", + "metric_id": "llm_stats.osworld-extended.score", + "metric_name": "OSWorld Extended score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "osworld-extended", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OSWorld Extended", + "raw_categories": "[\"multimodal\",\"reasoning\",\"general\",\"agents\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.22, + "details": { + "raw_score": "0.22", + "raw_score_field": "score", + "raw_model_id": "claude-3-5-sonnet-20241022", + "raw_benchmark_id": "osworld-extended", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-5-sonnet-20241022\",\"https://llm-stats.com/benchmarks/osworld-extended\",\"https://api.llm-stats.com/leaderboard/benchmarks/osworld-extended\"]", + "raw_score_id": "osworld-extended::claude-3-5-sonnet-20241022", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "osworld-screenshot-only::osworld-screenshot-only-claude-3-5-sonnet-20241022", + "evaluation_name": "llm_stats.osworld-screenshot-only", + "source_data": { + "dataset_name": "OSWorld Screenshot-only", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-5-sonnet-20241022", + "https://llm-stats.com/benchmarks/osworld-screenshot-only", + "https://api.llm-stats.com/leaderboard/benchmarks/osworld-screenshot-only" + ], + "additional_details": { + "raw_benchmark_id": "osworld-screenshot-only", + "raw_model_id": "claude-3-5-sonnet-20241022", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OSWorld Screenshot-only: A variant of the OSWorld benchmark that evaluates multimodal AI agents using only screenshot observations to complete open-ended computer tasks across real operating systems (Ubuntu, Windows, macOS). Tests agents' ability to perform complex workflows involving web apps, desktop applications, file I/O, and multi-application tasks through visual interface understanding and GUI grounding.", + "metric_id": "llm_stats.osworld-screenshot-only.score", + "metric_name": "OSWorld Screenshot-only score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "osworld-screenshot-only", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OSWorld Screenshot-only", + "raw_categories": "[\"multimodal\",\"general\",\"grounding\",\"agents\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.149, + "details": { + "raw_score": "0.149", + "raw_score_field": "score", + "raw_model_id": "claude-3-5-sonnet-20241022", + "raw_benchmark_id": "osworld-screenshot-only", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-5-sonnet-20241022\",\"https://llm-stats.com/benchmarks/osworld-screenshot-only\",\"https://api.llm-stats.com/leaderboard/benchmarks/osworld-screenshot-only\"]", + "raw_score_id": "osworld-screenshot-only::claude-3-5-sonnet-20241022", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "tau-bench-airline::tau-bench-airline-claude-3-5-sonnet-20241022", + "evaluation_name": "llm_stats.tau-bench-airline", + "source_data": { + "dataset_name": "TAU-bench Airline", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-5-sonnet-20241022", + "https://llm-stats.com/benchmarks/tau-bench-airline", + "https://api.llm-stats.com/leaderboard/benchmarks/tau-bench-airline" + ], + "additional_details": { + "raw_benchmark_id": "tau-bench-airline", + "raw_model_id": "claude-3-5-sonnet-20241022", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Part of τ-bench (TAU-bench), a benchmark for Tool-Agent-User interaction in real-world domains. The airline domain evaluates language agents' ability to interact with users through dynamic conversations while following domain-specific rules and using API tools. Agents must handle airline-related tasks and policies reliably.", + "metric_id": "llm_stats.tau-bench-airline.score", + "metric_name": "TAU-bench Airline score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "tau-bench-airline", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TAU-bench Airline", + "raw_categories": "[\"reasoning\",\"communication\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.46, + "details": { + "raw_score": "0.46", + "raw_score_field": "score", + "raw_model_id": "claude-3-5-sonnet-20241022", + "raw_benchmark_id": "tau-bench-airline", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-5-sonnet-20241022\",\"https://llm-stats.com/benchmarks/tau-bench-airline\",\"https://api.llm-stats.com/leaderboard/benchmarks/tau-bench-airline\"]", + "raw_score_id": "tau-bench-airline::claude-3-5-sonnet-20241022", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "tau-bench-retail::tau-bench-retail-claude-3-5-sonnet-20241022", + "evaluation_name": "llm_stats.tau-bench-retail", + "source_data": { + "dataset_name": "TAU-bench Retail", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-5-sonnet-20241022", + "https://llm-stats.com/benchmarks/tau-bench-retail", + "https://api.llm-stats.com/leaderboard/benchmarks/tau-bench-retail" + ], + "additional_details": { + "raw_benchmark_id": "tau-bench-retail", + "raw_model_id": "claude-3-5-sonnet-20241022", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A benchmark for evaluating tool-agent-user interaction in retail environments. Tests language agents' ability to handle dynamic conversations with users while using domain-specific API tools and following policy guidelines. Evaluates agents on tasks like order cancellations, address changes, and order status checks through multi-turn conversations.", + "metric_id": "llm_stats.tau-bench-retail.score", + "metric_name": "TAU-bench Retail score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "tau-bench-retail", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TAU-bench Retail", + "raw_categories": "[\"reasoning\",\"communication\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "25" + } + }, + "score_details": { + "score": 0.692, + "details": { + "raw_score": "0.692", + "raw_score_field": "score", + "raw_model_id": "claude-3-5-sonnet-20241022", + "raw_benchmark_id": "tau-bench-retail", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-5-sonnet-20241022\",\"https://llm-stats.com/benchmarks/tau-bench-retail\",\"https://api.llm-stats.com/leaderboard/benchmarks/tau-bench-retail\"]", + "raw_score_id": "tau-bench-retail::claude-3-5-sonnet-20241022", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/anthropic/claude-3-haiku-20240307/844dcc46-04ca-4182-891e-3817757c2ed5.json b/data/llm_stats/anthropic/claude-3-haiku-20240307/844dcc46-04ca-4182-891e-3817757c2ed5.json new file mode 100644 index 0000000000000000000000000000000000000000..40951426f36e25d098eb396b7e28210a199149d3 --- /dev/null +++ b/data/llm_stats/anthropic/claude-3-haiku-20240307/844dcc46-04ca-4182-891e-3817757c2ed5.json @@ -0,0 +1,306 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/anthropic_claude-3-haiku-20240307/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Claude 3 Haiku", + "id": "anthropic/claude-3-haiku-20240307", + "developer": "anthropic", + "additional_details": { + "raw_id": "claude-3-haiku-20240307", + "raw_name": "Claude 3 Haiku", + "raw_model_id": "claude-3-haiku-20240307", + "raw_model_name": "Claude 3 Haiku", + "raw_organization_id": "anthropic", + "raw_organization_name": "Anthropic", + "raw_release_date": "2024-03-13", + "raw_announcement_date": "2024-03-13", + "raw_multimodal": "true", + "raw_provider_slug": "anthropic", + "raw_provider_name": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "arc-c::arc-c-claude-3-haiku-20240307", + "evaluation_name": "llm_stats.arc-c", + "source_data": { + "dataset_name": "ARC-C", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-haiku-20240307", + "https://llm-stats.com/benchmarks/arc-c", + "https://api.llm-stats.com/leaderboard/benchmarks/arc-c" + ], + "additional_details": { + "raw_benchmark_id": "arc-c", + "raw_model_id": "claude-3-haiku-20240307", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The AI2 Reasoning Challenge (ARC) Challenge Set is a multiple-choice question-answering benchmark containing grade-school level science questions that require advanced reasoning capabilities. ARC-C specifically contains questions that were answered incorrectly by both retrieval-based and word co-occurrence algorithms, making it a particularly challenging subset designed to test commonsense reasoning abilities in AI systems.", + "metric_id": "llm_stats.arc-c.score", + "metric_name": "ARC-C score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arc-c", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ARC-C", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "34" + } + }, + "score_details": { + "score": 0.892, + "details": { + "raw_score": "0.892", + "raw_score_field": "score", + "raw_model_id": "claude-3-haiku-20240307", + "raw_benchmark_id": "arc-c", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-haiku-20240307\",\"https://llm-stats.com/benchmarks/arc-c\",\"https://api.llm-stats.com/leaderboard/benchmarks/arc-c\"]", + "raw_score_id": "arc-c::claude-3-haiku-20240307", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "big-bench-hard::big-bench-hard-claude-3-haiku-20240307", + "evaluation_name": "llm_stats.big-bench-hard", + "source_data": { + "dataset_name": "BIG-Bench Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-haiku-20240307", + "https://llm-stats.com/benchmarks/big-bench-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard" + ], + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_model_id": "claude-3-haiku-20240307", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIG-Bench Hard (BBH) is a subset of 23 challenging BIG-Bench tasks selected because prior language model evaluations did not outperform average human-rater performance. The benchmark contains 6,511 evaluation examples testing various forms of multi-step reasoning including arithmetic, logical reasoning (Boolean expressions, logical deduction), geometric reasoning, temporal reasoning, and language understanding. Tasks require capabilities such as causal judgment, object counting, navigation, pattern recognition, and complex problem solving.", + "metric_id": "llm_stats.big-bench-hard.score", + "metric_name": "BIG-Bench Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench Hard", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "21" + } + }, + "score_details": { + "score": 0.737, + "details": { + "raw_score": "0.737", + "raw_score_field": "score", + "raw_model_id": "claude-3-haiku-20240307", + "raw_benchmark_id": "big-bench-hard", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-haiku-20240307\",\"https://llm-stats.com/benchmarks/big-bench-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard\"]", + "raw_score_id": "big-bench-hard::claude-3-haiku-20240307", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "drop::drop-claude-3-haiku-20240307", + "evaluation_name": "llm_stats.drop", + "source_data": { + "dataset_name": "DROP", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-haiku-20240307", + "https://llm-stats.com/benchmarks/drop", + "https://api.llm-stats.com/leaderboard/benchmarks/drop" + ], + "additional_details": { + "raw_benchmark_id": "drop", + "raw_model_id": "claude-3-haiku-20240307", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "DROP (Discrete Reasoning Over Paragraphs) is a reading comprehension benchmark requiring discrete reasoning over paragraph content. It contains crowdsourced, adversarially-created questions that require resolving references and performing discrete operations like addition, counting, or sorting, demanding comprehensive paragraph understanding beyond paraphrase-and-entity-typing shortcuts.", + "metric_id": "llm_stats.drop.score", + "metric_name": "DROP score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "drop", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DROP", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "30" + } + }, + "score_details": { + "score": 0.784, + "details": { + "raw_score": "0.784", + "raw_score_field": "score", + "raw_model_id": "claude-3-haiku-20240307", + "raw_benchmark_id": "drop", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-haiku-20240307\",\"https://llm-stats.com/benchmarks/drop\",\"https://api.llm-stats.com/leaderboard/benchmarks/drop\"]", + "raw_score_id": "drop::claude-3-haiku-20240307", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hellaswag::hellaswag-claude-3-haiku-20240307", + "evaluation_name": "llm_stats.hellaswag", + "source_data": { + "dataset_name": "HellaSwag", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-haiku-20240307", + "https://llm-stats.com/benchmarks/hellaswag", + "https://api.llm-stats.com/leaderboard/benchmarks/hellaswag" + ], + "additional_details": { + "raw_benchmark_id": "hellaswag", + "raw_model_id": "claude-3-haiku-20240307", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A challenging commonsense natural language inference dataset that uses Adversarial Filtering to create questions trivial for humans (>95% accuracy) but difficult for state-of-the-art models, requiring completion of sentence endings based on physical situations and everyday activities", + "metric_id": "llm_stats.hellaswag.score", + "metric_name": "HellaSwag score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hellaswag", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HellaSwag", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.859, + "details": { + "raw_score": "0.859", + "raw_score_field": "score", + "raw_model_id": "claude-3-haiku-20240307", + "raw_benchmark_id": "hellaswag", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-haiku-20240307\",\"https://llm-stats.com/benchmarks/hellaswag\",\"https://api.llm-stats.com/leaderboard/benchmarks/hellaswag\"]", + "raw_score_id": "hellaswag::claude-3-haiku-20240307", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mgsm::mgsm-claude-3-haiku-20240307", + "evaluation_name": "llm_stats.mgsm", + "source_data": { + "dataset_name": "MGSM", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-haiku-20240307", + "https://llm-stats.com/benchmarks/mgsm", + "https://api.llm-stats.com/leaderboard/benchmarks/mgsm" + ], + "additional_details": { + "raw_benchmark_id": "mgsm", + "raw_model_id": "claude-3-haiku-20240307", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MGSM (Multilingual Grade School Math) is a benchmark of grade-school math problems. Contains 250 grade-school math problems manually translated from the GSM8K dataset into ten typologically diverse languages: Spanish, French, German, Russian, Chinese, Japanese, Thai, Swahili, Bengali, and Telugu. Evaluates multilingual mathematical reasoning capabilities.", + "metric_id": "llm_stats.mgsm.score", + "metric_name": "MGSM score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mgsm", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MGSM", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "31" + } + }, + "score_details": { + "score": 0.751, + "details": { + "raw_score": "0.751", + "raw_score_field": "score", + "raw_model_id": "claude-3-haiku-20240307", + "raw_benchmark_id": "mgsm", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-haiku-20240307\",\"https://llm-stats.com/benchmarks/mgsm\",\"https://api.llm-stats.com/leaderboard/benchmarks/mgsm\"]", + "raw_score_id": "mgsm::claude-3-haiku-20240307", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/anthropic/claude-3-opus-20240229/40448880-c8f0-4132-8bfe-f7261d89bc10.json b/data/llm_stats/anthropic/claude-3-opus-20240229/40448880-c8f0-4132-8bfe-f7261d89bc10.json new file mode 100644 index 0000000000000000000000000000000000000000..ffb721d4dab616f98a6d2e18d415ce1bf83a4d2f --- /dev/null +++ b/data/llm_stats/anthropic/claude-3-opus-20240229/40448880-c8f0-4132-8bfe-f7261d89bc10.json @@ -0,0 +1,358 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/anthropic_claude-3-opus-20240229/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Claude 3 Opus", + "id": "anthropic/claude-3-opus-20240229", + "developer": "anthropic", + "additional_details": { + "raw_id": "claude-3-opus-20240229", + "raw_name": "Claude 3 Opus", + "raw_model_id": "claude-3-opus-20240229", + "raw_model_name": "Claude 3 Opus", + "raw_organization_id": "anthropic", + "raw_organization_name": "Anthropic", + "raw_release_date": "2024-02-29", + "raw_announcement_date": "2024-02-29", + "raw_multimodal": "true", + "raw_provider_slug": "anthropic", + "raw_provider_name": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "arc-c::arc-c-claude-3-opus-20240229", + "evaluation_name": "llm_stats.arc-c", + "source_data": { + "dataset_name": "ARC-C", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-opus-20240229", + "https://llm-stats.com/benchmarks/arc-c", + "https://api.llm-stats.com/leaderboard/benchmarks/arc-c" + ], + "additional_details": { + "raw_benchmark_id": "arc-c", + "raw_model_id": "claude-3-opus-20240229", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The AI2 Reasoning Challenge (ARC) Challenge Set is a multiple-choice question-answering benchmark containing grade-school level science questions that require advanced reasoning capabilities. ARC-C specifically contains questions that were answered incorrectly by both retrieval-based and word co-occurrence algorithms, making it a particularly challenging subset designed to test commonsense reasoning abilities in AI systems.", + "metric_id": "llm_stats.arc-c.score", + "metric_name": "ARC-C score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arc-c", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ARC-C", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "34" + } + }, + "score_details": { + "score": 0.964, + "details": { + "raw_score": "0.964", + "raw_score_field": "score", + "raw_model_id": "claude-3-opus-20240229", + "raw_benchmark_id": "arc-c", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-opus-20240229\",\"https://llm-stats.com/benchmarks/arc-c\",\"https://api.llm-stats.com/leaderboard/benchmarks/arc-c\"]", + "raw_score_id": "arc-c::claude-3-opus-20240229", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "big-bench-hard::big-bench-hard-claude-3-opus-20240229", + "evaluation_name": "llm_stats.big-bench-hard", + "source_data": { + "dataset_name": "BIG-Bench Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-opus-20240229", + "https://llm-stats.com/benchmarks/big-bench-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard" + ], + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_model_id": "claude-3-opus-20240229", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIG-Bench Hard (BBH) is a subset of 23 challenging BIG-Bench tasks selected because prior language model evaluations did not outperform average human-rater performance. The benchmark contains 6,511 evaluation examples testing various forms of multi-step reasoning including arithmetic, logical reasoning (Boolean expressions, logical deduction), geometric reasoning, temporal reasoning, and language understanding. Tasks require capabilities such as causal judgment, object counting, navigation, pattern recognition, and complex problem solving.", + "metric_id": "llm_stats.big-bench-hard.score", + "metric_name": "BIG-Bench Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench Hard", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "21" + } + }, + "score_details": { + "score": 0.868, + "details": { + "raw_score": "0.868", + "raw_score_field": "score", + "raw_model_id": "claude-3-opus-20240229", + "raw_benchmark_id": "big-bench-hard", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-opus-20240229\",\"https://llm-stats.com/benchmarks/big-bench-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard\"]", + "raw_score_id": "big-bench-hard::claude-3-opus-20240229", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "drop::drop-claude-3-opus-20240229", + "evaluation_name": "llm_stats.drop", + "source_data": { + "dataset_name": "DROP", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-opus-20240229", + "https://llm-stats.com/benchmarks/drop", + "https://api.llm-stats.com/leaderboard/benchmarks/drop" + ], + "additional_details": { + "raw_benchmark_id": "drop", + "raw_model_id": "claude-3-opus-20240229", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "DROP (Discrete Reasoning Over Paragraphs) is a reading comprehension benchmark requiring discrete reasoning over paragraph content. It contains crowdsourced, adversarially-created questions that require resolving references and performing discrete operations like addition, counting, or sorting, demanding comprehensive paragraph understanding beyond paraphrase-and-entity-typing shortcuts.", + "metric_id": "llm_stats.drop.score", + "metric_name": "DROP score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "drop", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DROP", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "30" + } + }, + "score_details": { + "score": 0.831, + "details": { + "raw_score": "0.831", + "raw_score_field": "score", + "raw_model_id": "claude-3-opus-20240229", + "raw_benchmark_id": "drop", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-opus-20240229\",\"https://llm-stats.com/benchmarks/drop\",\"https://api.llm-stats.com/leaderboard/benchmarks/drop\"]", + "raw_score_id": "drop::claude-3-opus-20240229", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gsm8k::gsm8k-claude-3-opus-20240229", + "evaluation_name": "llm_stats.gsm8k", + "source_data": { + "dataset_name": "GSM8k", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-opus-20240229", + "https://llm-stats.com/benchmarks/gsm8k", + "https://api.llm-stats.com/leaderboard/benchmarks/gsm8k" + ], + "additional_details": { + "raw_benchmark_id": "gsm8k", + "raw_model_id": "claude-3-opus-20240229", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Grade School Math 8K, a dataset of 8.5K high-quality linguistically diverse grade school math word problems requiring multi-step reasoning and elementary arithmetic operations.", + "metric_id": "llm_stats.gsm8k.score", + "metric_name": "GSM8k score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gsm8k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GSM8k", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "48" + } + }, + "score_details": { + "score": 0.95, + "details": { + "raw_score": "0.95", + "raw_score_field": "score", + "raw_model_id": "claude-3-opus-20240229", + "raw_benchmark_id": "gsm8k", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-opus-20240229\",\"https://llm-stats.com/benchmarks/gsm8k\",\"https://api.llm-stats.com/leaderboard/benchmarks/gsm8k\"]", + "raw_score_id": "gsm8k::claude-3-opus-20240229", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hellaswag::hellaswag-claude-3-opus-20240229", + "evaluation_name": "llm_stats.hellaswag", + "source_data": { + "dataset_name": "HellaSwag", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-opus-20240229", + "https://llm-stats.com/benchmarks/hellaswag", + "https://api.llm-stats.com/leaderboard/benchmarks/hellaswag" + ], + "additional_details": { + "raw_benchmark_id": "hellaswag", + "raw_model_id": "claude-3-opus-20240229", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A challenging commonsense natural language inference dataset that uses Adversarial Filtering to create questions trivial for humans (>95% accuracy) but difficult for state-of-the-art models, requiring completion of sentence endings based on physical situations and everyday activities", + "metric_id": "llm_stats.hellaswag.score", + "metric_name": "HellaSwag score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hellaswag", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HellaSwag", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.954, + "details": { + "raw_score": "0.954", + "raw_score_field": "score", + "raw_model_id": "claude-3-opus-20240229", + "raw_benchmark_id": "hellaswag", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-opus-20240229\",\"https://llm-stats.com/benchmarks/hellaswag\",\"https://api.llm-stats.com/leaderboard/benchmarks/hellaswag\"]", + "raw_score_id": "hellaswag::claude-3-opus-20240229", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mgsm::mgsm-claude-3-opus-20240229", + "evaluation_name": "llm_stats.mgsm", + "source_data": { + "dataset_name": "MGSM", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-opus-20240229", + "https://llm-stats.com/benchmarks/mgsm", + "https://api.llm-stats.com/leaderboard/benchmarks/mgsm" + ], + "additional_details": { + "raw_benchmark_id": "mgsm", + "raw_model_id": "claude-3-opus-20240229", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MGSM (Multilingual Grade School Math) is a benchmark of grade-school math problems. Contains 250 grade-school math problems manually translated from the GSM8K dataset into ten typologically diverse languages: Spanish, French, German, Russian, Chinese, Japanese, Thai, Swahili, Bengali, and Telugu. Evaluates multilingual mathematical reasoning capabilities.", + "metric_id": "llm_stats.mgsm.score", + "metric_name": "MGSM score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mgsm", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MGSM", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "31" + } + }, + "score_details": { + "score": 0.907, + "details": { + "raw_score": "0.907", + "raw_score_field": "score", + "raw_model_id": "claude-3-opus-20240229", + "raw_benchmark_id": "mgsm", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-opus-20240229\",\"https://llm-stats.com/benchmarks/mgsm\",\"https://api.llm-stats.com/leaderboard/benchmarks/mgsm\"]", + "raw_score_id": "mgsm::claude-3-opus-20240229", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/anthropic/claude-3-sonnet-20240229/573dedd8-3809-4ddf-a186-cd6312900482.json b/data/llm_stats/anthropic/claude-3-sonnet-20240229/573dedd8-3809-4ddf-a186-cd6312900482.json new file mode 100644 index 0000000000000000000000000000000000000000..3f47c4ee21310987054393d951ce14ea8eb6d13d --- /dev/null +++ b/data/llm_stats/anthropic/claude-3-sonnet-20240229/573dedd8-3809-4ddf-a186-cd6312900482.json @@ -0,0 +1,358 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/anthropic_claude-3-sonnet-20240229/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Claude 3 Sonnet", + "id": "anthropic/claude-3-sonnet-20240229", + "developer": "anthropic", + "additional_details": { + "raw_id": "claude-3-sonnet-20240229", + "raw_name": "Claude 3 Sonnet", + "raw_model_id": "claude-3-sonnet-20240229", + "raw_model_name": "Claude 3 Sonnet", + "raw_organization_id": "anthropic", + "raw_organization_name": "Anthropic", + "raw_release_date": "2024-02-29", + "raw_announcement_date": "2024-02-29", + "raw_multimodal": "true", + "raw_provider_slug": "anthropic", + "raw_provider_name": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "arc-c::arc-c-claude-3-sonnet-20240229", + "evaluation_name": "llm_stats.arc-c", + "source_data": { + "dataset_name": "ARC-C", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-sonnet-20240229", + "https://llm-stats.com/benchmarks/arc-c", + "https://api.llm-stats.com/leaderboard/benchmarks/arc-c" + ], + "additional_details": { + "raw_benchmark_id": "arc-c", + "raw_model_id": "claude-3-sonnet-20240229", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The AI2 Reasoning Challenge (ARC) Challenge Set is a multiple-choice question-answering benchmark containing grade-school level science questions that require advanced reasoning capabilities. ARC-C specifically contains questions that were answered incorrectly by both retrieval-based and word co-occurrence algorithms, making it a particularly challenging subset designed to test commonsense reasoning abilities in AI systems.", + "metric_id": "llm_stats.arc-c.score", + "metric_name": "ARC-C score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arc-c", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ARC-C", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "34" + } + }, + "score_details": { + "score": 0.932, + "details": { + "raw_score": "0.932", + "raw_score_field": "score", + "raw_model_id": "claude-3-sonnet-20240229", + "raw_benchmark_id": "arc-c", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-sonnet-20240229\",\"https://llm-stats.com/benchmarks/arc-c\",\"https://api.llm-stats.com/leaderboard/benchmarks/arc-c\"]", + "raw_score_id": "arc-c::claude-3-sonnet-20240229", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "big-bench-hard::big-bench-hard-claude-3-sonnet-20240229", + "evaluation_name": "llm_stats.big-bench-hard", + "source_data": { + "dataset_name": "BIG-Bench Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-sonnet-20240229", + "https://llm-stats.com/benchmarks/big-bench-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard" + ], + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_model_id": "claude-3-sonnet-20240229", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIG-Bench Hard (BBH) is a subset of 23 challenging BIG-Bench tasks selected because prior language model evaluations did not outperform average human-rater performance. The benchmark contains 6,511 evaluation examples testing various forms of multi-step reasoning including arithmetic, logical reasoning (Boolean expressions, logical deduction), geometric reasoning, temporal reasoning, and language understanding. Tasks require capabilities such as causal judgment, object counting, navigation, pattern recognition, and complex problem solving.", + "metric_id": "llm_stats.big-bench-hard.score", + "metric_name": "BIG-Bench Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench Hard", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "21" + } + }, + "score_details": { + "score": 0.829, + "details": { + "raw_score": "0.829", + "raw_score_field": "score", + "raw_model_id": "claude-3-sonnet-20240229", + "raw_benchmark_id": "big-bench-hard", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-sonnet-20240229\",\"https://llm-stats.com/benchmarks/big-bench-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard\"]", + "raw_score_id": "big-bench-hard::claude-3-sonnet-20240229", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "drop::drop-claude-3-sonnet-20240229", + "evaluation_name": "llm_stats.drop", + "source_data": { + "dataset_name": "DROP", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-sonnet-20240229", + "https://llm-stats.com/benchmarks/drop", + "https://api.llm-stats.com/leaderboard/benchmarks/drop" + ], + "additional_details": { + "raw_benchmark_id": "drop", + "raw_model_id": "claude-3-sonnet-20240229", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "DROP (Discrete Reasoning Over Paragraphs) is a reading comprehension benchmark requiring discrete reasoning over paragraph content. It contains crowdsourced, adversarially-created questions that require resolving references and performing discrete operations like addition, counting, or sorting, demanding comprehensive paragraph understanding beyond paraphrase-and-entity-typing shortcuts.", + "metric_id": "llm_stats.drop.score", + "metric_name": "DROP score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "drop", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DROP", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "30" + } + }, + "score_details": { + "score": 0.789, + "details": { + "raw_score": "0.789", + "raw_score_field": "score", + "raw_model_id": "claude-3-sonnet-20240229", + "raw_benchmark_id": "drop", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-sonnet-20240229\",\"https://llm-stats.com/benchmarks/drop\",\"https://api.llm-stats.com/leaderboard/benchmarks/drop\"]", + "raw_score_id": "drop::claude-3-sonnet-20240229", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gsm8k::gsm8k-claude-3-sonnet-20240229", + "evaluation_name": "llm_stats.gsm8k", + "source_data": { + "dataset_name": "GSM8k", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-sonnet-20240229", + "https://llm-stats.com/benchmarks/gsm8k", + "https://api.llm-stats.com/leaderboard/benchmarks/gsm8k" + ], + "additional_details": { + "raw_benchmark_id": "gsm8k", + "raw_model_id": "claude-3-sonnet-20240229", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Grade School Math 8K, a dataset of 8.5K high-quality linguistically diverse grade school math word problems requiring multi-step reasoning and elementary arithmetic operations.", + "metric_id": "llm_stats.gsm8k.score", + "metric_name": "GSM8k score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gsm8k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GSM8k", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "48" + } + }, + "score_details": { + "score": 0.923, + "details": { + "raw_score": "0.923", + "raw_score_field": "score", + "raw_model_id": "claude-3-sonnet-20240229", + "raw_benchmark_id": "gsm8k", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-sonnet-20240229\",\"https://llm-stats.com/benchmarks/gsm8k\",\"https://api.llm-stats.com/leaderboard/benchmarks/gsm8k\"]", + "raw_score_id": "gsm8k::claude-3-sonnet-20240229", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hellaswag::hellaswag-claude-3-sonnet-20240229", + "evaluation_name": "llm_stats.hellaswag", + "source_data": { + "dataset_name": "HellaSwag", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-sonnet-20240229", + "https://llm-stats.com/benchmarks/hellaswag", + "https://api.llm-stats.com/leaderboard/benchmarks/hellaswag" + ], + "additional_details": { + "raw_benchmark_id": "hellaswag", + "raw_model_id": "claude-3-sonnet-20240229", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A challenging commonsense natural language inference dataset that uses Adversarial Filtering to create questions trivial for humans (>95% accuracy) but difficult for state-of-the-art models, requiring completion of sentence endings based on physical situations and everyday activities", + "metric_id": "llm_stats.hellaswag.score", + "metric_name": "HellaSwag score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hellaswag", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HellaSwag", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.89, + "details": { + "raw_score": "0.89", + "raw_score_field": "score", + "raw_model_id": "claude-3-sonnet-20240229", + "raw_benchmark_id": "hellaswag", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-sonnet-20240229\",\"https://llm-stats.com/benchmarks/hellaswag\",\"https://api.llm-stats.com/leaderboard/benchmarks/hellaswag\"]", + "raw_score_id": "hellaswag::claude-3-sonnet-20240229", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mgsm::mgsm-claude-3-sonnet-20240229", + "evaluation_name": "llm_stats.mgsm", + "source_data": { + "dataset_name": "MGSM", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-3-sonnet-20240229", + "https://llm-stats.com/benchmarks/mgsm", + "https://api.llm-stats.com/leaderboard/benchmarks/mgsm" + ], + "additional_details": { + "raw_benchmark_id": "mgsm", + "raw_model_id": "claude-3-sonnet-20240229", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MGSM (Multilingual Grade School Math) is a benchmark of grade-school math problems. Contains 250 grade-school math problems manually translated from the GSM8K dataset into ten typologically diverse languages: Spanish, French, German, Russian, Chinese, Japanese, Thai, Swahili, Bengali, and Telugu. Evaluates multilingual mathematical reasoning capabilities.", + "metric_id": "llm_stats.mgsm.score", + "metric_name": "MGSM score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mgsm", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MGSM", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "31" + } + }, + "score_details": { + "score": 0.835, + "details": { + "raw_score": "0.835", + "raw_score_field": "score", + "raw_model_id": "claude-3-sonnet-20240229", + "raw_benchmark_id": "mgsm", + "source_urls_json": "[\"https://llm-stats.com/models/claude-3-sonnet-20240229\",\"https://llm-stats.com/benchmarks/mgsm\",\"https://api.llm-stats.com/leaderboard/benchmarks/mgsm\"]", + "raw_score_id": "mgsm::claude-3-sonnet-20240229", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/anthropic/claude-mythos-preview/13f97aad-f8cf-417c-93db-8e2d0de57a1c.json b/data/llm_stats/anthropic/claude-mythos-preview/13f97aad-f8cf-417c-93db-8e2d0de57a1c.json new file mode 100644 index 0000000000000000000000000000000000000000..849427174b82c91980664fcab3ef5b884c795ee3 --- /dev/null +++ b/data/llm_stats/anthropic/claude-mythos-preview/13f97aad-f8cf-417c-93db-8e2d0de57a1c.json @@ -0,0 +1,877 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/anthropic_claude-mythos-preview/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Claude Mythos Preview", + "id": "anthropic/claude-mythos-preview", + "developer": "anthropic", + "additional_details": { + "raw_id": "claude-mythos-preview", + "raw_name": "Claude Mythos Preview", + "raw_model_id": "claude-mythos-preview", + "raw_model_name": "Claude Mythos Preview", + "raw_organization_id": "anthropic", + "raw_organization_name": "Anthropic", + "raw_announcement_date": "2026-04-07", + "raw_multimodal": "true", + "raw_provider_slug": "anthropic", + "raw_provider_name": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "browsecomp::browsecomp-claude-mythos-preview", + "evaluation_name": "llm_stats.browsecomp", + "source_data": { + "dataset_name": "BrowseComp", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-mythos-preview", + "https://llm-stats.com/benchmarks/browsecomp", + "https://api.llm-stats.com/leaderboard/benchmarks/browsecomp" + ], + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_model_id": "claude-mythos-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BrowseComp is a benchmark comprising 1,266 questions that challenge AI agents to persistently navigate the internet in search of hard-to-find, entangled information. The benchmark measures agents' ability to exercise persistence in information gathering, demonstrate creativity in web navigation, and find concise, verifiable answers. Despite the difficulty of the questions, BrowseComp is simple and easy-to-use, as predicted answers are short and easily verifiable against reference answers.", + "metric_id": "llm_stats.browsecomp.score", + "metric_name": "BrowseComp score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BrowseComp", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "52" + } + }, + "score_details": { + "score": 0.869, + "details": { + "raw_score": "0.869", + "raw_score_field": "score", + "raw_model_id": "claude-mythos-preview", + "raw_benchmark_id": "browsecomp", + "source_urls_json": "[\"https://llm-stats.com/models/claude-mythos-preview\",\"https://llm-stats.com/benchmarks/browsecomp\",\"https://api.llm-stats.com/leaderboard/benchmarks/browsecomp\"]", + "raw_score_id": "browsecomp::claude-mythos-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "charxiv-r::charxiv-r-claude-mythos-preview", + "evaluation_name": "llm_stats.charxiv-r", + "source_data": { + "dataset_name": "CharXiv-R", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-mythos-preview", + "https://llm-stats.com/benchmarks/charxiv-r", + "https://api.llm-stats.com/leaderboard/benchmarks/charxiv-r" + ], + "additional_details": { + "raw_benchmark_id": "charxiv-r", + "raw_model_id": "claude-mythos-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CharXiv-R is the reasoning component of the CharXiv benchmark, focusing on complex reasoning questions that require synthesizing information across visual chart elements. It evaluates multimodal large language models on their ability to understand and reason about scientific charts from arXiv papers through various reasoning tasks.", + "metric_id": "llm_stats.charxiv-r.score", + "metric_name": "CharXiv-R score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "charxiv-r", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CharXiv-R", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "43" + } + }, + "score_details": { + "score": 0.932, + "details": { + "raw_score": "0.932", + "raw_score_field": "score", + "raw_model_id": "claude-mythos-preview", + "raw_benchmark_id": "charxiv-r", + "source_urls_json": "[\"https://llm-stats.com/models/claude-mythos-preview\",\"https://llm-stats.com/benchmarks/charxiv-r\",\"https://api.llm-stats.com/leaderboard/benchmarks/charxiv-r\"]", + "raw_score_id": "charxiv-r::claude-mythos-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "cybench::cybench-claude-mythos-preview", + "evaluation_name": "llm_stats.cybench", + "source_data": { + "dataset_name": "CyBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-mythos-preview", + "https://llm-stats.com/benchmarks/cybench", + "https://api.llm-stats.com/leaderboard/benchmarks/cybench" + ], + "additional_details": { + "raw_benchmark_id": "cybench", + "raw_model_id": "claude-mythos-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CyBench is a suite of Capture-the-Flag (CTF) challenges measuring agentic cyber attack capabilities. It evaluates dual-use cybersecurity knowledge and measures the 'unguided success rate', where agents complete tasks end-to-end without guidance on appropriate subtasks.", + "metric_id": "llm_stats.cybench.score", + "metric_name": "CyBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "cybench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CyBench", + "raw_categories": "[\"safety\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 1.0, + "details": { + "raw_score": "1.0", + "raw_score_field": "score", + "raw_model_id": "claude-mythos-preview", + "raw_benchmark_id": "cybench", + "source_urls_json": "[\"https://llm-stats.com/models/claude-mythos-preview\",\"https://llm-stats.com/benchmarks/cybench\",\"https://api.llm-stats.com/leaderboard/benchmarks/cybench\"]", + "raw_score_id": "cybench::claude-mythos-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "cybergym::cybergym-claude-mythos-preview", + "evaluation_name": "llm_stats.cybergym", + "source_data": { + "dataset_name": "CyberGym", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-mythos-preview", + "https://llm-stats.com/benchmarks/cybergym", + "https://api.llm-stats.com/leaderboard/benchmarks/cybergym" + ], + "additional_details": { + "raw_benchmark_id": "cybergym", + "raw_model_id": "claude-mythos-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CyberGym is a benchmark for evaluating AI agents on cybersecurity tasks, testing their ability to identify vulnerabilities, perform security analysis, and complete security-related challenges in a controlled environment.", + "metric_id": "llm_stats.cybergym.score", + "metric_name": "CyberGym score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "cybergym", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CyberGym", + "raw_categories": "[\"safety\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.831, + "details": { + "raw_score": "0.831", + "raw_score_field": "score", + "raw_model_id": "claude-mythos-preview", + "raw_benchmark_id": "cybergym", + "source_urls_json": "[\"https://llm-stats.com/models/claude-mythos-preview\",\"https://llm-stats.com/benchmarks/cybergym\",\"https://api.llm-stats.com/leaderboard/benchmarks/cybergym\"]", + "raw_score_id": "cybergym::claude-mythos-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "figqa::figqa-claude-mythos-preview", + "evaluation_name": "llm_stats.figqa", + "source_data": { + "dataset_name": "FigQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-mythos-preview", + "https://llm-stats.com/benchmarks/figqa", + "https://api.llm-stats.com/leaderboard/benchmarks/figqa" + ], + "additional_details": { + "raw_benchmark_id": "figqa", + "raw_model_id": "claude-mythos-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "FigQA is a multiple-choice benchmark on interpreting scientific figures from biology papers. It evaluates dual-use biological knowledge and multimodal reasoning relevant to bioweapons development.", + "metric_id": "llm_stats.figqa.score", + "metric_name": "FigQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "figqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FigQA", + "raw_categories": "[\"safety\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.89, + "details": { + "raw_score": "0.89", + "raw_score_field": "score", + "raw_model_id": "claude-mythos-preview", + "raw_benchmark_id": "figqa", + "source_urls_json": "[\"https://llm-stats.com/models/claude-mythos-preview\",\"https://llm-stats.com/benchmarks/figqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/figqa\"]", + "raw_score_id": "figqa::claude-mythos-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gpqa::gpqa-claude-mythos-preview", + "evaluation_name": "llm_stats.gpqa", + "source_data": { + "dataset_name": "GPQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-mythos-preview", + "https://llm-stats.com/benchmarks/gpqa", + "https://api.llm-stats.com/leaderboard/benchmarks/gpqa" + ], + "additional_details": { + "raw_benchmark_id": "gpqa", + "raw_model_id": "claude-mythos-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A challenging dataset of 448 multiple-choice questions written by domain experts in biology, physics, and chemistry. Questions are Google-proof and extremely difficult, with PhD experts reaching 65% accuracy.", + "metric_id": "llm_stats.gpqa.score", + "metric_name": "GPQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gpqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GPQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\",\"biology\",\"chemistry\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "226" + } + }, + "score_details": { + "score": 0.946, + "details": { + "raw_score": "0.946", + "raw_score_field": "score", + "raw_model_id": "claude-mythos-preview", + "raw_benchmark_id": "gpqa", + "source_urls_json": "[\"https://llm-stats.com/models/claude-mythos-preview\",\"https://llm-stats.com/benchmarks/gpqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/gpqa\"]", + "raw_score_id": "gpqa::claude-mythos-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "graphwalks-bfs-128k::graphwalks-bfs-128k-claude-mythos-preview", + "evaluation_name": "llm_stats.graphwalks-bfs-128k", + "source_data": { + "dataset_name": "Graphwalks BFS >128k", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-mythos-preview", + "https://llm-stats.com/benchmarks/graphwalks-bfs-128k", + "https://api.llm-stats.com/leaderboard/benchmarks/graphwalks-bfs->128k" + ], + "additional_details": { + "raw_benchmark_id": "graphwalks-bfs->128k", + "raw_model_id": "claude-mythos-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A graph reasoning benchmark that evaluates language models' ability to perform breadth-first search (BFS) operations on graphs with context length over 128k tokens, testing long-context reasoning capabilities.", + "metric_id": "llm_stats.graphwalks-bfs-128k.score", + "metric_name": "Graphwalks BFS >128k score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "graphwalks-bfs->128k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Graphwalks BFS >128k", + "raw_categories": "[\"long_context\",\"reasoning\",\"spatial_reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.8, + "details": { + "raw_score": "0.8", + "raw_score_field": "score", + "raw_model_id": "claude-mythos-preview", + "raw_benchmark_id": "graphwalks-bfs->128k", + "source_urls_json": "[\"https://llm-stats.com/models/claude-mythos-preview\",\"https://llm-stats.com/benchmarks/graphwalks-bfs-128k\",\"https://api.llm-stats.com/leaderboard/benchmarks/graphwalks-bfs->128k\"]", + "raw_score_id": "graphwalks-bfs->128k::claude-mythos-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "humanity-s-last-exam::humanity-s-last-exam-claude-mythos-preview", + "evaluation_name": "llm_stats.humanity-s-last-exam", + "source_data": { + "dataset_name": "Humanity's Last Exam", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-mythos-preview", + "https://llm-stats.com/benchmarks/humanity-s-last-exam", + "https://api.llm-stats.com/leaderboard/benchmarks/humanity's-last-exam" + ], + "additional_details": { + "raw_benchmark_id": "humanity's-last-exam", + "raw_model_id": "claude-mythos-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Humanity's Last Exam (HLE) is a multi-modal academic benchmark with 2,500 questions across mathematics, humanities, and natural sciences, designed to test LLM capabilities at the frontier of human knowledge with unambiguous, verifiable solutions", + "metric_id": "llm_stats.humanity-s-last-exam.score", + "metric_name": "Humanity's Last Exam score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "humanity's-last-exam", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Humanity's Last Exam", + "raw_categories": "[\"math\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "88" + } + }, + "score_details": { + "score": 0.647, + "details": { + "raw_score": "0.647", + "raw_score_field": "score", + "raw_model_id": "claude-mythos-preview", + "raw_benchmark_id": "humanity's-last-exam", + "source_urls_json": "[\"https://llm-stats.com/models/claude-mythos-preview\",\"https://llm-stats.com/benchmarks/humanity-s-last-exam\",\"https://api.llm-stats.com/leaderboard/benchmarks/humanity's-last-exam\"]", + "raw_score_id": "humanity's-last-exam::claude-mythos-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmmlu::mmmlu-claude-mythos-preview", + "evaluation_name": "llm_stats.mmmlu", + "source_data": { + "dataset_name": "MMMLU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-mythos-preview", + "https://llm-stats.com/benchmarks/mmmlu", + "https://api.llm-stats.com/leaderboard/benchmarks/mmmlu" + ], + "additional_details": { + "raw_benchmark_id": "mmmlu", + "raw_model_id": "claude-mythos-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Multilingual Massive Multitask Language Understanding dataset released by OpenAI, featuring professionally translated MMLU test questions across 14 languages including Arabic, Bengali, German, Spanish, French, Hindi, Indonesian, Italian, Japanese, Korean, Portuguese, Swahili, Yoruba, and Chinese. Contains approximately 15,908 multiple-choice questions per language covering 57 subjects.", + "metric_id": "llm_stats.mmmlu.score", + "metric_name": "MMMLU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmmlu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMMLU", + "raw_categories": "[\"language\",\"math\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "49" + } + }, + "score_details": { + "score": 0.927, + "details": { + "raw_score": "0.927", + "raw_score_field": "score", + "raw_model_id": "claude-mythos-preview", + "raw_benchmark_id": "mmmlu", + "source_urls_json": "[\"https://llm-stats.com/models/claude-mythos-preview\",\"https://llm-stats.com/benchmarks/mmmlu\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmmlu\"]", + "raw_score_id": "mmmlu::claude-mythos-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "osworld-verified::osworld-verified-claude-mythos-preview", + "evaluation_name": "llm_stats.osworld-verified", + "source_data": { + "dataset_name": "OSWorld-Verified", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-mythos-preview", + "https://llm-stats.com/benchmarks/osworld-verified", + "https://api.llm-stats.com/leaderboard/benchmarks/osworld-verified" + ], + "additional_details": { + "raw_benchmark_id": "osworld-verified", + "raw_model_id": "claude-mythos-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OSWorld-Verified is a verified subset of OSWorld, a scalable real computer environment for multimodal agents supporting task setup, execution-based evaluation, and interactive learning across Ubuntu, Windows, and macOS.", + "metric_id": "llm_stats.osworld-verified.score", + "metric_name": "OSWorld-Verified score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "osworld-verified", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OSWorld-Verified", + "raw_categories": "[\"multimodal\",\"general\",\"agents\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.796, + "details": { + "raw_score": "0.796", + "raw_score_field": "score", + "raw_model_id": "claude-mythos-preview", + "raw_benchmark_id": "osworld-verified", + "source_urls_json": "[\"https://llm-stats.com/models/claude-mythos-preview\",\"https://llm-stats.com/benchmarks/osworld-verified\",\"https://api.llm-stats.com/leaderboard/benchmarks/osworld-verified\"]", + "raw_score_id": "osworld-verified::claude-mythos-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-multilingual::swe-bench-multilingual-claude-mythos-preview", + "evaluation_name": "llm_stats.swe-bench-multilingual", + "source_data": { + "dataset_name": "SWE-bench Multilingual", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-mythos-preview", + "https://llm-stats.com/benchmarks/swe-bench-multilingual", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-multilingual" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-multilingual", + "raw_model_id": "claude-mythos-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multilingual benchmark for issue resolving in software engineering that covers Java, TypeScript, JavaScript, Go, Rust, C, and C++. Contains 1,632 high-quality instances carefully annotated from 2,456 candidates by 68 expert annotators, designed to evaluate Large Language Models across diverse software ecosystems beyond Python.", + "metric_id": "llm_stats.swe-bench-multilingual.score", + "metric_name": "SWE-bench Multilingual score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-multilingual", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-bench Multilingual", + "raw_categories": "[\"reasoning\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "33" + } + }, + "score_details": { + "score": 0.873, + "details": { + "raw_score": "0.873", + "raw_score_field": "score", + "raw_model_id": "claude-mythos-preview", + "raw_benchmark_id": "swe-bench-multilingual", + "source_urls_json": "[\"https://llm-stats.com/models/claude-mythos-preview\",\"https://llm-stats.com/benchmarks/swe-bench-multilingual\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-multilingual\"]", + "raw_score_id": "swe-bench-multilingual::claude-mythos-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-multimodal::swe-bench-multimodal-claude-mythos-preview", + "evaluation_name": "llm_stats.swe-bench-multimodal", + "source_data": { + "dataset_name": "SWE-Bench Multimodal", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-mythos-preview", + "https://llm-stats.com/benchmarks/swe-bench-multimodal", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-multimodal" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-multimodal", + "raw_model_id": "claude-mythos-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SWE-Bench Multimodal extends SWE-Bench to evaluate language models on software engineering tasks that involve visual inputs such as screenshots, UI mockups, and diagrams alongside code understanding.", + "metric_id": "llm_stats.swe-bench-multimodal.score", + "metric_name": "SWE-Bench Multimodal score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-multimodal", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Bench Multimodal", + "raw_categories": "[\"multimodal\",\"reasoning\",\"agents\",\"code\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.59, + "details": { + "raw_score": "0.59", + "raw_score_field": "score", + "raw_model_id": "claude-mythos-preview", + "raw_benchmark_id": "swe-bench-multimodal", + "source_urls_json": "[\"https://llm-stats.com/models/claude-mythos-preview\",\"https://llm-stats.com/benchmarks/swe-bench-multimodal\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-multimodal\"]", + "raw_score_id": "swe-bench-multimodal::claude-mythos-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-pro::swe-bench-pro-claude-mythos-preview", + "evaluation_name": "llm_stats.swe-bench-pro", + "source_data": { + "dataset_name": "SWE-Bench Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-mythos-preview", + "https://llm-stats.com/benchmarks/swe-bench-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-pro" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-pro", + "raw_model_id": "claude-mythos-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SWE-Bench Pro is an advanced version of SWE-Bench that evaluates language models on complex, real-world software engineering tasks requiring extended reasoning and multi-step problem solving.", + "metric_id": "llm_stats.swe-bench-pro.score", + "metric_name": "SWE-Bench Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Bench Pro", + "raw_categories": "[\"reasoning\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "35" + } + }, + "score_details": { + "score": 0.778, + "details": { + "raw_score": "0.778", + "raw_score_field": "score", + "raw_model_id": "claude-mythos-preview", + "raw_benchmark_id": "swe-bench-pro", + "source_urls_json": "[\"https://llm-stats.com/models/claude-mythos-preview\",\"https://llm-stats.com/benchmarks/swe-bench-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-pro\"]", + "raw_score_id": "swe-bench-pro::claude-mythos-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-verified::swe-bench-verified-claude-mythos-preview", + "evaluation_name": "llm_stats.swe-bench-verified", + "source_data": { + "dataset_name": "SWE-Bench Verified", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-mythos-preview", + "https://llm-stats.com/benchmarks/swe-bench-verified", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-verified" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-verified", + "raw_model_id": "claude-mythos-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A verified subset of 500 software engineering problems from real GitHub issues, validated by human annotators for evaluating language models' ability to resolve real-world coding issues by generating patches for Python codebases.", + "metric_id": "llm_stats.swe-bench-verified.score", + "metric_name": "SWE-Bench Verified score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-verified", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Bench Verified", + "raw_categories": "[\"reasoning\",\"frontend_development\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "103" + } + }, + "score_details": { + "score": 0.939, + "details": { + "raw_score": "0.939", + "raw_score_field": "score", + "raw_model_id": "claude-mythos-preview", + "raw_benchmark_id": "swe-bench-verified", + "source_urls_json": "[\"https://llm-stats.com/models/claude-mythos-preview\",\"https://llm-stats.com/benchmarks/swe-bench-verified\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-verified\"]", + "raw_score_id": "swe-bench-verified::claude-mythos-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2::terminal-bench-2-claude-mythos-preview", + "evaluation_name": "llm_stats.terminal-bench-2", + "source_data": { + "dataset_name": "Terminal-Bench 2.0", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-mythos-preview", + "https://llm-stats.com/benchmarks/terminal-bench-2", + "https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2" + ], + "additional_details": { + "raw_benchmark_id": "terminal-bench-2", + "raw_model_id": "claude-mythos-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Terminal-Bench 2.0 is an updated benchmark for testing AI agents' tool use ability to operate a computer via terminal. It evaluates how well models can handle real-world, end-to-end tasks autonomously, including compiling code, training models, setting up servers, system administration, security tasks, data science workflows, and cybersecurity vulnerabilities.", + "metric_id": "llm_stats.terminal-bench-2.score", + "metric_name": "Terminal-Bench 2.0 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "terminal-bench-2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Terminal-Bench 2.0", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "49" + } + }, + "score_details": { + "score": 0.82, + "details": { + "raw_score": "0.82", + "raw_score_field": "score", + "raw_model_id": "claude-mythos-preview", + "raw_benchmark_id": "terminal-bench-2", + "source_urls_json": "[\"https://llm-stats.com/models/claude-mythos-preview\",\"https://llm-stats.com/benchmarks/terminal-bench-2\",\"https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2\"]", + "raw_score_id": "terminal-bench-2::claude-mythos-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "usamo25::usamo25-claude-mythos-preview", + "evaluation_name": "llm_stats.usamo25", + "source_data": { + "dataset_name": "USAMO25", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-mythos-preview", + "https://llm-stats.com/benchmarks/usamo25", + "https://api.llm-stats.com/leaderboard/benchmarks/usamo25" + ], + "additional_details": { + "raw_benchmark_id": "usamo25", + "raw_model_id": "claude-mythos-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The 2025 United States of America Mathematical Olympiad (USAMO) benchmark consists of six challenging mathematical problems requiring rigorous proof-based reasoning. USAMO is the most prestigious high school mathematics competition in the United States, serving as the final round of the American Mathematics Competitions series. This benchmark evaluates models on mathematical problem-solving capabilities beyond simple numerical computation, focusing on formal mathematical reasoning and proof generation.", + "metric_id": "llm_stats.usamo25.score", + "metric_name": "USAMO25 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "usamo25", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "USAMO25", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.976, + "details": { + "raw_score": "0.976", + "raw_score_field": "score", + "raw_model_id": "claude-mythos-preview", + "raw_benchmark_id": "usamo25", + "source_urls_json": "[\"https://llm-stats.com/models/claude-mythos-preview\",\"https://llm-stats.com/benchmarks/usamo25\",\"https://api.llm-stats.com/leaderboard/benchmarks/usamo25\"]", + "raw_score_id": "usamo25::claude-mythos-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/anthropic/claude-opus-4-6/76548115-4eec-445c-a227-07a8368c55a5.json b/data/llm_stats/anthropic/claude-opus-4-6/76548115-4eec-445c-a227-07a8368c55a5.json new file mode 100644 index 0000000000000000000000000000000000000000..415fe969a788ce4b82679e87ae4ad8b32124eace --- /dev/null +++ b/data/llm_stats/anthropic/claude-opus-4-6/76548115-4eec-445c-a227-07a8368c55a5.json @@ -0,0 +1,1296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/anthropic_claude-opus-4-6/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-6", + "id": "anthropic/claude-opus-4-6", + "developer": "anthropic", + "additional_details": { + "raw_id": "claude-opus-4-6", + "raw_model_name": "claude-opus-4-6", + "raw_display_name": "Claude Opus 4.6", + "raw_organization_id": "anthropic", + "raw_organization_name": "Anthropic", + "raw_context_length": "1000000", + "raw_input_modalities": "[\"image\",\"text\"]", + "raw_output_modalities": "[\"text\"]", + "raw_input_price": "5.0", + "raw_output_price": "25.0", + "raw_model_id": "claude-opus-4-6", + "raw_provider_slug": "anthropic", + "raw_provider_name": "ZeroEval" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aime-2025::aime-2025-claude-opus-4-6", + "evaluation_name": "llm_stats.aime-2025", + "source_data": { + "dataset_name": "AIME 2025", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-6", + "https://llm-stats.com/benchmarks/aime-2025", + "https://api.llm-stats.com/leaderboard/benchmarks/aime-2025" + ], + "additional_details": { + "raw_benchmark_id": "aime-2025", + "raw_model_id": "claude-opus-4-6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "All 30 problems from the 2025 American Invitational Mathematics Examination (AIME I and AIME II), testing olympiad-level mathematical reasoning with integer answers from 000-999. Used as an AI benchmark to evaluate large language models' ability to solve complex mathematical problems requiring multi-step logical deductions and structured symbolic reasoning.", + "metric_id": "llm_stats.aime-2025.score", + "metric_name": "AIME 2025 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aime-2025", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AIME 2025", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "114" + } + }, + "score_details": { + "score": 0.9979, + "details": { + "raw_score": "0.9979", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-6", + "raw_benchmark_id": "aime-2025", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-6\",\"https://llm-stats.com/benchmarks/aime-2025\",\"https://api.llm-stats.com/leaderboard/benchmarks/aime-2025\"]", + "raw_score_id": "aime-2025::claude-opus-4-6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "arc-agi-v2::arc-agi-v2-claude-opus-4-6", + "evaluation_name": "llm_stats.arc-agi-v2", + "source_data": { + "dataset_name": "ARC-AGI v2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-6", + "https://llm-stats.com/benchmarks/arc-agi-v2", + "https://api.llm-stats.com/leaderboard/benchmarks/arc-agi-v2" + ], + "additional_details": { + "raw_benchmark_id": "arc-agi-v2", + "raw_model_id": "claude-opus-4-6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ARC-AGI-2 is an upgraded benchmark for measuring abstract reasoning and problem-solving abilities in AI systems through visual grid transformation tasks. It evaluates fluid intelligence via input-output grid pairs (1x1 to 30x30) using colored cells (0-9), requiring models to identify underlying transformation rules from demonstration examples and apply them to test cases. Designed to be easy for humans but challenging for AI, focusing on core cognitive abilities like spatial reasoning, pattern recognition, and compositional generalization.", + "metric_id": "llm_stats.arc-agi-v2.score", + "metric_name": "ARC-AGI v2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arc-agi-v2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ARC-AGI v2", + "raw_categories": "[\"reasoning\",\"spatial_reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.688, + "details": { + "raw_score": "0.688", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-6", + "raw_benchmark_id": "arc-agi-v2", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-6\",\"https://llm-stats.com/benchmarks/arc-agi-v2\",\"https://api.llm-stats.com/leaderboard/benchmarks/arc-agi-v2\"]", + "raw_score_id": "arc-agi-v2::claude-opus-4-6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "browsecomp::browsecomp-claude-opus-4-6", + "evaluation_name": "llm_stats.browsecomp", + "source_data": { + "dataset_name": "BrowseComp", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-6", + "https://llm-stats.com/benchmarks/browsecomp", + "https://api.llm-stats.com/leaderboard/benchmarks/browsecomp" + ], + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_model_id": "claude-opus-4-6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BrowseComp is a benchmark comprising 1,266 questions that challenge AI agents to persistently navigate the internet in search of hard-to-find, entangled information. The benchmark measures agents' ability to exercise persistence in information gathering, demonstrate creativity in web navigation, and find concise, verifiable answers. Despite the difficulty of the questions, BrowseComp is simple and easy-to-use, as predicted answers are short and easily verifiable against reference answers.", + "metric_id": "llm_stats.browsecomp.score", + "metric_name": "BrowseComp score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BrowseComp", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "52" + } + }, + "score_details": { + "score": 0.84, + "details": { + "raw_score": "0.84", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-6", + "raw_benchmark_id": "browsecomp", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-6\",\"https://llm-stats.com/benchmarks/browsecomp\",\"https://api.llm-stats.com/leaderboard/benchmarks/browsecomp\"]", + "raw_score_id": "browsecomp::claude-opus-4-6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "cybergym::cybergym-claude-opus-4-6", + "evaluation_name": "llm_stats.cybergym", + "source_data": { + "dataset_name": "CyberGym", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-6", + "https://llm-stats.com/benchmarks/cybergym", + "https://api.llm-stats.com/leaderboard/benchmarks/cybergym" + ], + "additional_details": { + "raw_benchmark_id": "cybergym", + "raw_model_id": "claude-opus-4-6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CyberGym is a benchmark for evaluating AI agents on cybersecurity tasks, testing their ability to identify vulnerabilities, perform security analysis, and complete security-related challenges in a controlled environment.", + "metric_id": "llm_stats.cybergym.score", + "metric_name": "CyberGym score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "cybergym", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CyberGym", + "raw_categories": "[\"safety\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.738, + "details": { + "raw_score": "0.738", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-6", + "raw_benchmark_id": "cybergym", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-6\",\"https://llm-stats.com/benchmarks/cybergym\",\"https://api.llm-stats.com/leaderboard/benchmarks/cybergym\"]", + "raw_score_id": "cybergym::claude-opus-4-6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "deepsearchqa::deepsearchqa-claude-opus-4-6", + "evaluation_name": "llm_stats.deepsearchqa", + "source_data": { + "dataset_name": "DeepSearchQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-6", + "https://llm-stats.com/benchmarks/deepsearchqa", + "https://api.llm-stats.com/leaderboard/benchmarks/deepsearchqa" + ], + "additional_details": { + "raw_benchmark_id": "deepsearchqa", + "raw_model_id": "claude-opus-4-6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "DeepSearchQA is a benchmark for evaluating deep search and question-answering capabilities, testing models' ability to perform multi-hop reasoning and information retrieval across complex knowledge domains.", + "metric_id": "llm_stats.deepsearchqa.score", + "metric_name": "DeepSearchQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "deepsearchqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DeepSearchQA", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.913, + "details": { + "raw_score": "0.913", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-6", + "raw_benchmark_id": "deepsearchqa", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-6\",\"https://llm-stats.com/benchmarks/deepsearchqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/deepsearchqa\"]", + "raw_score_id": "deepsearchqa::claude-opus-4-6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "figqa::figqa-claude-opus-4-6", + "evaluation_name": "llm_stats.figqa", + "source_data": { + "dataset_name": "FigQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-6", + "https://llm-stats.com/benchmarks/figqa", + "https://api.llm-stats.com/leaderboard/benchmarks/figqa" + ], + "additional_details": { + "raw_benchmark_id": "figqa", + "raw_model_id": "claude-opus-4-6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "FigQA is a multiple-choice benchmark on interpreting scientific figures from biology papers. It evaluates dual-use biological knowledge and multimodal reasoning relevant to bioweapons development.", + "metric_id": "llm_stats.figqa.score", + "metric_name": "FigQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "figqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FigQA", + "raw_categories": "[\"safety\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.783, + "details": { + "raw_score": "0.783", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-6", + "raw_benchmark_id": "figqa", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-6\",\"https://llm-stats.com/benchmarks/figqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/figqa\"]", + "raw_score_id": "figqa::claude-opus-4-6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "finance-agent::finance-agent-claude-opus-4-6", + "evaluation_name": "llm_stats.finance-agent", + "source_data": { + "dataset_name": "Finance Agent", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-6", + "https://llm-stats.com/benchmarks/finance-agent", + "https://api.llm-stats.com/leaderboard/benchmarks/finance-agent" + ], + "additional_details": { + "raw_benchmark_id": "finance-agent", + "raw_model_id": "claude-opus-4-6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Finance Agent is a benchmark for evaluating AI models on agentic financial analysis tasks, testing their ability to process financial data, perform calculations, and generate accurate analyses across various financial domains.", + "metric_id": "llm_stats.finance-agent.score", + "metric_name": "Finance Agent score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "finance-agent", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Finance Agent", + "raw_categories": "[\"reasoning\",\"finance\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.607, + "details": { + "raw_score": "0.607", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-6", + "raw_benchmark_id": "finance-agent", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-6\",\"https://llm-stats.com/benchmarks/finance-agent\",\"https://api.llm-stats.com/leaderboard/benchmarks/finance-agent\"]", + "raw_score_id": "finance-agent::claude-opus-4-6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gdpval-aa::gdpval-aa-claude-opus-4-6", + "evaluation_name": "llm_stats.gdpval-aa", + "source_data": { + "dataset_name": "GDPval-AA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-6", + "https://llm-stats.com/benchmarks/gdpval-aa", + "https://api.llm-stats.com/leaderboard/benchmarks/gdpval-aa" + ], + "additional_details": { + "raw_benchmark_id": "gdpval-aa", + "raw_model_id": "claude-opus-4-6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "GDPval-AA is an evaluation of AI model performance on economically valuable knowledge work tasks across professional domains including finance, legal, and other sectors. Run independently by Artificial Analysis, it uses Elo scoring to rank models on real-world work task performance.", + "metric_id": "llm_stats.gdpval-aa.score", + "metric_name": "GDPval-AA score", + "metric_kind": "benchmark_score", + "metric_unit": "points", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1606.0, + "additional_details": { + "raw_benchmark_id": "gdpval-aa", + "raw_score_field": "score", + "bound_strategy": "observed_value", + "raw_name": "GDPval-AA", + "raw_categories": "[\"legal\",\"reasoning\",\"finance\",\"general\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "34" + } + }, + "score_details": { + "score": 1606.0, + "details": { + "raw_score": "1606.0", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-6", + "raw_benchmark_id": "gdpval-aa", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-6\",\"https://llm-stats.com/benchmarks/gdpval-aa\",\"https://api.llm-stats.com/leaderboard/benchmarks/gdpval-aa\"]", + "raw_score_id": "gdpval-aa::claude-opus-4-6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gpqa::gpqa-claude-opus-4-6", + "evaluation_name": "llm_stats.gpqa", + "source_data": { + "dataset_name": "GPQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-6", + "https://llm-stats.com/benchmarks/gpqa", + "https://api.llm-stats.com/leaderboard/benchmarks/gpqa" + ], + "additional_details": { + "raw_benchmark_id": "gpqa", + "raw_model_id": "claude-opus-4-6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A challenging dataset of 448 multiple-choice questions written by domain experts in biology, physics, and chemistry. Questions are Google-proof and extremely difficult, with PhD experts reaching 65% accuracy.", + "metric_id": "llm_stats.gpqa.score", + "metric_name": "GPQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gpqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GPQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\",\"biology\",\"chemistry\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "226" + } + }, + "score_details": { + "score": 0.913, + "details": { + "raw_score": "0.913", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-6", + "raw_benchmark_id": "gpqa", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-6\",\"https://llm-stats.com/benchmarks/gpqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/gpqa\"]", + "raw_score_id": "gpqa::claude-opus-4-6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "graphwalks-bfs-128k::graphwalks-bfs-128k-claude-opus-4-6", + "evaluation_name": "llm_stats.graphwalks-bfs-128k", + "source_data": { + "dataset_name": "Graphwalks BFS >128k", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-6", + "https://llm-stats.com/benchmarks/graphwalks-bfs-128k", + "https://api.llm-stats.com/leaderboard/benchmarks/graphwalks-bfs->128k" + ], + "additional_details": { + "raw_benchmark_id": "graphwalks-bfs->128k", + "raw_model_id": "claude-opus-4-6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A graph reasoning benchmark that evaluates language models' ability to perform breadth-first search (BFS) operations on graphs with context length over 128k tokens, testing long-context reasoning capabilities.", + "metric_id": "llm_stats.graphwalks-bfs-128k.score", + "metric_name": "Graphwalks BFS >128k score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "graphwalks-bfs->128k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Graphwalks BFS >128k", + "raw_categories": "[\"long_context\",\"reasoning\",\"spatial_reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.615, + "details": { + "raw_score": "0.615", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-6", + "raw_benchmark_id": "graphwalks-bfs->128k", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-6\",\"https://llm-stats.com/benchmarks/graphwalks-bfs-128k\",\"https://api.llm-stats.com/leaderboard/benchmarks/graphwalks-bfs->128k\"]", + "raw_score_id": "graphwalks-bfs->128k::claude-opus-4-6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "graphwalks-parents-128k::graphwalks-parents-128k-claude-opus-4-6", + "evaluation_name": "llm_stats.graphwalks-parents-128k", + "source_data": { + "dataset_name": "Graphwalks parents >128k", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-6", + "https://llm-stats.com/benchmarks/graphwalks-parents-128k", + "https://api.llm-stats.com/leaderboard/benchmarks/graphwalks-parents->128k" + ], + "additional_details": { + "raw_benchmark_id": "graphwalks-parents->128k", + "raw_model_id": "claude-opus-4-6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A graph reasoning benchmark that evaluates language models' ability to find parent nodes in graphs with context length over 128k tokens, testing long-context reasoning and graph structure understanding.", + "metric_id": "llm_stats.graphwalks-parents-128k.score", + "metric_name": "Graphwalks parents >128k score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "graphwalks-parents->128k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Graphwalks parents >128k", + "raw_categories": "[\"long_context\",\"reasoning\",\"spatial_reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.954, + "details": { + "raw_score": "0.954", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-6", + "raw_benchmark_id": "graphwalks-parents->128k", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-6\",\"https://llm-stats.com/benchmarks/graphwalks-parents-128k\",\"https://api.llm-stats.com/leaderboard/benchmarks/graphwalks-parents->128k\"]", + "raw_score_id": "graphwalks-parents->128k::claude-opus-4-6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "humanity-s-last-exam::humanity-s-last-exam-claude-opus-4-6", + "evaluation_name": "llm_stats.humanity-s-last-exam", + "source_data": { + "dataset_name": "Humanity's Last Exam", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-6", + "https://llm-stats.com/benchmarks/humanity-s-last-exam", + "https://api.llm-stats.com/leaderboard/benchmarks/humanity's-last-exam" + ], + "additional_details": { + "raw_benchmark_id": "humanity's-last-exam", + "raw_model_id": "claude-opus-4-6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Humanity's Last Exam (HLE) is a multi-modal academic benchmark with 2,500 questions across mathematics, humanities, and natural sciences, designed to test LLM capabilities at the frontier of human knowledge with unambiguous, verifiable solutions", + "metric_id": "llm_stats.humanity-s-last-exam.score", + "metric_name": "Humanity's Last Exam score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "humanity's-last-exam", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Humanity's Last Exam", + "raw_categories": "[\"math\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "88" + } + }, + "score_details": { + "score": 0.531, + "details": { + "raw_score": "0.531", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-6", + "raw_benchmark_id": "humanity's-last-exam", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-6\",\"https://llm-stats.com/benchmarks/humanity-s-last-exam\",\"https://api.llm-stats.com/leaderboard/benchmarks/humanity's-last-exam\"]", + "raw_score_id": "humanity's-last-exam::claude-opus-4-6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mcp-atlas::mcp-atlas-claude-opus-4-6", + "evaluation_name": "llm_stats.mcp-atlas", + "source_data": { + "dataset_name": "MCP Atlas", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-6", + "https://llm-stats.com/benchmarks/mcp-atlas", + "https://api.llm-stats.com/leaderboard/benchmarks/mcp-atlas" + ], + "additional_details": { + "raw_benchmark_id": "mcp-atlas", + "raw_model_id": "claude-opus-4-6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MCP Atlas is a benchmark for evaluating AI models on scaled tool use capabilities, measuring how well models can coordinate and utilize multiple tools across complex multi-step tasks.", + "metric_id": "llm_stats.mcp-atlas.score", + "metric_name": "MCP Atlas score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mcp-atlas", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MCP Atlas", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.627, + "details": { + "raw_score": "0.627", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-6", + "raw_benchmark_id": "mcp-atlas", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-6\",\"https://llm-stats.com/benchmarks/mcp-atlas\",\"https://api.llm-stats.com/leaderboard/benchmarks/mcp-atlas\"]", + "raw_score_id": "mcp-atlas::claude-opus-4-6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmmlu::mmmlu-claude-opus-4-6", + "evaluation_name": "llm_stats.mmmlu", + "source_data": { + "dataset_name": "MMMLU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-6", + "https://llm-stats.com/benchmarks/mmmlu", + "https://api.llm-stats.com/leaderboard/benchmarks/mmmlu" + ], + "additional_details": { + "raw_benchmark_id": "mmmlu", + "raw_model_id": "claude-opus-4-6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Multilingual Massive Multitask Language Understanding dataset released by OpenAI, featuring professionally translated MMLU test questions across 14 languages including Arabic, Bengali, German, Spanish, French, Hindi, Indonesian, Italian, Japanese, Korean, Portuguese, Swahili, Yoruba, and Chinese. Contains approximately 15,908 multiple-choice questions per language covering 57 subjects.", + "metric_id": "llm_stats.mmmlu.score", + "metric_name": "MMMLU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmmlu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMMLU", + "raw_categories": "[\"language\",\"math\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "49" + } + }, + "score_details": { + "score": 0.911, + "details": { + "raw_score": "0.911", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-6", + "raw_benchmark_id": "mmmlu", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-6\",\"https://llm-stats.com/benchmarks/mmmlu\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmmlu\"]", + "raw_score_id": "mmmlu::claude-opus-4-6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmmu-pro::mmmu-pro-claude-opus-4-6", + "evaluation_name": "llm_stats.mmmu-pro", + "source_data": { + "dataset_name": "MMMU-Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-6", + "https://llm-stats.com/benchmarks/mmmu-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/mmmu-pro" + ], + "additional_details": { + "raw_benchmark_id": "mmmu-pro", + "raw_model_id": "claude-opus-4-6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A more robust multi-discipline multimodal understanding benchmark that enhances MMMU through a three-step process: filtering text-only answerable questions, augmenting candidate options, and introducing vision-only input settings. Achieves significantly lower model performance (16.8-26.9%) compared to original MMMU, providing more rigorous evaluation that closely mimics real-world scenarios.", + "metric_id": "llm_stats.mmmu-pro.score", + "metric_name": "MMMU-Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmmu-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMMU-Pro", + "raw_categories": "[\"multimodal\",\"reasoning\",\"general\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "60" + } + }, + "score_details": { + "score": 0.773, + "details": { + "raw_score": "0.773", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-6", + "raw_benchmark_id": "mmmu-pro", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-6\",\"https://llm-stats.com/benchmarks/mmmu-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmmu-pro\"]", + "raw_score_id": "mmmu-pro::claude-opus-4-6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mrcr-v2-8-needle::mrcr-v2-8-needle-claude-opus-4-6", + "evaluation_name": "llm_stats.mrcr-v2-8-needle", + "source_data": { + "dataset_name": "MRCR v2 (8-needle)", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-6", + "https://llm-stats.com/benchmarks/mrcr-v2-8-needle", + "https://api.llm-stats.com/leaderboard/benchmarks/mrcr-v2-(8-needle)" + ], + "additional_details": { + "raw_benchmark_id": "mrcr-v2-(8-needle)", + "raw_model_id": "claude-opus-4-6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MRCR v2 (8-needle) is a variant of the Multi-Round Coreference Resolution benchmark that includes 8 needle items to retrieve from long contexts. This tests models' ability to simultaneously track and reason about multiple pieces of information across extended conversations.", + "metric_id": "llm_stats.mrcr-v2-8-needle.score", + "metric_name": "MRCR v2 (8-needle) score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mrcr-v2-(8-needle)", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MRCR v2 (8-needle)", + "raw_categories": "[\"long_context\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.76, + "details": { + "raw_score": "0.76", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-6", + "raw_benchmark_id": "mrcr-v2-(8-needle)", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-6\",\"https://llm-stats.com/benchmarks/mrcr-v2-8-needle\",\"https://api.llm-stats.com/leaderboard/benchmarks/mrcr-v2-(8-needle)\"]", + "raw_score_id": "mrcr-v2-(8-needle)::claude-opus-4-6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "openrca::openrca-claude-opus-4-6", + "evaluation_name": "llm_stats.openrca", + "source_data": { + "dataset_name": "OpenRCA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-6", + "https://llm-stats.com/benchmarks/openrca", + "https://api.llm-stats.com/leaderboard/benchmarks/openrca" + ], + "additional_details": { + "raw_benchmark_id": "openrca", + "raw_model_id": "claude-opus-4-6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OpenRCA is a benchmark for evaluating AI models on root cause analysis tasks. For each failure case, the model receives 1 point if all generated root-cause elements match the ground-truth ones, and 0 points if any mismatch is identified. The overall accuracy is the average score across all failure cases.", + "metric_id": "llm_stats.openrca.score", + "metric_name": "OpenRCA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "openrca", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OpenRCA", + "raw_categories": "[\"reasoning\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.349, + "details": { + "raw_score": "0.349", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-6", + "raw_benchmark_id": "openrca", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-6\",\"https://llm-stats.com/benchmarks/openrca\",\"https://api.llm-stats.com/leaderboard/benchmarks/openrca\"]", + "raw_score_id": "openrca::claude-opus-4-6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "osworld::osworld-claude-opus-4-6", + "evaluation_name": "llm_stats.osworld", + "source_data": { + "dataset_name": "OSWorld", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-6", + "https://llm-stats.com/benchmarks/osworld", + "https://api.llm-stats.com/leaderboard/benchmarks/osworld" + ], + "additional_details": { + "raw_benchmark_id": "osworld", + "raw_model_id": "claude-opus-4-6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OSWorld: The first-of-its-kind scalable, real computer environment for multimodal agents, supporting task setup, execution-based evaluation, and interactive learning across Ubuntu, Windows, and macOS with 369 computer tasks involving real web and desktop applications, OS file I/O, and multi-application workflows", + "metric_id": "llm_stats.osworld.score", + "metric_name": "OSWorld score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "osworld", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OSWorld", + "raw_categories": "[\"multimodal\",\"general\",\"agents\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "20" + } + }, + "score_details": { + "score": 0.727, + "details": { + "raw_score": "0.727", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-6", + "raw_benchmark_id": "osworld", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-6\",\"https://llm-stats.com/benchmarks/osworld\",\"https://api.llm-stats.com/leaderboard/benchmarks/osworld\"]", + "raw_score_id": "osworld::claude-opus-4-6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-multilingual::swe-bench-multilingual-claude-opus-4-6", + "evaluation_name": "llm_stats.swe-bench-multilingual", + "source_data": { + "dataset_name": "SWE-bench Multilingual", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-6", + "https://llm-stats.com/benchmarks/swe-bench-multilingual", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-multilingual" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-multilingual", + "raw_model_id": "claude-opus-4-6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multilingual benchmark for issue resolving in software engineering that covers Java, TypeScript, JavaScript, Go, Rust, C, and C++. Contains 1,632 high-quality instances carefully annotated from 2,456 candidates by 68 expert annotators, designed to evaluate Large Language Models across diverse software ecosystems beyond Python.", + "metric_id": "llm_stats.swe-bench-multilingual.score", + "metric_name": "SWE-bench Multilingual score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-multilingual", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-bench Multilingual", + "raw_categories": "[\"reasoning\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "33" + } + }, + "score_details": { + "score": 0.7783, + "details": { + "raw_score": "0.7783", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-6", + "raw_benchmark_id": "swe-bench-multilingual", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-6\",\"https://llm-stats.com/benchmarks/swe-bench-multilingual\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-multilingual\"]", + "raw_score_id": "swe-bench-multilingual::claude-opus-4-6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-verified::swe-bench-verified-claude-opus-4-6", + "evaluation_name": "llm_stats.swe-bench-verified", + "source_data": { + "dataset_name": "SWE-Bench Verified", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-6", + "https://llm-stats.com/benchmarks/swe-bench-verified", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-verified" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-verified", + "raw_model_id": "claude-opus-4-6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A verified subset of 500 software engineering problems from real GitHub issues, validated by human annotators for evaluating language models' ability to resolve real-world coding issues by generating patches for Python codebases.", + "metric_id": "llm_stats.swe-bench-verified.score", + "metric_name": "SWE-Bench Verified score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-verified", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Bench Verified", + "raw_categories": "[\"reasoning\",\"frontend_development\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "103" + } + }, + "score_details": { + "score": 0.808, + "details": { + "raw_score": "0.808", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-6", + "raw_benchmark_id": "swe-bench-verified", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-6\",\"https://llm-stats.com/benchmarks/swe-bench-verified\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-verified\"]", + "raw_score_id": "swe-bench-verified::claude-opus-4-6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "tau2-retail::tau2-retail-claude-opus-4-6", + "evaluation_name": "llm_stats.tau2-retail", + "source_data": { + "dataset_name": "Tau2 Retail", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-6", + "https://llm-stats.com/benchmarks/tau2-retail", + "https://api.llm-stats.com/leaderboard/benchmarks/tau2-retail" + ], + "additional_details": { + "raw_benchmark_id": "tau2-retail", + "raw_model_id": "claude-opus-4-6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "τ²-bench retail domain evaluates conversational AI agents in customer service scenarios within a dual-control environment where both agent and user can interact with tools. Tests tool-agent-user interaction, rule adherence, and task consistency in retail customer support contexts.", + "metric_id": "llm_stats.tau2-retail.score", + "metric_name": "Tau2 Retail score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "tau2-retail", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Tau2 Retail", + "raw_categories": "[\"reasoning\",\"communication\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "26" + } + }, + "score_details": { + "score": 0.919, + "details": { + "raw_score": "0.919", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-6", + "raw_benchmark_id": "tau2-retail", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-6\",\"https://llm-stats.com/benchmarks/tau2-retail\",\"https://api.llm-stats.com/leaderboard/benchmarks/tau2-retail\"]", + "raw_score_id": "tau2-retail::claude-opus-4-6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "tau2-telecom::tau2-telecom-claude-opus-4-6", + "evaluation_name": "llm_stats.tau2-telecom", + "source_data": { + "dataset_name": "Tau2 Telecom", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-6", + "https://llm-stats.com/benchmarks/tau2-telecom", + "https://api.llm-stats.com/leaderboard/benchmarks/tau2-telecom" + ], + "additional_details": { + "raw_benchmark_id": "tau2-telecom", + "raw_model_id": "claude-opus-4-6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "τ²-Bench telecom domain evaluates conversational agents in a dual-control environment modeled as a Dec-POMDP, where both agent and user use tools in shared telecommunications troubleshooting scenarios that test coordination and communication capabilities.", + "metric_id": "llm_stats.tau2-telecom.score", + "metric_name": "Tau2 Telecom score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "tau2-telecom", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Tau2 Telecom", + "raw_categories": "[\"reasoning\",\"communication\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "35" + } + }, + "score_details": { + "score": 0.993, + "details": { + "raw_score": "0.993", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-6", + "raw_benchmark_id": "tau2-telecom", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-6\",\"https://llm-stats.com/benchmarks/tau2-telecom\",\"https://api.llm-stats.com/leaderboard/benchmarks/tau2-telecom\"]", + "raw_score_id": "tau2-telecom::claude-opus-4-6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2::terminal-bench-2-claude-opus-4-6", + "evaluation_name": "llm_stats.terminal-bench-2", + "source_data": { + "dataset_name": "Terminal-Bench 2.0", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-6", + "https://llm-stats.com/benchmarks/terminal-bench-2", + "https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2" + ], + "additional_details": { + "raw_benchmark_id": "terminal-bench-2", + "raw_model_id": "claude-opus-4-6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Terminal-Bench 2.0 is an updated benchmark for testing AI agents' tool use ability to operate a computer via terminal. It evaluates how well models can handle real-world, end-to-end tasks autonomously, including compiling code, training models, setting up servers, system administration, security tasks, data science workflows, and cybersecurity vulnerabilities.", + "metric_id": "llm_stats.terminal-bench-2.score", + "metric_name": "Terminal-Bench 2.0 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "terminal-bench-2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Terminal-Bench 2.0", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "49" + } + }, + "score_details": { + "score": 0.654, + "details": { + "raw_score": "0.654", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-6", + "raw_benchmark_id": "terminal-bench-2", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-6\",\"https://llm-stats.com/benchmarks/terminal-bench-2\",\"https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2\"]", + "raw_score_id": "terminal-bench-2::claude-opus-4-6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "vending-bench-2::vending-bench-2-claude-opus-4-6", + "evaluation_name": "llm_stats.vending-bench-2", + "source_data": { + "dataset_name": "Vending-Bench 2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-6", + "https://llm-stats.com/benchmarks/vending-bench-2", + "https://api.llm-stats.com/leaderboard/benchmarks/vending-bench-2" + ], + "additional_details": { + "raw_benchmark_id": "vending-bench-2", + "raw_model_id": "claude-opus-4-6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Vending-Bench 2 tests longer horizon planning capabilities by evaluating how well AI models can manage a simulated vending machine business over extended periods. The benchmark measures a model's ability to maintain consistent tool usage and decision-making for a full simulated year of operation, driving higher returns without drifting off task.", + "metric_id": "llm_stats.vending-bench-2.score", + "metric_name": "Vending-Bench 2 score", + "metric_kind": "benchmark_score", + "metric_unit": "points", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 8017.59, + "additional_details": { + "raw_benchmark_id": "vending-bench-2", + "raw_score_field": "score", + "bound_strategy": "observed_value", + "raw_name": "Vending-Bench 2", + "raw_categories": "[\"reasoning\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 8017.59, + "details": { + "raw_score": "8017.59", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-6", + "raw_benchmark_id": "vending-bench-2", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-6\",\"https://llm-stats.com/benchmarks/vending-bench-2\",\"https://api.llm-stats.com/leaderboard/benchmarks/vending-bench-2\"]", + "raw_score_id": "vending-bench-2::claude-opus-4-6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/anthropic/claude-opus-4-7/0a60815d-8aa1-409c-a8f2-54c28e7a8107.json b/data/llm_stats/anthropic/claude-opus-4-7/0a60815d-8aa1-409c-a8f2-54c28e7a8107.json new file mode 100644 index 0000000000000000000000000000000000000000..00c482179e43d5cdd5091cf3722da0cebe378fbe --- /dev/null +++ b/data/llm_stats/anthropic/claude-opus-4-7/0a60815d-8aa1-409c-a8f2-54c28e7a8107.json @@ -0,0 +1,672 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/anthropic_claude-opus-4-7/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-7", + "id": "anthropic/claude-opus-4-7", + "developer": "anthropic", + "additional_details": { + "raw_id": "claude-opus-4-7", + "raw_model_name": "claude-opus-4-7", + "raw_display_name": "Claude Opus 4.7", + "raw_organization_id": "anthropic", + "raw_organization_name": "Anthropic", + "raw_context_length": "1000000", + "raw_input_modalities": "[\"image\",\"text\"]", + "raw_output_modalities": "[\"text\"]", + "raw_input_price": "5.0", + "raw_output_price": "25.0", + "raw_model_id": "claude-opus-4-7", + "raw_provider_slug": "anthropic", + "raw_provider_name": "ZeroEval" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "browsecomp::browsecomp-claude-opus-4-7", + "evaluation_name": "llm_stats.browsecomp", + "source_data": { + "dataset_name": "BrowseComp", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-7", + "https://llm-stats.com/benchmarks/browsecomp", + "https://api.llm-stats.com/leaderboard/benchmarks/browsecomp" + ], + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_model_id": "claude-opus-4-7", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BrowseComp is a benchmark comprising 1,266 questions that challenge AI agents to persistently navigate the internet in search of hard-to-find, entangled information. The benchmark measures agents' ability to exercise persistence in information gathering, demonstrate creativity in web navigation, and find concise, verifiable answers. Despite the difficulty of the questions, BrowseComp is simple and easy-to-use, as predicted answers are short and easily verifiable against reference answers.", + "metric_id": "llm_stats.browsecomp.score", + "metric_name": "BrowseComp score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BrowseComp", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "52" + } + }, + "score_details": { + "score": 0.793, + "details": { + "raw_score": "0.793", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-7", + "raw_benchmark_id": "browsecomp", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-7\",\"https://llm-stats.com/benchmarks/browsecomp\",\"https://api.llm-stats.com/leaderboard/benchmarks/browsecomp\"]", + "raw_score_id": "browsecomp::claude-opus-4-7", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "charxiv-r::charxiv-r-claude-opus-4-7", + "evaluation_name": "llm_stats.charxiv-r", + "source_data": { + "dataset_name": "CharXiv-R", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-7", + "https://llm-stats.com/benchmarks/charxiv-r", + "https://api.llm-stats.com/leaderboard/benchmarks/charxiv-r" + ], + "additional_details": { + "raw_benchmark_id": "charxiv-r", + "raw_model_id": "claude-opus-4-7", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CharXiv-R is the reasoning component of the CharXiv benchmark, focusing on complex reasoning questions that require synthesizing information across visual chart elements. It evaluates multimodal large language models on their ability to understand and reason about scientific charts from arXiv papers through various reasoning tasks.", + "metric_id": "llm_stats.charxiv-r.score", + "metric_name": "CharXiv-R score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "charxiv-r", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CharXiv-R", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "43" + } + }, + "score_details": { + "score": 0.91, + "details": { + "raw_score": "0.91", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-7", + "raw_benchmark_id": "charxiv-r", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-7\",\"https://llm-stats.com/benchmarks/charxiv-r\",\"https://api.llm-stats.com/leaderboard/benchmarks/charxiv-r\"]", + "raw_score_id": "charxiv-r::claude-opus-4-7", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "cybergym::cybergym-claude-opus-4-7", + "evaluation_name": "llm_stats.cybergym", + "source_data": { + "dataset_name": "CyberGym", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-7", + "https://llm-stats.com/benchmarks/cybergym", + "https://api.llm-stats.com/leaderboard/benchmarks/cybergym" + ], + "additional_details": { + "raw_benchmark_id": "cybergym", + "raw_model_id": "claude-opus-4-7", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CyberGym is a benchmark for evaluating AI agents on cybersecurity tasks, testing their ability to identify vulnerabilities, perform security analysis, and complete security-related challenges in a controlled environment.", + "metric_id": "llm_stats.cybergym.score", + "metric_name": "CyberGym score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "cybergym", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CyberGym", + "raw_categories": "[\"safety\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.731, + "details": { + "raw_score": "0.731", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-7", + "raw_benchmark_id": "cybergym", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-7\",\"https://llm-stats.com/benchmarks/cybergym\",\"https://api.llm-stats.com/leaderboard/benchmarks/cybergym\"]", + "raw_score_id": "cybergym::claude-opus-4-7", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "finance-agent::finance-agent-claude-opus-4-7", + "evaluation_name": "llm_stats.finance-agent", + "source_data": { + "dataset_name": "Finance Agent", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-7", + "https://llm-stats.com/benchmarks/finance-agent", + "https://api.llm-stats.com/leaderboard/benchmarks/finance-agent" + ], + "additional_details": { + "raw_benchmark_id": "finance-agent", + "raw_model_id": "claude-opus-4-7", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Finance Agent is a benchmark for evaluating AI models on agentic financial analysis tasks, testing their ability to process financial data, perform calculations, and generate accurate analyses across various financial domains.", + "metric_id": "llm_stats.finance-agent.score", + "metric_name": "Finance Agent score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "finance-agent", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Finance Agent", + "raw_categories": "[\"reasoning\",\"finance\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.644, + "details": { + "raw_score": "0.644", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-7", + "raw_benchmark_id": "finance-agent", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-7\",\"https://llm-stats.com/benchmarks/finance-agent\",\"https://api.llm-stats.com/leaderboard/benchmarks/finance-agent\"]", + "raw_score_id": "finance-agent::claude-opus-4-7", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gpqa::gpqa-claude-opus-4-7", + "evaluation_name": "llm_stats.gpqa", + "source_data": { + "dataset_name": "GPQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-7", + "https://llm-stats.com/benchmarks/gpqa", + "https://api.llm-stats.com/leaderboard/benchmarks/gpqa" + ], + "additional_details": { + "raw_benchmark_id": "gpqa", + "raw_model_id": "claude-opus-4-7", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A challenging dataset of 448 multiple-choice questions written by domain experts in biology, physics, and chemistry. Questions are Google-proof and extremely difficult, with PhD experts reaching 65% accuracy.", + "metric_id": "llm_stats.gpqa.score", + "metric_name": "GPQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gpqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GPQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\",\"biology\",\"chemistry\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "226" + } + }, + "score_details": { + "score": 0.942, + "details": { + "raw_score": "0.942", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-7", + "raw_benchmark_id": "gpqa", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-7\",\"https://llm-stats.com/benchmarks/gpqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/gpqa\"]", + "raw_score_id": "gpqa::claude-opus-4-7", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "humanity-s-last-exam::humanity-s-last-exam-claude-opus-4-7", + "evaluation_name": "llm_stats.humanity-s-last-exam", + "source_data": { + "dataset_name": "Humanity's Last Exam", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-7", + "https://llm-stats.com/benchmarks/humanity-s-last-exam", + "https://api.llm-stats.com/leaderboard/benchmarks/humanity's-last-exam" + ], + "additional_details": { + "raw_benchmark_id": "humanity's-last-exam", + "raw_model_id": "claude-opus-4-7", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Humanity's Last Exam (HLE) is a multi-modal academic benchmark with 2,500 questions across mathematics, humanities, and natural sciences, designed to test LLM capabilities at the frontier of human knowledge with unambiguous, verifiable solutions", + "metric_id": "llm_stats.humanity-s-last-exam.score", + "metric_name": "Humanity's Last Exam score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "humanity's-last-exam", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Humanity's Last Exam", + "raw_categories": "[\"math\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "88" + } + }, + "score_details": { + "score": 0.547, + "details": { + "raw_score": "0.547", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-7", + "raw_benchmark_id": "humanity's-last-exam", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-7\",\"https://llm-stats.com/benchmarks/humanity-s-last-exam\",\"https://api.llm-stats.com/leaderboard/benchmarks/humanity's-last-exam\"]", + "raw_score_id": "humanity's-last-exam::claude-opus-4-7", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mcp-atlas::mcp-atlas-claude-opus-4-7", + "evaluation_name": "llm_stats.mcp-atlas", + "source_data": { + "dataset_name": "MCP Atlas", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-7", + "https://llm-stats.com/benchmarks/mcp-atlas", + "https://api.llm-stats.com/leaderboard/benchmarks/mcp-atlas" + ], + "additional_details": { + "raw_benchmark_id": "mcp-atlas", + "raw_model_id": "claude-opus-4-7", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MCP Atlas is a benchmark for evaluating AI models on scaled tool use capabilities, measuring how well models can coordinate and utilize multiple tools across complex multi-step tasks.", + "metric_id": "llm_stats.mcp-atlas.score", + "metric_name": "MCP Atlas score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mcp-atlas", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MCP Atlas", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.773, + "details": { + "raw_score": "0.773", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-7", + "raw_benchmark_id": "mcp-atlas", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-7\",\"https://llm-stats.com/benchmarks/mcp-atlas\",\"https://api.llm-stats.com/leaderboard/benchmarks/mcp-atlas\"]", + "raw_score_id": "mcp-atlas::claude-opus-4-7", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmmlu::mmmlu-claude-opus-4-7", + "evaluation_name": "llm_stats.mmmlu", + "source_data": { + "dataset_name": "MMMLU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-7", + "https://llm-stats.com/benchmarks/mmmlu", + "https://api.llm-stats.com/leaderboard/benchmarks/mmmlu" + ], + "additional_details": { + "raw_benchmark_id": "mmmlu", + "raw_model_id": "claude-opus-4-7", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Multilingual Massive Multitask Language Understanding dataset released by OpenAI, featuring professionally translated MMLU test questions across 14 languages including Arabic, Bengali, German, Spanish, French, Hindi, Indonesian, Italian, Japanese, Korean, Portuguese, Swahili, Yoruba, and Chinese. Contains approximately 15,908 multiple-choice questions per language covering 57 subjects.", + "metric_id": "llm_stats.mmmlu.score", + "metric_name": "MMMLU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmmlu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMMLU", + "raw_categories": "[\"language\",\"math\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "49" + } + }, + "score_details": { + "score": 0.915, + "details": { + "raw_score": "0.915", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-7", + "raw_benchmark_id": "mmmlu", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-7\",\"https://llm-stats.com/benchmarks/mmmlu\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmmlu\"]", + "raw_score_id": "mmmlu::claude-opus-4-7", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "osworld-verified::osworld-verified-claude-opus-4-7", + "evaluation_name": "llm_stats.osworld-verified", + "source_data": { + "dataset_name": "OSWorld-Verified", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-7", + "https://llm-stats.com/benchmarks/osworld-verified", + "https://api.llm-stats.com/leaderboard/benchmarks/osworld-verified" + ], + "additional_details": { + "raw_benchmark_id": "osworld-verified", + "raw_model_id": "claude-opus-4-7", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OSWorld-Verified is a verified subset of OSWorld, a scalable real computer environment for multimodal agents supporting task setup, execution-based evaluation, and interactive learning across Ubuntu, Windows, and macOS.", + "metric_id": "llm_stats.osworld-verified.score", + "metric_name": "OSWorld-Verified score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "osworld-verified", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OSWorld-Verified", + "raw_categories": "[\"multimodal\",\"general\",\"agents\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.78, + "details": { + "raw_score": "0.78", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-7", + "raw_benchmark_id": "osworld-verified", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-7\",\"https://llm-stats.com/benchmarks/osworld-verified\",\"https://api.llm-stats.com/leaderboard/benchmarks/osworld-verified\"]", + "raw_score_id": "osworld-verified::claude-opus-4-7", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-pro::swe-bench-pro-claude-opus-4-7", + "evaluation_name": "llm_stats.swe-bench-pro", + "source_data": { + "dataset_name": "SWE-Bench Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-7", + "https://llm-stats.com/benchmarks/swe-bench-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-pro" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-pro", + "raw_model_id": "claude-opus-4-7", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SWE-Bench Pro is an advanced version of SWE-Bench that evaluates language models on complex, real-world software engineering tasks requiring extended reasoning and multi-step problem solving.", + "metric_id": "llm_stats.swe-bench-pro.score", + "metric_name": "SWE-Bench Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Bench Pro", + "raw_categories": "[\"reasoning\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "35" + } + }, + "score_details": { + "score": 0.643, + "details": { + "raw_score": "0.643", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-7", + "raw_benchmark_id": "swe-bench-pro", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-7\",\"https://llm-stats.com/benchmarks/swe-bench-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-pro\"]", + "raw_score_id": "swe-bench-pro::claude-opus-4-7", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-verified::swe-bench-verified-claude-opus-4-7", + "evaluation_name": "llm_stats.swe-bench-verified", + "source_data": { + "dataset_name": "SWE-Bench Verified", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-7", + "https://llm-stats.com/benchmarks/swe-bench-verified", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-verified" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-verified", + "raw_model_id": "claude-opus-4-7", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A verified subset of 500 software engineering problems from real GitHub issues, validated by human annotators for evaluating language models' ability to resolve real-world coding issues by generating patches for Python codebases.", + "metric_id": "llm_stats.swe-bench-verified.score", + "metric_name": "SWE-Bench Verified score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-verified", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Bench Verified", + "raw_categories": "[\"reasoning\",\"frontend_development\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "103" + } + }, + "score_details": { + "score": 0.876, + "details": { + "raw_score": "0.876", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-7", + "raw_benchmark_id": "swe-bench-verified", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-7\",\"https://llm-stats.com/benchmarks/swe-bench-verified\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-verified\"]", + "raw_score_id": "swe-bench-verified::claude-opus-4-7", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2::terminal-bench-2-claude-opus-4-7", + "evaluation_name": "llm_stats.terminal-bench-2", + "source_data": { + "dataset_name": "Terminal-Bench 2.0", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-7", + "https://llm-stats.com/benchmarks/terminal-bench-2", + "https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2" + ], + "additional_details": { + "raw_benchmark_id": "terminal-bench-2", + "raw_model_id": "claude-opus-4-7", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Terminal-Bench 2.0 is an updated benchmark for testing AI agents' tool use ability to operate a computer via terminal. It evaluates how well models can handle real-world, end-to-end tasks autonomously, including compiling code, training models, setting up servers, system administration, security tasks, data science workflows, and cybersecurity vulnerabilities.", + "metric_id": "llm_stats.terminal-bench-2.score", + "metric_name": "Terminal-Bench 2.0 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "terminal-bench-2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Terminal-Bench 2.0", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "49" + } + }, + "score_details": { + "score": 0.694, + "details": { + "raw_score": "0.694", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-7", + "raw_benchmark_id": "terminal-bench-2", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-7\",\"https://llm-stats.com/benchmarks/terminal-bench-2\",\"https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2\"]", + "raw_score_id": "terminal-bench-2::claude-opus-4-7", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/anthropic/claude-opus-4-8/d862f24e-98f1-4535-8744-2926c9d82064.json b/data/llm_stats/anthropic/claude-opus-4-8/d862f24e-98f1-4535-8744-2926c9d82064.json new file mode 100644 index 0000000000000000000000000000000000000000..8a522c6571f345faae5ce5591d9c56f2a577b443 --- /dev/null +++ b/data/llm_stats/anthropic/claude-opus-4-8/d862f24e-98f1-4535-8744-2926c9d82064.json @@ -0,0 +1,1140 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/anthropic_claude-opus-4-8/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-8", + "id": "anthropic/claude-opus-4-8", + "developer": "anthropic", + "additional_details": { + "raw_id": "claude-opus-4-8", + "raw_model_name": "claude-opus-4-8", + "raw_display_name": "Claude Opus 4.8", + "raw_organization_id": "anthropic", + "raw_organization_name": "Anthropic", + "raw_context_length": "1000000", + "raw_input_modalities": "[\"image\",\"text\"]", + "raw_output_modalities": "[\"text\"]", + "raw_input_price": "5.0", + "raw_output_price": "25.0", + "raw_model_id": "claude-opus-4-8", + "raw_provider_slug": "anthropic", + "raw_provider_name": "ZeroEval" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "browsecomp::browsecomp-claude-opus-4-8", + "evaluation_name": "llm_stats.browsecomp", + "source_data": { + "dataset_name": "BrowseComp", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-8", + "https://llm-stats.com/benchmarks/browsecomp", + "https://api.llm-stats.com/leaderboard/benchmarks/browsecomp" + ], + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_model_id": "claude-opus-4-8", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BrowseComp is a benchmark comprising 1,266 questions that challenge AI agents to persistently navigate the internet in search of hard-to-find, entangled information. The benchmark measures agents' ability to exercise persistence in information gathering, demonstrate creativity in web navigation, and find concise, verifiable answers. Despite the difficulty of the questions, BrowseComp is simple and easy-to-use, as predicted answers are short and easily verifiable against reference answers.", + "metric_id": "llm_stats.browsecomp.score", + "metric_name": "BrowseComp score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BrowseComp", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "52" + } + }, + "score_details": { + "score": 0.843, + "details": { + "raw_score": "0.843", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-8", + "raw_benchmark_id": "browsecomp", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-8\",\"https://llm-stats.com/benchmarks/browsecomp\",\"https://api.llm-stats.com/leaderboard/benchmarks/browsecomp\"]", + "raw_score_id": "browsecomp::claude-opus-4-8", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "charxiv-r::charxiv-r-claude-opus-4-8", + "evaluation_name": "llm_stats.charxiv-r", + "source_data": { + "dataset_name": "CharXiv-R", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-8", + "https://llm-stats.com/benchmarks/charxiv-r", + "https://api.llm-stats.com/leaderboard/benchmarks/charxiv-r" + ], + "additional_details": { + "raw_benchmark_id": "charxiv-r", + "raw_model_id": "claude-opus-4-8", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CharXiv-R is the reasoning component of the CharXiv benchmark, focusing on complex reasoning questions that require synthesizing information across visual chart elements. It evaluates multimodal large language models on their ability to understand and reason about scientific charts from arXiv papers through various reasoning tasks.", + "metric_id": "llm_stats.charxiv-r.score", + "metric_name": "CharXiv-R score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "charxiv-r", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CharXiv-R", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "43" + } + }, + "score_details": { + "score": 0.899, + "details": { + "raw_score": "0.899", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-8", + "raw_benchmark_id": "charxiv-r", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-8\",\"https://llm-stats.com/benchmarks/charxiv-r\",\"https://api.llm-stats.com/leaderboard/benchmarks/charxiv-r\"]", + "raw_score_id": "charxiv-r::claude-opus-4-8", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "cybergym::cybergym-claude-opus-4-8", + "evaluation_name": "llm_stats.cybergym", + "source_data": { + "dataset_name": "CyberGym", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-8", + "https://llm-stats.com/benchmarks/cybergym", + "https://api.llm-stats.com/leaderboard/benchmarks/cybergym" + ], + "additional_details": { + "raw_benchmark_id": "cybergym", + "raw_model_id": "claude-opus-4-8", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CyberGym is a benchmark for evaluating AI agents on cybersecurity tasks, testing their ability to identify vulnerabilities, perform security analysis, and complete security-related challenges in a controlled environment.", + "metric_id": "llm_stats.cybergym.score", + "metric_name": "CyberGym score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "cybergym", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CyberGym", + "raw_categories": "[\"safety\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.788, + "details": { + "raw_score": "0.788", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-8", + "raw_benchmark_id": "cybergym", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-8\",\"https://llm-stats.com/benchmarks/cybergym\",\"https://api.llm-stats.com/leaderboard/benchmarks/cybergym\"]", + "raw_score_id": "cybergym::claude-opus-4-8", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "deepsearchqa::deepsearchqa-claude-opus-4-8", + "evaluation_name": "llm_stats.deepsearchqa", + "source_data": { + "dataset_name": "DeepSearchQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-8", + "https://llm-stats.com/benchmarks/deepsearchqa", + "https://api.llm-stats.com/leaderboard/benchmarks/deepsearchqa" + ], + "additional_details": { + "raw_benchmark_id": "deepsearchqa", + "raw_model_id": "claude-opus-4-8", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "DeepSearchQA is a benchmark for evaluating deep search and question-answering capabilities, testing models' ability to perform multi-hop reasoning and information retrieval across complex knowledge domains.", + "metric_id": "llm_stats.deepsearchqa.score", + "metric_name": "DeepSearchQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "deepsearchqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DeepSearchQA", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.931, + "details": { + "raw_score": "0.931", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-8", + "raw_benchmark_id": "deepsearchqa", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-8\",\"https://llm-stats.com/benchmarks/deepsearchqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/deepsearchqa\"]", + "raw_score_id": "deepsearchqa::claude-opus-4-8", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "finance-agent::finance-agent-claude-opus-4-8", + "evaluation_name": "llm_stats.finance-agent", + "source_data": { + "dataset_name": "Finance Agent", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-8", + "https://llm-stats.com/benchmarks/finance-agent", + "https://api.llm-stats.com/leaderboard/benchmarks/finance-agent" + ], + "additional_details": { + "raw_benchmark_id": "finance-agent", + "raw_model_id": "claude-opus-4-8", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Finance Agent is a benchmark for evaluating AI models on agentic financial analysis tasks, testing their ability to process financial data, perform calculations, and generate accurate analyses across various financial domains.", + "metric_id": "llm_stats.finance-agent.score", + "metric_name": "Finance Agent score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "finance-agent", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Finance Agent", + "raw_categories": "[\"reasoning\",\"finance\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.539, + "details": { + "raw_score": "0.539", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-8", + "raw_benchmark_id": "finance-agent", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-8\",\"https://llm-stats.com/benchmarks/finance-agent\",\"https://api.llm-stats.com/leaderboard/benchmarks/finance-agent\"]", + "raw_score_id": "finance-agent::claude-opus-4-8", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gpqa::gpqa-claude-opus-4-8", + "evaluation_name": "llm_stats.gpqa", + "source_data": { + "dataset_name": "GPQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-8", + "https://llm-stats.com/benchmarks/gpqa", + "https://api.llm-stats.com/leaderboard/benchmarks/gpqa" + ], + "additional_details": { + "raw_benchmark_id": "gpqa", + "raw_model_id": "claude-opus-4-8", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A challenging dataset of 448 multiple-choice questions written by domain experts in biology, physics, and chemistry. Questions are Google-proof and extremely difficult, with PhD experts reaching 65% accuracy.", + "metric_id": "llm_stats.gpqa.score", + "metric_name": "GPQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gpqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GPQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\",\"biology\",\"chemistry\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "226" + } + }, + "score_details": { + "score": 0.936, + "details": { + "raw_score": "0.936", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-8", + "raw_benchmark_id": "gpqa", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-8\",\"https://llm-stats.com/benchmarks/gpqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/gpqa\"]", + "raw_score_id": "gpqa::claude-opus-4-8", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "graphwalks-bfs-128k::graphwalks-bfs-128k-claude-opus-4-8", + "evaluation_name": "llm_stats.graphwalks-bfs-128k", + "source_data": { + "dataset_name": "Graphwalks BFS >128k", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-8", + "https://llm-stats.com/benchmarks/graphwalks-bfs-128k", + "https://api.llm-stats.com/leaderboard/benchmarks/graphwalks-bfs->128k" + ], + "additional_details": { + "raw_benchmark_id": "graphwalks-bfs->128k", + "raw_model_id": "claude-opus-4-8", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A graph reasoning benchmark that evaluates language models' ability to perform breadth-first search (BFS) operations on graphs with context length over 128k tokens, testing long-context reasoning capabilities.", + "metric_id": "llm_stats.graphwalks-bfs-128k.score", + "metric_name": "Graphwalks BFS >128k score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "graphwalks-bfs->128k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Graphwalks BFS >128k", + "raw_categories": "[\"long_context\",\"reasoning\",\"spatial_reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.681, + "details": { + "raw_score": "0.681", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-8", + "raw_benchmark_id": "graphwalks-bfs->128k", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-8\",\"https://llm-stats.com/benchmarks/graphwalks-bfs-128k\",\"https://api.llm-stats.com/leaderboard/benchmarks/graphwalks-bfs->128k\"]", + "raw_score_id": "graphwalks-bfs->128k::claude-opus-4-8", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "graphwalks-parents-128k::graphwalks-parents-128k-claude-opus-4-8", + "evaluation_name": "llm_stats.graphwalks-parents-128k", + "source_data": { + "dataset_name": "Graphwalks parents >128k", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-8", + "https://llm-stats.com/benchmarks/graphwalks-parents-128k", + "https://api.llm-stats.com/leaderboard/benchmarks/graphwalks-parents->128k" + ], + "additional_details": { + "raw_benchmark_id": "graphwalks-parents->128k", + "raw_model_id": "claude-opus-4-8", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A graph reasoning benchmark that evaluates language models' ability to find parent nodes in graphs with context length over 128k tokens, testing long-context reasoning and graph structure understanding.", + "metric_id": "llm_stats.graphwalks-parents-128k.score", + "metric_name": "Graphwalks parents >128k score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "graphwalks-parents->128k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Graphwalks parents >128k", + "raw_categories": "[\"long_context\",\"reasoning\",\"spatial_reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.833, + "details": { + "raw_score": "0.833", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-8", + "raw_benchmark_id": "graphwalks-parents->128k", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-8\",\"https://llm-stats.com/benchmarks/graphwalks-parents-128k\",\"https://api.llm-stats.com/leaderboard/benchmarks/graphwalks-parents->128k\"]", + "raw_score_id": "graphwalks-parents->128k::claude-opus-4-8", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "healthbench-professional::healthbench-professional-claude-opus-4-8", + "evaluation_name": "llm_stats.healthbench-professional", + "source_data": { + "dataset_name": "HealthBench Professional", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-8", + "https://llm-stats.com/benchmarks/healthbench-professional", + "https://api.llm-stats.com/leaderboard/benchmarks/healthbench-professional" + ], + "additional_details": { + "raw_benchmark_id": "healthbench-professional", + "raw_model_id": "claude-opus-4-8", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "HealthBench Professional evaluates model capability and safety for clinician use cases using real clinician-style chats and physician-authored grading rubrics.", + "metric_id": "llm_stats.healthbench-professional.score", + "metric_name": "HealthBench Professional score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "healthbench-professional", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HealthBench Professional", + "raw_categories": "[\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.558, + "details": { + "raw_score": "0.558", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-8", + "raw_benchmark_id": "healthbench-professional", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-8\",\"https://llm-stats.com/benchmarks/healthbench-professional\",\"https://api.llm-stats.com/leaderboard/benchmarks/healthbench-professional\"]", + "raw_score_id": "healthbench-professional::claude-opus-4-8", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "humanity-s-last-exam::humanity-s-last-exam-claude-opus-4-8", + "evaluation_name": "llm_stats.humanity-s-last-exam", + "source_data": { + "dataset_name": "Humanity's Last Exam", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-8", + "https://llm-stats.com/benchmarks/humanity-s-last-exam", + "https://api.llm-stats.com/leaderboard/benchmarks/humanity's-last-exam" + ], + "additional_details": { + "raw_benchmark_id": "humanity's-last-exam", + "raw_model_id": "claude-opus-4-8", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Humanity's Last Exam (HLE) is a multi-modal academic benchmark with 2,500 questions across mathematics, humanities, and natural sciences, designed to test LLM capabilities at the frontier of human knowledge with unambiguous, verifiable solutions", + "metric_id": "llm_stats.humanity-s-last-exam.score", + "metric_name": "Humanity's Last Exam score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "humanity's-last-exam", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Humanity's Last Exam", + "raw_categories": "[\"math\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "88" + } + }, + "score_details": { + "score": 0.579, + "details": { + "raw_score": "0.579", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-8", + "raw_benchmark_id": "humanity's-last-exam", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-8\",\"https://llm-stats.com/benchmarks/humanity-s-last-exam\",\"https://api.llm-stats.com/leaderboard/benchmarks/humanity's-last-exam\"]", + "raw_score_id": "humanity's-last-exam::claude-opus-4-8", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "include::include-claude-opus-4-8", + "evaluation_name": "llm_stats.include", + "source_data": { + "dataset_name": "Include", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-8", + "https://llm-stats.com/benchmarks/include", + "https://api.llm-stats.com/leaderboard/benchmarks/include" + ], + "additional_details": { + "raw_benchmark_id": "include", + "raw_model_id": "claude-opus-4-8", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Include benchmark - specific documentation not found in official sources", + "metric_id": "llm_stats.include.score", + "metric_name": "Include score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "include", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Include", + "raw_categories": "[\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "31" + } + }, + "score_details": { + "score": 0.876, + "details": { + "raw_score": "0.876", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-8", + "raw_benchmark_id": "include", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-8\",\"https://llm-stats.com/benchmarks/include\",\"https://api.llm-stats.com/leaderboard/benchmarks/include\"]", + "raw_score_id": "include::claude-opus-4-8", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mcp-atlas::mcp-atlas-claude-opus-4-8", + "evaluation_name": "llm_stats.mcp-atlas", + "source_data": { + "dataset_name": "MCP Atlas", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-8", + "https://llm-stats.com/benchmarks/mcp-atlas", + "https://api.llm-stats.com/leaderboard/benchmarks/mcp-atlas" + ], + "additional_details": { + "raw_benchmark_id": "mcp-atlas", + "raw_model_id": "claude-opus-4-8", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MCP Atlas is a benchmark for evaluating AI models on scaled tool use capabilities, measuring how well models can coordinate and utilize multiple tools across complex multi-step tasks.", + "metric_id": "llm_stats.mcp-atlas.score", + "metric_name": "MCP Atlas score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mcp-atlas", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MCP Atlas", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.822, + "details": { + "raw_score": "0.822", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-8", + "raw_benchmark_id": "mcp-atlas", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-8\",\"https://llm-stats.com/benchmarks/mcp-atlas\",\"https://api.llm-stats.com/leaderboard/benchmarks/mcp-atlas\"]", + "raw_score_id": "mcp-atlas::claude-opus-4-8", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "officeqa-pro::officeqa-pro-claude-opus-4-8", + "evaluation_name": "llm_stats.officeqa-pro", + "source_data": { + "dataset_name": "OfficeQA Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-8", + "https://llm-stats.com/benchmarks/officeqa-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/officeqa-pro" + ], + "additional_details": { + "raw_benchmark_id": "officeqa-pro", + "raw_model_id": "claude-opus-4-8", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OfficeQA Pro evaluates AI models on professional knowledge-work questions and tasks drawn from real office workflows, including document analysis, spreadsheet reasoning, and information synthesis across business domains.", + "metric_id": "llm_stats.officeqa-pro.score", + "metric_name": "OfficeQA Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "officeqa-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OfficeQA Pro", + "raw_categories": "[\"reasoning\",\"general\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.662, + "details": { + "raw_score": "0.662", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-8", + "raw_benchmark_id": "officeqa-pro", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-8\",\"https://llm-stats.com/benchmarks/officeqa-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/officeqa-pro\"]", + "raw_score_id": "officeqa-pro::claude-opus-4-8", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "osworld-verified::osworld-verified-claude-opus-4-8", + "evaluation_name": "llm_stats.osworld-verified", + "source_data": { + "dataset_name": "OSWorld-Verified", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-8", + "https://llm-stats.com/benchmarks/osworld-verified", + "https://api.llm-stats.com/leaderboard/benchmarks/osworld-verified" + ], + "additional_details": { + "raw_benchmark_id": "osworld-verified", + "raw_model_id": "claude-opus-4-8", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OSWorld-Verified is a verified subset of OSWorld, a scalable real computer environment for multimodal agents supporting task setup, execution-based evaluation, and interactive learning across Ubuntu, Windows, and macOS.", + "metric_id": "llm_stats.osworld-verified.score", + "metric_name": "OSWorld-Verified score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "osworld-verified", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OSWorld-Verified", + "raw_categories": "[\"multimodal\",\"general\",\"agents\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.834, + "details": { + "raw_score": "0.834", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-8", + "raw_benchmark_id": "osworld-verified", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-8\",\"https://llm-stats.com/benchmarks/osworld-verified\",\"https://api.llm-stats.com/leaderboard/benchmarks/osworld-verified\"]", + "raw_score_id": "osworld-verified::claude-opus-4-8", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "screenspot-pro::screenspot-pro-claude-opus-4-8", + "evaluation_name": "llm_stats.screenspot-pro", + "source_data": { + "dataset_name": "ScreenSpot Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-8", + "https://llm-stats.com/benchmarks/screenspot-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/screenspot-pro" + ], + "additional_details": { + "raw_benchmark_id": "screenspot-pro", + "raw_model_id": "claude-opus-4-8", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ScreenSpot-Pro is a novel GUI grounding benchmark designed to rigorously evaluate the grounding capabilities of multimodal large language models (MLLMs) in professional high-resolution computing environments. The benchmark comprises 1,581 instructions across 23 applications spanning 5 industries and 3 operating systems, featuring authentic high-resolution images from professional domains with expert annotations. Unlike previous benchmarks that focus on cropped screenshots in consumer applications, ScreenSpot-Pro addresses the complexity and diversity of real-world professional software scenarios, revealing significant performance gaps in current MLLM GUI perception capabilities.", + "metric_id": "llm_stats.screenspot-pro.score", + "metric_name": "ScreenSpot Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "screenspot-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ScreenSpot Pro", + "raw_categories": "[\"multimodal\",\"spatial_reasoning\",\"grounding\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.879, + "details": { + "raw_score": "0.879", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-8", + "raw_benchmark_id": "screenspot-pro", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-8\",\"https://llm-stats.com/benchmarks/screenspot-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/screenspot-pro\"]", + "raw_score_id": "screenspot-pro::claude-opus-4-8", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-multilingual::swe-bench-multilingual-claude-opus-4-8", + "evaluation_name": "llm_stats.swe-bench-multilingual", + "source_data": { + "dataset_name": "SWE-bench Multilingual", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-8", + "https://llm-stats.com/benchmarks/swe-bench-multilingual", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-multilingual" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-multilingual", + "raw_model_id": "claude-opus-4-8", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multilingual benchmark for issue resolving in software engineering that covers Java, TypeScript, JavaScript, Go, Rust, C, and C++. Contains 1,632 high-quality instances carefully annotated from 2,456 candidates by 68 expert annotators, designed to evaluate Large Language Models across diverse software ecosystems beyond Python.", + "metric_id": "llm_stats.swe-bench-multilingual.score", + "metric_name": "SWE-bench Multilingual score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-multilingual", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-bench Multilingual", + "raw_categories": "[\"reasoning\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "33" + } + }, + "score_details": { + "score": 0.844, + "details": { + "raw_score": "0.844", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-8", + "raw_benchmark_id": "swe-bench-multilingual", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-8\",\"https://llm-stats.com/benchmarks/swe-bench-multilingual\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-multilingual\"]", + "raw_score_id": "swe-bench-multilingual::claude-opus-4-8", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-multimodal::swe-bench-multimodal-claude-opus-4-8", + "evaluation_name": "llm_stats.swe-bench-multimodal", + "source_data": { + "dataset_name": "SWE-Bench Multimodal", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-8", + "https://llm-stats.com/benchmarks/swe-bench-multimodal", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-multimodal" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-multimodal", + "raw_model_id": "claude-opus-4-8", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SWE-Bench Multimodal extends SWE-Bench to evaluate language models on software engineering tasks that involve visual inputs such as screenshots, UI mockups, and diagrams alongside code understanding.", + "metric_id": "llm_stats.swe-bench-multimodal.score", + "metric_name": "SWE-Bench Multimodal score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-multimodal", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Bench Multimodal", + "raw_categories": "[\"multimodal\",\"reasoning\",\"agents\",\"code\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.384, + "details": { + "raw_score": "0.384", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-8", + "raw_benchmark_id": "swe-bench-multimodal", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-8\",\"https://llm-stats.com/benchmarks/swe-bench-multimodal\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-multimodal\"]", + "raw_score_id": "swe-bench-multimodal::claude-opus-4-8", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-pro::swe-bench-pro-claude-opus-4-8", + "evaluation_name": "llm_stats.swe-bench-pro", + "source_data": { + "dataset_name": "SWE-Bench Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-8", + "https://llm-stats.com/benchmarks/swe-bench-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-pro" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-pro", + "raw_model_id": "claude-opus-4-8", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SWE-Bench Pro is an advanced version of SWE-Bench that evaluates language models on complex, real-world software engineering tasks requiring extended reasoning and multi-step problem solving.", + "metric_id": "llm_stats.swe-bench-pro.score", + "metric_name": "SWE-Bench Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Bench Pro", + "raw_categories": "[\"reasoning\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "35" + } + }, + "score_details": { + "score": 0.692, + "details": { + "raw_score": "0.692", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-8", + "raw_benchmark_id": "swe-bench-pro", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-8\",\"https://llm-stats.com/benchmarks/swe-bench-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-pro\"]", + "raw_score_id": "swe-bench-pro::claude-opus-4-8", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-verified::swe-bench-verified-claude-opus-4-8", + "evaluation_name": "llm_stats.swe-bench-verified", + "source_data": { + "dataset_name": "SWE-Bench Verified", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-8", + "https://llm-stats.com/benchmarks/swe-bench-verified", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-verified" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-verified", + "raw_model_id": "claude-opus-4-8", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A verified subset of 500 software engineering problems from real GitHub issues, validated by human annotators for evaluating language models' ability to resolve real-world coding issues by generating patches for Python codebases.", + "metric_id": "llm_stats.swe-bench-verified.score", + "metric_name": "SWE-Bench Verified score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-verified", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Bench Verified", + "raw_categories": "[\"reasoning\",\"frontend_development\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "103" + } + }, + "score_details": { + "score": 0.886, + "details": { + "raw_score": "0.886", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-8", + "raw_benchmark_id": "swe-bench-verified", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-8\",\"https://llm-stats.com/benchmarks/swe-bench-verified\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-verified\"]", + "raw_score_id": "swe-bench-verified::claude-opus-4-8", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2::terminal-bench-2-claude-opus-4-8", + "evaluation_name": "llm_stats.terminal-bench-2", + "source_data": { + "dataset_name": "Terminal-Bench 2.0", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-8", + "https://llm-stats.com/benchmarks/terminal-bench-2", + "https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2" + ], + "additional_details": { + "raw_benchmark_id": "terminal-bench-2", + "raw_model_id": "claude-opus-4-8", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Terminal-Bench 2.0 is an updated benchmark for testing AI agents' tool use ability to operate a computer via terminal. It evaluates how well models can handle real-world, end-to-end tasks autonomously, including compiling code, training models, setting up servers, system administration, security tasks, data science workflows, and cybersecurity vulnerabilities.", + "metric_id": "llm_stats.terminal-bench-2.score", + "metric_name": "Terminal-Bench 2.0 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "terminal-bench-2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Terminal-Bench 2.0", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "49" + } + }, + "score_details": { + "score": 0.746, + "details": { + "raw_score": "0.746", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-8", + "raw_benchmark_id": "terminal-bench-2", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-8\",\"https://llm-stats.com/benchmarks/terminal-bench-2\",\"https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2\"]", + "raw_score_id": "terminal-bench-2::claude-opus-4-8", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "toolathlon::toolathlon-claude-opus-4-8", + "evaluation_name": "llm_stats.toolathlon", + "source_data": { + "dataset_name": "Toolathlon", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-opus-4-8", + "https://llm-stats.com/benchmarks/toolathlon", + "https://api.llm-stats.com/leaderboard/benchmarks/toolathlon" + ], + "additional_details": { + "raw_benchmark_id": "toolathlon", + "raw_model_id": "claude-opus-4-8", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Tool Decathlon is a comprehensive benchmark for evaluating AI agents' ability to use multiple tools across diverse task categories. It measures proficiency in tool selection, sequencing, and execution across ten different tool-use scenarios.", + "metric_id": "llm_stats.toolathlon.score", + "metric_name": "Toolathlon score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "toolathlon", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Toolathlon", + "raw_categories": "[\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "24" + } + }, + "score_details": { + "score": 0.599, + "details": { + "raw_score": "0.599", + "raw_score_field": "score", + "raw_model_id": "claude-opus-4-8", + "raw_benchmark_id": "toolathlon", + "source_urls_json": "[\"https://llm-stats.com/models/claude-opus-4-8\",\"https://llm-stats.com/benchmarks/toolathlon\",\"https://api.llm-stats.com/leaderboard/benchmarks/toolathlon\"]", + "raw_score_id": "toolathlon::claude-opus-4-8", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/anthropic/claude-sonnet-5/558cef4c-fe80-4578-9551-9b669a489102.json b/data/llm_stats/anthropic/claude-sonnet-5/558cef4c-fe80-4578-9551-9b669a489102.json new file mode 100644 index 0000000000000000000000000000000000000000..60176c9a2b6ca735b3ac1b605161aaf96363e418 --- /dev/null +++ b/data/llm_stats/anthropic/claude-sonnet-5/558cef4c-fe80-4578-9551-9b669a489102.json @@ -0,0 +1,1088 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/anthropic_claude-sonnet-5/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-5", + "id": "anthropic/claude-sonnet-5", + "developer": "anthropic", + "additional_details": { + "raw_id": "claude-sonnet-5", + "raw_model_name": "claude-sonnet-5", + "raw_display_name": "Claude Sonnet 5", + "raw_organization_id": "anthropic", + "raw_organization_name": "Anthropic", + "raw_context_length": "1000000", + "raw_input_modalities": "[\"image\",\"text\"]", + "raw_output_modalities": "[\"text\"]", + "raw_input_price": "3.0", + "raw_output_price": "15.0", + "raw_model_id": "claude-sonnet-5", + "raw_provider_slug": "anthropic", + "raw_provider_name": "ZeroEval" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "arxivmath::arxivmath-claude-sonnet-5", + "evaluation_name": "llm_stats.arxivmath", + "source_data": { + "dataset_name": "ArXivMath", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-sonnet-5", + "https://llm-stats.com/benchmarks/arxivmath", + "https://api.llm-stats.com/leaderboard/benchmarks/arxivmath" + ], + "additional_details": { + "raw_benchmark_id": "arxivmath", + "raw_model_id": "claude-sonnet-5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ArXivMath is a final-answer benchmark of research-level mathematics maintained by MathArena. Problems are extracted monthly from recent arXiv paper abstracts, then filtered through automated and manual checks to ensure they are self-contained, non-trivial, and verifiable. Because problems are drawn from active research, the benchmark is more realistic and more closely connected to mathematical research than contest or olympiad benchmarks.", + "metric_id": "llm_stats.arxivmath.score", + "metric_name": "ArXivMath score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arxivmath", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ArXivMath", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.722, + "details": { + "raw_score": "0.722", + "raw_score_field": "score", + "raw_model_id": "claude-sonnet-5", + "raw_benchmark_id": "arxivmath", + "source_urls_json": "[\"https://llm-stats.com/models/claude-sonnet-5\",\"https://llm-stats.com/benchmarks/arxivmath\",\"https://api.llm-stats.com/leaderboard/benchmarks/arxivmath\"]", + "raw_score_id": "arxivmath::claude-sonnet-5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "automationbench::automationbench-claude-sonnet-5", + "evaluation_name": "llm_stats.automationbench", + "source_data": { + "dataset_name": "AutomationBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-sonnet-5", + "https://llm-stats.com/benchmarks/automationbench", + "https://api.llm-stats.com/leaderboard/benchmarks/automationbench" + ], + "additional_details": { + "raw_benchmark_id": "automationbench", + "raw_model_id": "claude-sonnet-5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "AutomationBench is a tool-use benchmark that evaluates AI agents on automating real-world workflows, testing their ability to orchestrate tools and complete multi-step automation tasks.", + "metric_id": "llm_stats.automationbench.score", + "metric_name": "AutomationBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "automationbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AutomationBench", + "raw_categories": "[\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.135, + "details": { + "raw_score": "0.135", + "raw_score_field": "score", + "raw_model_id": "claude-sonnet-5", + "raw_benchmark_id": "automationbench", + "source_urls_json": "[\"https://llm-stats.com/models/claude-sonnet-5\",\"https://llm-stats.com/benchmarks/automationbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/automationbench\"]", + "raw_score_id": "automationbench::claude-sonnet-5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "benchcad::benchcad-claude-sonnet-5", + "evaluation_name": "llm_stats.benchcad", + "source_data": { + "dataset_name": "BenchCAD", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-sonnet-5", + "https://llm-stats.com/benchmarks/benchcad", + "https://api.llm-stats.com/leaderboard/benchmarks/benchcad" + ], + "additional_details": { + "raw_benchmark_id": "benchcad", + "raw_model_id": "claude-sonnet-5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BenchCAD is a benchmark for programmatic CAD reasoning built from 17,900 execution-verified CadQuery programs spanning 106 industrial part families, roughly half anchored to real ISO, DIN, EN, ASME, and IEC specification tables. It decomposes CAD capability into matched tasks; the Vision2Code task requires models to generate CadQuery code from multi-view renders, scored by voxel IoU against the reference geometry.", + "metric_id": "llm_stats.benchcad.score", + "metric_name": "BenchCAD score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "benchcad", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BenchCAD", + "raw_categories": "[\"multimodal\",\"reasoning\",\"code\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.373, + "details": { + "raw_score": "0.373", + "raw_score_field": "score", + "raw_model_id": "claude-sonnet-5", + "raw_benchmark_id": "benchcad", + "source_urls_json": "[\"https://llm-stats.com/models/claude-sonnet-5\",\"https://llm-stats.com/benchmarks/benchcad\",\"https://api.llm-stats.com/leaderboard/benchmarks/benchcad\"]", + "raw_score_id": "benchcad::claude-sonnet-5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "browsecomp::browsecomp-claude-sonnet-5", + "evaluation_name": "llm_stats.browsecomp", + "source_data": { + "dataset_name": "BrowseComp", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-sonnet-5", + "https://llm-stats.com/benchmarks/browsecomp", + "https://api.llm-stats.com/leaderboard/benchmarks/browsecomp" + ], + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_model_id": "claude-sonnet-5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BrowseComp is a benchmark comprising 1,266 questions that challenge AI agents to persistently navigate the internet in search of hard-to-find, entangled information. The benchmark measures agents' ability to exercise persistence in information gathering, demonstrate creativity in web navigation, and find concise, verifiable answers. Despite the difficulty of the questions, BrowseComp is simple and easy-to-use, as predicted answers are short and easily verifiable against reference answers.", + "metric_id": "llm_stats.browsecomp.score", + "metric_name": "BrowseComp score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BrowseComp", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "52" + } + }, + "score_details": { + "score": 0.847, + "details": { + "raw_score": "0.847", + "raw_score_field": "score", + "raw_model_id": "claude-sonnet-5", + "raw_benchmark_id": "browsecomp", + "source_urls_json": "[\"https://llm-stats.com/models/claude-sonnet-5\",\"https://llm-stats.com/benchmarks/browsecomp\",\"https://api.llm-stats.com/leaderboard/benchmarks/browsecomp\"]", + "raw_score_id": "browsecomp::claude-sonnet-5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "chartmuseum::chartmuseum-claude-sonnet-5", + "evaluation_name": "llm_stats.chartmuseum", + "source_data": { + "dataset_name": "ChartMuseum", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-sonnet-5", + "https://llm-stats.com/benchmarks/chartmuseum", + "https://api.llm-stats.com/leaderboard/benchmarks/chartmuseum" + ], + "additional_details": { + "raw_benchmark_id": "chartmuseum", + "raw_model_id": "claude-sonnet-5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ChartMuseum is a chart question-answering benchmark of 1,162 expert-annotated questions over real-world chart images drawn from 184 sources, including academic figures, infographics, and unconventional chart designs. It specifically targets questions that require visual reasoning, such as comparing unlabeled visual elements, tracking trajectories, and judging spatial relationships.", + "metric_id": "llm_stats.chartmuseum.score", + "metric_name": "ChartMuseum score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "chartmuseum", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ChartMuseum", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.867, + "details": { + "raw_score": "0.867", + "raw_score_field": "score", + "raw_model_id": "claude-sonnet-5", + "raw_benchmark_id": "chartmuseum", + "source_urls_json": "[\"https://llm-stats.com/models/claude-sonnet-5\",\"https://llm-stats.com/benchmarks/chartmuseum\",\"https://api.llm-stats.com/leaderboard/benchmarks/chartmuseum\"]", + "raw_score_id": "chartmuseum::claude-sonnet-5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "charxiv-r::charxiv-r-claude-sonnet-5", + "evaluation_name": "llm_stats.charxiv-r", + "source_data": { + "dataset_name": "CharXiv-R", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-sonnet-5", + "https://llm-stats.com/benchmarks/charxiv-r", + "https://api.llm-stats.com/leaderboard/benchmarks/charxiv-r" + ], + "additional_details": { + "raw_benchmark_id": "charxiv-r", + "raw_model_id": "claude-sonnet-5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CharXiv-R is the reasoning component of the CharXiv benchmark, focusing on complex reasoning questions that require synthesizing information across visual chart elements. It evaluates multimodal large language models on their ability to understand and reason about scientific charts from arXiv papers through various reasoning tasks.", + "metric_id": "llm_stats.charxiv-r.score", + "metric_name": "CharXiv-R score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "charxiv-r", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CharXiv-R", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "43" + } + }, + "score_details": { + "score": 0.883, + "details": { + "raw_score": "0.883", + "raw_score_field": "score", + "raw_model_id": "claude-sonnet-5", + "raw_benchmark_id": "charxiv-r", + "source_urls_json": "[\"https://llm-stats.com/models/claude-sonnet-5\",\"https://llm-stats.com/benchmarks/charxiv-r\",\"https://api.llm-stats.com/leaderboard/benchmarks/charxiv-r\"]", + "raw_score_id": "charxiv-r::claude-sonnet-5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "frontiercode::frontiercode-claude-sonnet-5", + "evaluation_name": "llm_stats.frontiercode", + "source_data": { + "dataset_name": "FrontierCode", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-sonnet-5", + "https://llm-stats.com/benchmarks/frontiercode", + "https://api.llm-stats.com/leaderboard/benchmarks/frontiercode" + ], + "additional_details": { + "raw_benchmark_id": "frontiercode", + "raw_model_id": "claude-sonnet-5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "FrontierCode is Cognition's coding evaluation that tests whether models can pass difficult coding tasks while meeting the standards of high-quality production codebases. The Diamond subset contains the hardest problems.", + "metric_id": "llm_stats.frontiercode.score", + "metric_name": "FrontierCode score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "frontiercode", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FrontierCode", + "raw_categories": "[\"reasoning\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.388, + "details": { + "raw_score": "0.388", + "raw_score_field": "score", + "raw_model_id": "claude-sonnet-5", + "raw_benchmark_id": "frontiercode", + "source_urls_json": "[\"https://llm-stats.com/models/claude-sonnet-5\",\"https://llm-stats.com/benchmarks/frontiercode\",\"https://api.llm-stats.com/leaderboard/benchmarks/frontiercode\"]", + "raw_score_id": "frontiercode::claude-sonnet-5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gdp-pdf::gdp-pdf-claude-sonnet-5", + "evaluation_name": "llm_stats.gdp-pdf", + "source_data": { + "dataset_name": "GDP.pdf", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-sonnet-5", + "https://llm-stats.com/benchmarks/gdp-pdf", + "https://api.llm-stats.com/leaderboard/benchmarks/gdp-pdf" + ], + "additional_details": { + "raw_benchmark_id": "gdp-pdf", + "raw_model_id": "claude-sonnet-5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "GDP.pdf is a knowledge-work vision benchmark that evaluates models on economically valuable professional tasks presented as visual documents (PDFs), testing document-based reasoning, chart and table interpretation, and problem solving without tools.", + "metric_id": "llm_stats.gdp-pdf.score", + "metric_name": "GDP.pdf score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gdp-pdf", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GDP.pdf", + "raw_categories": "[\"multimodal\",\"reasoning\",\"general\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.816, + "details": { + "raw_score": "0.816", + "raw_score_field": "score", + "raw_model_id": "claude-sonnet-5", + "raw_benchmark_id": "gdp-pdf", + "source_urls_json": "[\"https://llm-stats.com/models/claude-sonnet-5\",\"https://llm-stats.com/benchmarks/gdp-pdf\",\"https://api.llm-stats.com/leaderboard/benchmarks/gdp-pdf\"]", + "raw_score_id": "gdp-pdf::claude-sonnet-5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "healthbench-professional::healthbench-professional-claude-sonnet-5", + "evaluation_name": "llm_stats.healthbench-professional", + "source_data": { + "dataset_name": "HealthBench Professional", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-sonnet-5", + "https://llm-stats.com/benchmarks/healthbench-professional", + "https://api.llm-stats.com/leaderboard/benchmarks/healthbench-professional" + ], + "additional_details": { + "raw_benchmark_id": "healthbench-professional", + "raw_model_id": "claude-sonnet-5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "HealthBench Professional evaluates model capability and safety for clinician use cases using real clinician-style chats and physician-authored grading rubrics.", + "metric_id": "llm_stats.healthbench-professional.score", + "metric_name": "HealthBench Professional score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "healthbench-professional", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HealthBench Professional", + "raw_categories": "[\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.578, + "details": { + "raw_score": "0.578", + "raw_score_field": "score", + "raw_model_id": "claude-sonnet-5", + "raw_benchmark_id": "healthbench-professional", + "source_urls_json": "[\"https://llm-stats.com/models/claude-sonnet-5\",\"https://llm-stats.com/benchmarks/healthbench-professional\",\"https://api.llm-stats.com/leaderboard/benchmarks/healthbench-professional\"]", + "raw_score_id": "healthbench-professional::claude-sonnet-5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "humanity-s-last-exam::humanity-s-last-exam-claude-sonnet-5", + "evaluation_name": "llm_stats.humanity-s-last-exam", + "source_data": { + "dataset_name": "Humanity's Last Exam", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-sonnet-5", + "https://llm-stats.com/benchmarks/humanity-s-last-exam", + "https://api.llm-stats.com/leaderboard/benchmarks/humanity's-last-exam" + ], + "additional_details": { + "raw_benchmark_id": "humanity's-last-exam", + "raw_model_id": "claude-sonnet-5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Humanity's Last Exam (HLE) is a multi-modal academic benchmark with 2,500 questions across mathematics, humanities, and natural sciences, designed to test LLM capabilities at the frontier of human knowledge with unambiguous, verifiable solutions", + "metric_id": "llm_stats.humanity-s-last-exam.score", + "metric_name": "Humanity's Last Exam score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "humanity's-last-exam", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Humanity's Last Exam", + "raw_categories": "[\"math\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "88" + } + }, + "score_details": { + "score": 0.574, + "details": { + "raw_score": "0.574", + "raw_score_field": "score", + "raw_model_id": "claude-sonnet-5", + "raw_benchmark_id": "humanity's-last-exam", + "source_urls_json": "[\"https://llm-stats.com/models/claude-sonnet-5\",\"https://llm-stats.com/benchmarks/humanity-s-last-exam\",\"https://api.llm-stats.com/leaderboard/benchmarks/humanity's-last-exam\"]", + "raw_score_id": "humanity's-last-exam::claude-sonnet-5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "legal-agent-benchmark::legal-agent-benchmark-claude-sonnet-5", + "evaluation_name": "llm_stats.legal-agent-benchmark", + "source_data": { + "dataset_name": "Legal Agent Benchmark", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-sonnet-5", + "https://llm-stats.com/benchmarks/legal-agent-benchmark", + "https://api.llm-stats.com/leaderboard/benchmarks/legal-agent-benchmark" + ], + "additional_details": { + "raw_benchmark_id": "legal-agent-benchmark", + "raw_model_id": "claude-sonnet-5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The Legal Agent Benchmark (LAB) is Harvey's open-source benchmark for evaluating AI agents on complex, long-horizon legal work. Tasks are scored under an all-pass standard against expert-curated rubrics, where a task passes only if every required rubric criterion (facts, conclusions, citations, structure, and analytical moves) passes.", + "metric_id": "llm_stats.legal-agent-benchmark.score", + "metric_name": "Legal Agent Benchmark score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "legal-agent-benchmark", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Legal Agent Benchmark", + "raw_categories": "[\"legal\",\"reasoning\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.058, + "details": { + "raw_score": "0.058", + "raw_score_field": "score", + "raw_model_id": "claude-sonnet-5", + "raw_benchmark_id": "legal-agent-benchmark", + "source_urls_json": "[\"https://llm-stats.com/models/claude-sonnet-5\",\"https://llm-stats.com/benchmarks/legal-agent-benchmark\",\"https://api.llm-stats.com/leaderboard/benchmarks/legal-agent-benchmark\"]", + "raw_score_id": "legal-agent-benchmark::claude-sonnet-5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "officeqa-pro::officeqa-pro-claude-sonnet-5", + "evaluation_name": "llm_stats.officeqa-pro", + "source_data": { + "dataset_name": "OfficeQA Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-sonnet-5", + "https://llm-stats.com/benchmarks/officeqa-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/officeqa-pro" + ], + "additional_details": { + "raw_benchmark_id": "officeqa-pro", + "raw_model_id": "claude-sonnet-5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OfficeQA Pro evaluates AI models on professional knowledge-work questions and tasks drawn from real office workflows, including document analysis, spreadsheet reasoning, and information synthesis across business domains.", + "metric_id": "llm_stats.officeqa-pro.score", + "metric_name": "OfficeQA Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "officeqa-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OfficeQA Pro", + "raw_categories": "[\"reasoning\",\"general\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.594, + "details": { + "raw_score": "0.594", + "raw_score_field": "score", + "raw_model_id": "claude-sonnet-5", + "raw_benchmark_id": "officeqa-pro", + "source_urls_json": "[\"https://llm-stats.com/models/claude-sonnet-5\",\"https://llm-stats.com/benchmarks/officeqa-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/officeqa-pro\"]", + "raw_score_id": "officeqa-pro::claude-sonnet-5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "osworld-verified::osworld-verified-claude-sonnet-5", + "evaluation_name": "llm_stats.osworld-verified", + "source_data": { + "dataset_name": "OSWorld-Verified", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-sonnet-5", + "https://llm-stats.com/benchmarks/osworld-verified", + "https://api.llm-stats.com/leaderboard/benchmarks/osworld-verified" + ], + "additional_details": { + "raw_benchmark_id": "osworld-verified", + "raw_model_id": "claude-sonnet-5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OSWorld-Verified is a verified subset of OSWorld, a scalable real computer environment for multimodal agents supporting task setup, execution-based evaluation, and interactive learning across Ubuntu, Windows, and macOS.", + "metric_id": "llm_stats.osworld-verified.score", + "metric_name": "OSWorld-Verified score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "osworld-verified", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OSWorld-Verified", + "raw_categories": "[\"multimodal\",\"general\",\"agents\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.812, + "details": { + "raw_score": "0.812", + "raw_score_field": "score", + "raw_model_id": "claude-sonnet-5", + "raw_benchmark_id": "osworld-verified", + "source_urls_json": "[\"https://llm-stats.com/models/claude-sonnet-5\",\"https://llm-stats.com/benchmarks/osworld-verified\",\"https://api.llm-stats.com/leaderboard/benchmarks/osworld-verified\"]", + "raw_score_id": "osworld-verified::claude-sonnet-5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-multilingual::swe-bench-multilingual-claude-sonnet-5", + "evaluation_name": "llm_stats.swe-bench-multilingual", + "source_data": { + "dataset_name": "SWE-bench Multilingual", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-sonnet-5", + "https://llm-stats.com/benchmarks/swe-bench-multilingual", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-multilingual" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-multilingual", + "raw_model_id": "claude-sonnet-5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multilingual benchmark for issue resolving in software engineering that covers Java, TypeScript, JavaScript, Go, Rust, C, and C++. Contains 1,632 high-quality instances carefully annotated from 2,456 candidates by 68 expert annotators, designed to evaluate Large Language Models across diverse software ecosystems beyond Python.", + "metric_id": "llm_stats.swe-bench-multilingual.score", + "metric_name": "SWE-bench Multilingual score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-multilingual", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-bench Multilingual", + "raw_categories": "[\"reasoning\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "33" + } + }, + "score_details": { + "score": 0.783, + "details": { + "raw_score": "0.783", + "raw_score_field": "score", + "raw_model_id": "claude-sonnet-5", + "raw_benchmark_id": "swe-bench-multilingual", + "source_urls_json": "[\"https://llm-stats.com/models/claude-sonnet-5\",\"https://llm-stats.com/benchmarks/swe-bench-multilingual\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-multilingual\"]", + "raw_score_id": "swe-bench-multilingual::claude-sonnet-5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-multimodal::swe-bench-multimodal-claude-sonnet-5", + "evaluation_name": "llm_stats.swe-bench-multimodal", + "source_data": { + "dataset_name": "SWE-Bench Multimodal", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-sonnet-5", + "https://llm-stats.com/benchmarks/swe-bench-multimodal", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-multimodal" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-multimodal", + "raw_model_id": "claude-sonnet-5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SWE-Bench Multimodal extends SWE-Bench to evaluate language models on software engineering tasks that involve visual inputs such as screenshots, UI mockups, and diagrams alongside code understanding.", + "metric_id": "llm_stats.swe-bench-multimodal.score", + "metric_name": "SWE-Bench Multimodal score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-multimodal", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Bench Multimodal", + "raw_categories": "[\"multimodal\",\"reasoning\",\"agents\",\"code\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.281, + "details": { + "raw_score": "0.281", + "raw_score_field": "score", + "raw_model_id": "claude-sonnet-5", + "raw_benchmark_id": "swe-bench-multimodal", + "source_urls_json": "[\"https://llm-stats.com/models/claude-sonnet-5\",\"https://llm-stats.com/benchmarks/swe-bench-multimodal\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-multimodal\"]", + "raw_score_id": "swe-bench-multimodal::claude-sonnet-5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-pro::swe-bench-pro-claude-sonnet-5", + "evaluation_name": "llm_stats.swe-bench-pro", + "source_data": { + "dataset_name": "SWE-Bench Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-sonnet-5", + "https://llm-stats.com/benchmarks/swe-bench-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-pro" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-pro", + "raw_model_id": "claude-sonnet-5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SWE-Bench Pro is an advanced version of SWE-Bench that evaluates language models on complex, real-world software engineering tasks requiring extended reasoning and multi-step problem solving.", + "metric_id": "llm_stats.swe-bench-pro.score", + "metric_name": "SWE-Bench Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Bench Pro", + "raw_categories": "[\"reasoning\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "35" + } + }, + "score_details": { + "score": 0.632, + "details": { + "raw_score": "0.632", + "raw_score_field": "score", + "raw_model_id": "claude-sonnet-5", + "raw_benchmark_id": "swe-bench-pro", + "source_urls_json": "[\"https://llm-stats.com/models/claude-sonnet-5\",\"https://llm-stats.com/benchmarks/swe-bench-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-pro\"]", + "raw_score_id": "swe-bench-pro::claude-sonnet-5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-verified::swe-bench-verified-claude-sonnet-5", + "evaluation_name": "llm_stats.swe-bench-verified", + "source_data": { + "dataset_name": "SWE-Bench Verified", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-sonnet-5", + "https://llm-stats.com/benchmarks/swe-bench-verified", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-verified" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-verified", + "raw_model_id": "claude-sonnet-5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A verified subset of 500 software engineering problems from real GitHub issues, validated by human annotators for evaluating language models' ability to resolve real-world coding issues by generating patches for Python codebases.", + "metric_id": "llm_stats.swe-bench-verified.score", + "metric_name": "SWE-Bench Verified score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-verified", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Bench Verified", + "raw_categories": "[\"reasoning\",\"frontend_development\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "103" + } + }, + "score_details": { + "score": 0.852, + "details": { + "raw_score": "0.852", + "raw_score_field": "score", + "raw_model_id": "claude-sonnet-5", + "raw_benchmark_id": "swe-bench-verified", + "source_urls_json": "[\"https://llm-stats.com/models/claude-sonnet-5\",\"https://llm-stats.com/benchmarks/swe-bench-verified\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-verified\"]", + "raw_score_id": "swe-bench-verified::claude-sonnet-5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2::terminal-bench-2-claude-sonnet-5", + "evaluation_name": "llm_stats.terminal-bench-2", + "source_data": { + "dataset_name": "Terminal-Bench 2.0", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-sonnet-5", + "https://llm-stats.com/benchmarks/terminal-bench-2", + "https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2" + ], + "additional_details": { + "raw_benchmark_id": "terminal-bench-2", + "raw_model_id": "claude-sonnet-5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Terminal-Bench 2.0 is an updated benchmark for testing AI agents' tool use ability to operate a computer via terminal. It evaluates how well models can handle real-world, end-to-end tasks autonomously, including compiling code, training models, setting up servers, system administration, security tasks, data science workflows, and cybersecurity vulnerabilities.", + "metric_id": "llm_stats.terminal-bench-2.score", + "metric_name": "Terminal-Bench 2.0 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "terminal-bench-2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Terminal-Bench 2.0", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "49" + } + }, + "score_details": { + "score": 0.804, + "details": { + "raw_score": "0.804", + "raw_score_field": "score", + "raw_model_id": "claude-sonnet-5", + "raw_benchmark_id": "terminal-bench-2", + "source_urls_json": "[\"https://llm-stats.com/models/claude-sonnet-5\",\"https://llm-stats.com/benchmarks/terminal-bench-2\",\"https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2\"]", + "raw_score_id": "terminal-bench-2::claude-sonnet-5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "toolathlon::toolathlon-claude-sonnet-5", + "evaluation_name": "llm_stats.toolathlon", + "source_data": { + "dataset_name": "Toolathlon", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-sonnet-5", + "https://llm-stats.com/benchmarks/toolathlon", + "https://api.llm-stats.com/leaderboard/benchmarks/toolathlon" + ], + "additional_details": { + "raw_benchmark_id": "toolathlon", + "raw_model_id": "claude-sonnet-5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Tool Decathlon is a comprehensive benchmark for evaluating AI agents' ability to use multiple tools across diverse task categories. It measures proficiency in tool selection, sequencing, and execution across ten different tool-use scenarios.", + "metric_id": "llm_stats.toolathlon.score", + "metric_name": "Toolathlon score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "toolathlon", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Toolathlon", + "raw_categories": "[\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "24" + } + }, + "score_details": { + "score": 0.543, + "details": { + "raw_score": "0.543", + "raw_score_field": "score", + "raw_model_id": "claude-sonnet-5", + "raw_benchmark_id": "toolathlon", + "source_urls_json": "[\"https://llm-stats.com/models/claude-sonnet-5\",\"https://llm-stats.com/benchmarks/toolathlon\",\"https://api.llm-stats.com/leaderboard/benchmarks/toolathlon\"]", + "raw_score_id": "toolathlon::claude-sonnet-5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "usamo-2026::usamo-2026-claude-sonnet-5", + "evaluation_name": "llm_stats.usamo-2026", + "source_data": { + "dataset_name": "USAMO 2026", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/claude-sonnet-5", + "https://llm-stats.com/benchmarks/usamo-2026", + "https://api.llm-stats.com/leaderboard/benchmarks/usamo-2026" + ], + "additional_details": { + "raw_benchmark_id": "usamo-2026", + "raw_model_id": "claude-sonnet-5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "USAMO 2026 evaluates models on the six problems from the 2026 United States of America Mathematical Olympiad, requiring rigorous proof-based reasoning. Following the MathArena methodology, proofs are graded against human expert rubrics by dual strong judge models, with the minimum of the two scores taken as final. Maximum score is 42 points.", + "metric_id": "llm_stats.usamo-2026.score", + "metric_name": "USAMO 2026 score", + "metric_kind": "benchmark_score", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "raw_benchmark_id": "usamo-2026", + "raw_score_field": "score", + "bound_strategy": "inferred_percent", + "raw_name": "USAMO 2026", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 33.39, + "details": { + "raw_score": "33.39", + "raw_score_field": "score", + "raw_model_id": "claude-sonnet-5", + "raw_benchmark_id": "usamo-2026", + "source_urls_json": "[\"https://llm-stats.com/models/claude-sonnet-5\",\"https://llm-stats.com/benchmarks/usamo-2026\",\"https://api.llm-stats.com/leaderboard/benchmarks/usamo-2026\"]", + "raw_score_id": "usamo-2026::claude-sonnet-5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/baidu/ernie-4.5/5fb50c3a-cd86-4e76-9793-ee18d3107b66.json b/data/llm_stats/baidu/ernie-4.5/5fb50c3a-cd86-4e76-9793-ee18d3107b66.json new file mode 100644 index 0000000000000000000000000000000000000000..f47d63f0a73f9d2e2284fa5eef8e0f84457e2a68 --- /dev/null +++ b/data/llm_stats/baidu/ernie-4.5/5fb50c3a-cd86-4e76-9793-ee18d3107b66.json @@ -0,0 +1,515 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/baidu_ernie-4.5/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "ERNIE 4.5", + "id": "baidu/ernie-4.5", + "developer": "baidu", + "additional_details": { + "raw_id": "ernie-4.5", + "raw_name": "ERNIE 4.5", + "raw_model_id": "ernie-4.5", + "raw_model_name": "ERNIE 4.5", + "raw_organization_id": "baidu", + "raw_organization_name": "Baidu", + "raw_release_date": "2025-06-25", + "raw_announcement_date": "2025-06-25", + "raw_multimodal": "false", + "raw_param_count": "21000000000", + "raw_provider_slug": "baidu", + "raw_provider_name": "Baidu" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "agieval::agieval-ernie-4.5", + "evaluation_name": "llm_stats.agieval", + "source_data": { + "dataset_name": "AGIEval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/ernie-4.5", + "https://llm-stats.com/benchmarks/agieval", + "https://api.llm-stats.com/leaderboard/benchmarks/agieval" + ], + "additional_details": { + "raw_benchmark_id": "agieval", + "raw_model_id": "ernie-4.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A human-centric benchmark for evaluating foundation models on standardized exams including college entrance exams (Gaokao, SAT), law school admission tests (LSAT), math competitions, lawyer qualification tests, and civil service exams. Contains 20 tasks (18 multiple-choice, 2 cloze) designed to assess understanding, knowledge, reasoning, and calculation abilities in real-world academic and professional contexts.", + "metric_id": "llm_stats.agieval.score", + "metric_name": "AGIEval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "agieval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AGIEval", + "raw_categories": "[\"legal\",\"math\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.285, + "details": { + "raw_score": "0.285", + "raw_score_field": "score", + "raw_model_id": "ernie-4.5", + "raw_benchmark_id": "agieval", + "source_urls_json": "[\"https://llm-stats.com/models/ernie-4.5\",\"https://llm-stats.com/benchmarks/agieval\",\"https://api.llm-stats.com/leaderboard/benchmarks/agieval\"]", + "raw_score_id": "agieval::ernie-4.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "arc-e::arc-e-ernie-4.5", + "evaluation_name": "llm_stats.arc-e", + "source_data": { + "dataset_name": "ARC-E", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/ernie-4.5", + "https://llm-stats.com/benchmarks/arc-e", + "https://api.llm-stats.com/leaderboard/benchmarks/arc-e" + ], + "additional_details": { + "raw_benchmark_id": "arc-e", + "raw_model_id": "ernie-4.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ARC-E (AI2 Reasoning Challenge - Easy Set) is a subset of grade-school level, multiple-choice science questions that requires knowledge and reasoning capabilities. Part of the AI2 Reasoning Challenge dataset containing 5,197 questions that test scientific reasoning and factual knowledge. The Easy Set contains questions that are answerable by retrieval-based and word co-occurrence algorithms, making them more accessible than the Challenge Set.", + "metric_id": "llm_stats.arc-e.score", + "metric_name": "ARC-E score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arc-e", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ARC-E", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.607, + "details": { + "raw_score": "0.607", + "raw_score_field": "score", + "raw_model_id": "ernie-4.5", + "raw_benchmark_id": "arc-e", + "source_urls_json": "[\"https://llm-stats.com/models/ernie-4.5\",\"https://llm-stats.com/benchmarks/arc-e\",\"https://api.llm-stats.com/leaderboard/benchmarks/arc-e\"]", + "raw_score_id": "arc-e::ernie-4.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bbh::bbh-ernie-4.5", + "evaluation_name": "llm_stats.bbh", + "source_data": { + "dataset_name": "BBH", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/ernie-4.5", + "https://llm-stats.com/benchmarks/bbh", + "https://api.llm-stats.com/leaderboard/benchmarks/bbh" + ], + "additional_details": { + "raw_benchmark_id": "bbh", + "raw_model_id": "ernie-4.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Big-Bench Hard (BBH) is a suite of 23 challenging tasks selected from BIG-Bench for which prior language model evaluations did not outperform the average human-rater. These tasks require multi-step reasoning across diverse domains including arithmetic, logical reasoning, reading comprehension, and commonsense reasoning. The benchmark was designed to test capabilities believed to be beyond current language models and focuses on evaluating complex reasoning skills including temporal understanding, spatial reasoning, causal understanding, and deductive logical reasoning.", + "metric_id": "llm_stats.bbh.score", + "metric_name": "BBH score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bbh", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BBH", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.304, + "details": { + "raw_score": "0.304", + "raw_score_field": "score", + "raw_model_id": "ernie-4.5", + "raw_benchmark_id": "bbh", + "source_urls_json": "[\"https://llm-stats.com/models/ernie-4.5\",\"https://llm-stats.com/benchmarks/bbh\",\"https://api.llm-stats.com/leaderboard/benchmarks/bbh\"]", + "raw_score_id": "bbh::ernie-4.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "c-eval::c-eval-ernie-4.5", + "evaluation_name": "llm_stats.c-eval", + "source_data": { + "dataset_name": "C-Eval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/ernie-4.5", + "https://llm-stats.com/benchmarks/c-eval", + "https://api.llm-stats.com/leaderboard/benchmarks/c-eval" + ], + "additional_details": { + "raw_benchmark_id": "c-eval", + "raw_model_id": "ernie-4.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "C-Eval is a comprehensive Chinese evaluation suite designed to assess advanced knowledge and reasoning abilities of foundation models in a Chinese context. It comprises 13,948 multiple-choice questions across 52 diverse disciplines spanning humanities, science, and engineering, with four difficulty levels: middle school, high school, college, and professional. The benchmark includes C-Eval Hard, a subset of very challenging subjects requiring advanced reasoning abilities.", + "metric_id": "llm_stats.c-eval.score", + "metric_name": "C-Eval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "c-eval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "C-Eval", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.407, + "details": { + "raw_score": "0.407", + "raw_score_field": "score", + "raw_model_id": "ernie-4.5", + "raw_benchmark_id": "c-eval", + "source_urls_json": "[\"https://llm-stats.com/models/ernie-4.5\",\"https://llm-stats.com/benchmarks/c-eval\",\"https://api.llm-stats.com/leaderboard/benchmarks/c-eval\"]", + "raw_score_id": "c-eval::ernie-4.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "cluewsc::cluewsc-ernie-4.5", + "evaluation_name": "llm_stats.cluewsc", + "source_data": { + "dataset_name": "CLUEWSC", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/ernie-4.5", + "https://llm-stats.com/benchmarks/cluewsc", + "https://api.llm-stats.com/leaderboard/benchmarks/cluewsc" + ], + "additional_details": { + "raw_benchmark_id": "cluewsc", + "raw_model_id": "ernie-4.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CLUEWSC2020 is the Chinese version of the Winograd Schema Challenge, part of the CLUE benchmark. It focuses on pronoun disambiguation and coreference resolution, requiring models to determine which noun a pronoun refers to in a sentence. The dataset contains 1,244 training samples and 304 development samples extracted from contemporary Chinese literature.", + "metric_id": "llm_stats.cluewsc.score", + "metric_name": "CLUEWSC score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "cluewsc", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CLUEWSC", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.486, + "details": { + "raw_score": "0.486", + "raw_score_field": "score", + "raw_model_id": "ernie-4.5", + "raw_benchmark_id": "cluewsc", + "source_urls_json": "[\"https://llm-stats.com/models/ernie-4.5\",\"https://llm-stats.com/benchmarks/cluewsc\",\"https://api.llm-stats.com/leaderboard/benchmarks/cluewsc\"]", + "raw_score_id": "cluewsc::ernie-4.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "cmmlu::cmmlu-ernie-4.5", + "evaluation_name": "llm_stats.cmmlu", + "source_data": { + "dataset_name": "CMMLU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/ernie-4.5", + "https://llm-stats.com/benchmarks/cmmlu", + "https://api.llm-stats.com/leaderboard/benchmarks/cmmlu" + ], + "additional_details": { + "raw_benchmark_id": "cmmlu", + "raw_model_id": "ernie-4.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CMMLU (Chinese Massive Multitask Language Understanding) is a comprehensive Chinese benchmark that evaluates the knowledge and reasoning capabilities of large language models across 67 different subject topics. The benchmark covers natural sciences, social sciences, engineering, and humanities with multiple-choice questions ranging from basic to advanced professional levels.", + "metric_id": "llm_stats.cmmlu.score", + "metric_name": "CMMLU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "cmmlu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CMMLU", + "raw_categories": "[\"language\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.398, + "details": { + "raw_score": "0.398", + "raw_score_field": "score", + "raw_model_id": "ernie-4.5", + "raw_benchmark_id": "cmmlu", + "source_urls_json": "[\"https://llm-stats.com/models/ernie-4.5\",\"https://llm-stats.com/benchmarks/cmmlu\",\"https://api.llm-stats.com/leaderboard/benchmarks/cmmlu\"]", + "raw_score_id": "cmmlu::ernie-4.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "humaneval::humaneval-ernie-4.5", + "evaluation_name": "llm_stats.humaneval", + "source_data": { + "dataset_name": "HumanEval+", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/ernie-4.5", + "https://llm-stats.com/benchmarks/humaneval", + "https://api.llm-stats.com/leaderboard/benchmarks/humaneval+" + ], + "additional_details": { + "raw_benchmark_id": "humaneval+", + "raw_model_id": "ernie-4.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Enhanced version of HumanEval that extends the original test cases by 80x using EvalPlus framework for rigorous evaluation of LLM-synthesized code functional correctness, detecting previously undetected wrong code", + "metric_id": "llm_stats.humaneval.score", + "metric_name": "HumanEval+ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "humaneval+", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HumanEval+", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.25, + "details": { + "raw_score": "0.25", + "raw_score_field": "score", + "raw_model_id": "ernie-4.5", + "raw_benchmark_id": "humaneval+", + "source_urls_json": "[\"https://llm-stats.com/models/ernie-4.5\",\"https://llm-stats.com/benchmarks/humaneval\",\"https://api.llm-stats.com/leaderboard/benchmarks/humaneval+\"]", + "raw_score_id": "humaneval+::ernie-4.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mbpp::mbpp-ernie-4.5", + "evaluation_name": "llm_stats.mbpp", + "source_data": { + "dataset_name": "MBPP+", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/ernie-4.5", + "https://llm-stats.com/benchmarks/mbpp", + "https://api.llm-stats.com/leaderboard/benchmarks/mbpp+" + ], + "additional_details": { + "raw_benchmark_id": "mbpp+", + "raw_model_id": "ernie-4.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MBPP+ is an enhanced version of MBPP (Mostly Basic Python Problems) with significantly more test cases (35x) for more rigorous evaluation. MBPP is a benchmark of 974 crowd-sourced Python programming problems designed to be solvable by entry-level programmers, covering programming fundamentals and standard library functionality.", + "metric_id": "llm_stats.mbpp.score", + "metric_name": "MBPP+ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mbpp+", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MBPP+", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.402, + "details": { + "raw_score": "0.402", + "raw_score_field": "score", + "raw_model_id": "ernie-4.5", + "raw_benchmark_id": "mbpp+", + "source_urls_json": "[\"https://llm-stats.com/models/ernie-4.5\",\"https://llm-stats.com/benchmarks/mbpp\",\"https://api.llm-stats.com/leaderboard/benchmarks/mbpp+\"]", + "raw_score_id": "mbpp+::ernie-4.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "piqa::piqa-ernie-4.5", + "evaluation_name": "llm_stats.piqa", + "source_data": { + "dataset_name": "PIQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/ernie-4.5", + "https://llm-stats.com/benchmarks/piqa", + "https://api.llm-stats.com/leaderboard/benchmarks/piqa" + ], + "additional_details": { + "raw_benchmark_id": "piqa", + "raw_model_id": "ernie-4.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "PIQA (Physical Interaction: Question Answering) is a benchmark dataset for physical commonsense reasoning in natural language. It tests AI systems' ability to answer questions requiring physical world knowledge through multiple choice questions with everyday situations, focusing on atypical solutions inspired by instructables.com. The dataset contains 21,000 multiple choice questions where models must choose the most appropriate solution for physical interactions.", + "metric_id": "llm_stats.piqa.score", + "metric_name": "PIQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "piqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PIQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.552, + "details": { + "raw_score": "0.552", + "raw_score_field": "score", + "raw_model_id": "ernie-4.5", + "raw_benchmark_id": "piqa", + "source_urls_json": "[\"https://llm-stats.com/models/ernie-4.5\",\"https://llm-stats.com/benchmarks/piqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/piqa\"]", + "raw_score_id": "piqa::ernie-4.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/bytedance/seed-2.0-pro/73725ec2-b498-454d-aad3-5b84571897f2.json b/data/llm_stats/bytedance/seed-2.0-pro/73725ec2-b498-454d-aad3-5b84571897f2.json new file mode 100644 index 0000000000000000000000000000000000000000..153a87767c98597a4d6236866a95004b2a55ba47 --- /dev/null +++ b/data/llm_stats/bytedance/seed-2.0-pro/73725ec2-b498-454d-aad3-5b84571897f2.json @@ -0,0 +1,308 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/bytedance_seed-2.0-pro/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "seed-2.0-pro", + "id": "bytedance/seed-2.0-pro", + "developer": "bytedance", + "additional_details": { + "raw_id": "seed-2.0-pro", + "raw_model_name": "seed-2.0-pro", + "raw_display_name": "Seed 2.0 Pro", + "raw_organization_id": "bytedance", + "raw_organization_name": "ByteDance", + "raw_context_length": "256000", + "raw_input_modalities": "[\"image\",\"text\",\"video\"]", + "raw_output_modalities": "[\"text\"]", + "raw_input_price": "0.5", + "raw_output_price": "3.0", + "raw_model_id": "seed-2.0-pro", + "raw_provider_slug": "bytedance", + "raw_provider_name": "ZeroEval" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aime-2025::aime-2025-seed-2.0-pro", + "evaluation_name": "llm_stats.aime-2025", + "source_data": { + "dataset_name": "AIME 2025", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.0-pro", + "https://llm-stats.com/benchmarks/aime-2025", + "https://api.llm-stats.com/leaderboard/benchmarks/aime-2025" + ], + "additional_details": { + "raw_benchmark_id": "aime-2025", + "raw_model_id": "seed-2.0-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "All 30 problems from the 2025 American Invitational Mathematics Examination (AIME I and AIME II), testing olympiad-level mathematical reasoning with integer answers from 000-999. Used as an AI benchmark to evaluate large language models' ability to solve complex mathematical problems requiring multi-step logical deductions and structured symbolic reasoning.", + "metric_id": "llm_stats.aime-2025.score", + "metric_name": "AIME 2025 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aime-2025", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AIME 2025", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "114" + } + }, + "score_details": { + "score": 0.983, + "details": { + "raw_score": "0.983", + "raw_score_field": "score", + "raw_model_id": "seed-2.0-pro", + "raw_benchmark_id": "aime-2025", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.0-pro\",\"https://llm-stats.com/benchmarks/aime-2025\",\"https://api.llm-stats.com/leaderboard/benchmarks/aime-2025\"]", + "raw_score_id": "aime-2025::seed-2.0-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "aime-2026::aime-2026-seed-2.0-pro", + "evaluation_name": "llm_stats.aime-2026", + "source_data": { + "dataset_name": "AIME 2026", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.0-pro", + "https://llm-stats.com/benchmarks/aime-2026", + "https://api.llm-stats.com/leaderboard/benchmarks/aime-2026" + ], + "additional_details": { + "raw_benchmark_id": "aime-2026", + "raw_model_id": "seed-2.0-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "All 30 problems from the 2026 American Invitational Mathematics Examination (AIME I and AIME II), testing olympiad-level mathematical reasoning with integer answers from 000-999. Used as an AI benchmark to evaluate large language models' ability to solve complex mathematical problems requiring multi-step logical deductions and structured symbolic reasoning.", + "metric_id": "llm_stats.aime-2026.score", + "metric_name": "AIME 2026 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aime-2026", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AIME 2026", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "17" + } + }, + "score_details": { + "score": 0.942, + "details": { + "raw_score": "0.942", + "raw_score_field": "score", + "raw_model_id": "seed-2.0-pro", + "raw_benchmark_id": "aime-2026", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.0-pro\",\"https://llm-stats.com/benchmarks/aime-2026\",\"https://api.llm-stats.com/leaderboard/benchmarks/aime-2026\"]", + "raw_score_id": "aime-2026::seed-2.0-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "browsecomp::browsecomp-seed-2.0-pro", + "evaluation_name": "llm_stats.browsecomp", + "source_data": { + "dataset_name": "BrowseComp", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.0-pro", + "https://llm-stats.com/benchmarks/browsecomp", + "https://api.llm-stats.com/leaderboard/benchmarks/browsecomp" + ], + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_model_id": "seed-2.0-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BrowseComp is a benchmark comprising 1,266 questions that challenge AI agents to persistently navigate the internet in search of hard-to-find, entangled information. The benchmark measures agents' ability to exercise persistence in information gathering, demonstrate creativity in web navigation, and find concise, verifiable answers. Despite the difficulty of the questions, BrowseComp is simple and easy-to-use, as predicted answers are short and easily verifiable against reference answers.", + "metric_id": "llm_stats.browsecomp.score", + "metric_name": "BrowseComp score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BrowseComp", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "52" + } + }, + "score_details": { + "score": 0.773, + "details": { + "raw_score": "0.773", + "raw_score_field": "score", + "raw_model_id": "seed-2.0-pro", + "raw_benchmark_id": "browsecomp", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.0-pro\",\"https://llm-stats.com/benchmarks/browsecomp\",\"https://api.llm-stats.com/leaderboard/benchmarks/browsecomp\"]", + "raw_score_id": "browsecomp::seed-2.0-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gpqa::gpqa-seed-2.0-pro", + "evaluation_name": "llm_stats.gpqa", + "source_data": { + "dataset_name": "GPQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.0-pro", + "https://llm-stats.com/benchmarks/gpqa", + "https://api.llm-stats.com/leaderboard/benchmarks/gpqa" + ], + "additional_details": { + "raw_benchmark_id": "gpqa", + "raw_model_id": "seed-2.0-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A challenging dataset of 448 multiple-choice questions written by domain experts in biology, physics, and chemistry. Questions are Google-proof and extremely difficult, with PhD experts reaching 65% accuracy.", + "metric_id": "llm_stats.gpqa.score", + "metric_name": "GPQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gpqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GPQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\",\"biology\",\"chemistry\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "226" + } + }, + "score_details": { + "score": 0.889, + "details": { + "raw_score": "0.889", + "raw_score_field": "score", + "raw_model_id": "seed-2.0-pro", + "raw_benchmark_id": "gpqa", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.0-pro\",\"https://llm-stats.com/benchmarks/gpqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/gpqa\"]", + "raw_score_id": "gpqa::seed-2.0-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "livecodebench-v6::livecodebench-v6-seed-2.0-pro", + "evaluation_name": "llm_stats.livecodebench-v6", + "source_data": { + "dataset_name": "LiveCodeBench v6", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.0-pro", + "https://llm-stats.com/benchmarks/livecodebench-v6", + "https://api.llm-stats.com/leaderboard/benchmarks/livecodebench-v6" + ], + "additional_details": { + "raw_benchmark_id": "livecodebench-v6", + "raw_model_id": "seed-2.0-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LiveCodeBench is a holistic and contamination-free evaluation benchmark for large language models for code. It continuously collects new problems from programming contests (LeetCode, AtCoder, CodeForces) and evaluates four different scenarios: code generation, self-repair, code execution, and test output prediction. Problems are annotated with release dates to enable evaluation on unseen problems released after a model's training cutoff.", + "metric_id": "llm_stats.livecodebench-v6.score", + "metric_name": "LiveCodeBench v6 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "livecodebench-v6", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LiveCodeBench v6", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "53" + } + }, + "score_details": { + "score": 0.878, + "details": { + "raw_score": "0.878", + "raw_score_field": "score", + "raw_model_id": "seed-2.0-pro", + "raw_benchmark_id": "livecodebench-v6", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.0-pro\",\"https://llm-stats.com/benchmarks/livecodebench-v6\",\"https://api.llm-stats.com/leaderboard/benchmarks/livecodebench-v6\"]", + "raw_score_id": "livecodebench-v6::seed-2.0-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/bytedance/seed-2.1-pro/801b46d0-ea40-4355-9d01-7cf47a16941b.json b/data/llm_stats/bytedance/seed-2.1-pro/801b46d0-ea40-4355-9d01-7cf47a16941b.json new file mode 100644 index 0000000000000000000000000000000000000000..d5d4ca9e19f42e7975a1e358a5fbdbba856722cf --- /dev/null +++ b/data/llm_stats/bytedance/seed-2.1-pro/801b46d0-ea40-4355-9d01-7cf47a16941b.json @@ -0,0 +1,4934 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/bytedance_seed-2.1-pro/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Seed 2.1 Pro", + "id": "bytedance/seed-2.1-pro", + "developer": "bytedance", + "additional_details": { + "raw_id": "seed-2.1-pro", + "raw_name": "Seed 2.1 Pro", + "raw_model_id": "seed-2.1-pro", + "raw_model_name": "Seed 2.1 Pro", + "raw_organization_id": "bytedance", + "raw_organization_name": "ByteDance", + "raw_release_date": "2026-06-24", + "raw_announcement_date": "2026-06-24", + "raw_multimodal": "true", + "raw_provider_slug": "bytedance", + "raw_provider_name": "ByteDance" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aethercode::aethercode-seed-2.1-pro", + "evaluation_name": "llm_stats.aethercode", + "source_data": { + "dataset_name": "AetherCode", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/aethercode", + "https://api.llm-stats.com/leaderboard/benchmarks/aethercode" + ], + "additional_details": { + "raw_benchmark_id": "aethercode", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "AetherCode is a competitive-programming benchmark of olympiad-level algorithmic coding problems.", + "metric_id": "llm_stats.aethercode.score", + "metric_name": "AetherCode score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aethercode", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AetherCode", + "raw_categories": "[\"reasoning\",\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.658, + "details": { + "raw_score": "0.658", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "aethercode", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/aethercode\",\"https://api.llm-stats.com/leaderboard/benchmarks/aethercode\"]", + "raw_score_id": "aethercode::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "agent-startup-bench::agent-startup-bench-seed-2.1-pro", + "evaluation_name": "llm_stats.agent-startup-bench", + "source_data": { + "dataset_name": "Agent Startup Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/agent-startup-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/agent-startup-bench" + ], + "additional_details": { + "raw_benchmark_id": "agent-startup-bench", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Agent Startup Bench measures AI agents on high-economic-value, startup-style tasks that require autonomous planning and execution to deliver practical, verifiable results.", + "metric_id": "llm_stats.agent-startup-bench.score", + "metric_name": "Agent Startup Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "agent-startup-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Agent Startup Bench", + "raw_categories": "[\"reasoning\",\"general\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.688, + "details": { + "raw_score": "0.688", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "agent-startup-bench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/agent-startup-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/agent-startup-bench\"]", + "raw_score_id": "agent-startup-bench::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "agents-last-exam::agents-last-exam-seed-2.1-pro", + "evaluation_name": "llm_stats.agents-last-exam", + "source_data": { + "dataset_name": "Agents' Last Exam", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/agents-last-exam", + "https://api.llm-stats.com/leaderboard/benchmarks/agents-last-exam" + ], + "additional_details": { + "raw_benchmark_id": "agents-last-exam", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Agents' Last Exam is a challenging benchmark for AI agents on hard, long-horizon tasks that test sustained reasoning, planning, and tool use, reported with and without tool access.", + "metric_id": "llm_stats.agents-last-exam.score", + "metric_name": "Agents' Last Exam score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "agents-last-exam", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Agents' Last Exam", + "raw_categories": "[\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.414, + "details": { + "raw_score": "0.414", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "agents-last-exam", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/agents-last-exam\",\"https://api.llm-stats.com/leaderboard/benchmarks/agents-last-exam\"]", + "raw_score_id": "agents-last-exam::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "apex-agents::apex-agents-seed-2.1-pro", + "evaluation_name": "llm_stats.apex-agents", + "source_data": { + "dataset_name": "APEX-Agents", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/apex-agents", + "https://api.llm-stats.com/leaderboard/benchmarks/apex-agents" + ], + "additional_details": { + "raw_benchmark_id": "apex-agents", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "APEX-Agents is a benchmark evaluating AI agents on long horizon professional tasks that require sustained reasoning, planning, and execution across complex multi-step workflows.", + "metric_id": "llm_stats.apex-agents.score", + "metric_name": "APEX-Agents score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "apex-agents", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "APEX-Agents", + "raw_categories": "[\"reasoning\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.338, + "details": { + "raw_score": "0.338", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "apex-agents", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/apex-agents\",\"https://api.llm-stats.com/leaderboard/benchmarks/apex-agents\"]", + "raw_score_id": "apex-agents::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "arcagi2::arcagi2-seed-2.1-pro", + "evaluation_name": "llm_stats.arcagi2", + "source_data": { + "dataset_name": "ArcAGI2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/arcagi2", + "https://api.llm-stats.com/leaderboard/benchmarks/arcagi2" + ], + "additional_details": { + "raw_benchmark_id": "arcagi2", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ARC-AGI-2 is the second-generation Abstraction and Reasoning Corpus benchmark measuring fluid, general reasoning and abstraction.", + "metric_id": "llm_stats.arcagi2.score", + "metric_name": "ArcAGI2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arcagi2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ArcAGI2", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.625, + "details": { + "raw_score": "0.625", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "arcagi2", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/arcagi2\",\"https://api.llm-stats.com/leaderboard/benchmarks/arcagi2\"]", + "raw_score_id": "arcagi2::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "artifacts-bench::artifacts-bench-seed-2.1-pro", + "evaluation_name": "llm_stats.artifacts-bench", + "source_data": { + "dataset_name": "Artifacts Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/artifacts-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/artifacts-bench" + ], + "additional_details": { + "raw_benchmark_id": "artifacts-bench", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Artifacts Bench evaluates a model's ability to generate visual code artifacts, measuring the quality of generated interactive and visual front-end outputs from natural-language requests.", + "metric_id": "llm_stats.artifacts-bench.score", + "metric_name": "Artifacts Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "artifacts-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Artifacts Bench", + "raw_categories": "[\"frontend_development\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.51, + "details": { + "raw_score": "0.51", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "artifacts-bench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/artifacts-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/artifacts-bench\"]", + "raw_score_id": "artifacts-bench::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "babyvision::babyvision-seed-2.1-pro", + "evaluation_name": "llm_stats.babyvision", + "source_data": { + "dataset_name": "BabyVision", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/babyvision", + "https://api.llm-stats.com/leaderboard/benchmarks/babyvision" + ], + "additional_details": { + "raw_benchmark_id": "babyvision", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A benchmark for early-stage visual reasoning and perception on child-like vision tasks.", + "metric_id": "llm_stats.babyvision.score", + "metric_name": "BabyVision score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "babyvision", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BabyVision", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.737, + "details": { + "raw_score": "0.737", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "babyvision", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/babyvision\",\"https://api.llm-stats.com/leaderboard/benchmarks/babyvision\"]", + "raw_score_id": "babyvision::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "beyond-aime::beyond-aime-seed-2.1-pro", + "evaluation_name": "llm_stats.beyond-aime", + "source_data": { + "dataset_name": "Beyond AIME", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/beyond-aime", + "https://api.llm-stats.com/leaderboard/benchmarks/beyond-aime" + ], + "additional_details": { + "raw_benchmark_id": "beyond-aime", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Beyond AIME is a difficult mathematical reasoning benchmark designed to test deeper reasoning chains and harder decomposition than standard AIME-style problem sets.", + "metric_id": "llm_stats.beyond-aime.score", + "metric_name": "Beyond AIME score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "beyond-aime", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Beyond AIME", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.87, + "details": { + "raw_score": "0.87", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "beyond-aime", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/beyond-aime\",\"https://api.llm-stats.com/leaderboard/benchmarks/beyond-aime\"]", + "raw_score_id": "beyond-aime::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "blink::blink-seed-2.1-pro", + "evaluation_name": "llm_stats.blink", + "source_data": { + "dataset_name": "BLINK", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/blink", + "https://api.llm-stats.com/leaderboard/benchmarks/blink" + ], + "additional_details": { + "raw_benchmark_id": "blink", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BLINK: Multimodal Large Language Models Can See but Not Perceive. A benchmark for multimodal language models focusing on core visual perception abilities. Reformats 14 classic computer vision tasks into 3,807 multiple-choice questions paired with single or multiple images and visual prompting. Tasks include relative depth estimation, visual correspondence, forensics detection, multi-view reasoning, counting, object localization, and spatial reasoning that humans can solve 'within a blink'.", + "metric_id": "llm_stats.blink.score", + "metric_name": "BLINK score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "blink", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BLINK", + "raw_categories": "[\"multimodal\",\"reasoning\",\"spatial_reasoning\",\"3d\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.814, + "details": { + "raw_score": "0.814", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "blink", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/blink\",\"https://api.llm-stats.com/leaderboard/benchmarks/blink\"]", + "raw_score_id": "blink::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "browsecomp::browsecomp-seed-2.1-pro", + "evaluation_name": "llm_stats.browsecomp", + "source_data": { + "dataset_name": "BrowseComp", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/browsecomp", + "https://api.llm-stats.com/leaderboard/benchmarks/browsecomp" + ], + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BrowseComp is a benchmark comprising 1,266 questions that challenge AI agents to persistently navigate the internet in search of hard-to-find, entangled information. The benchmark measures agents' ability to exercise persistence in information gathering, demonstrate creativity in web navigation, and find concise, verifiable answers. Despite the difficulty of the questions, BrowseComp is simple and easy-to-use, as predicted answers are short and easily verifiable against reference answers.", + "metric_id": "llm_stats.browsecomp.score", + "metric_name": "BrowseComp score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BrowseComp", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "52" + } + }, + "score_details": { + "score": 0.862, + "details": { + "raw_score": "0.862", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "browsecomp", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/browsecomp\",\"https://api.llm-stats.com/leaderboard/benchmarks/browsecomp\"]", + "raw_score_id": "browsecomp::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "chartqapro::chartqapro-seed-2.1-pro", + "evaluation_name": "llm_stats.chartqapro", + "source_data": { + "dataset_name": "ChartQAPro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/chartqapro", + "https://api.llm-stats.com/leaderboard/benchmarks/chartqapro" + ], + "additional_details": { + "raw_benchmark_id": "chartqapro", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ChartQAPro is a challenging benchmark for question answering over diverse, real-world charts and infographics.", + "metric_id": "llm_stats.chartqapro.score", + "metric_name": "ChartQAPro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "chartqapro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ChartQAPro", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.709, + "details": { + "raw_score": "0.709", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "chartqapro", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/chartqapro\",\"https://api.llm-stats.com/leaderboard/benchmarks/chartqapro\"]", + "raw_score_id": "chartqapro::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "charxiv-d::charxiv-d-seed-2.1-pro", + "evaluation_name": "llm_stats.charxiv-d", + "source_data": { + "dataset_name": "CharXiv-D", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/charxiv-d", + "https://api.llm-stats.com/leaderboard/benchmarks/charxiv-d" + ], + "additional_details": { + "raw_benchmark_id": "charxiv-d", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CharXiv-D is the descriptive questions subset of the CharXiv benchmark, designed to assess multimodal large language models' ability to extract basic information from scientific charts. It contains descriptive questions covering information extraction, enumeration, pattern recognition, and counting across 2,323 diverse charts from arXiv papers, all curated and verified by human experts.", + "metric_id": "llm_stats.charxiv-d.score", + "metric_name": "CharXiv-D score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "charxiv-d", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CharXiv-D", + "raw_categories": "[\"multimodal\",\"reasoning\",\"structured_output\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.955, + "details": { + "raw_score": "0.955", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "charxiv-d", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/charxiv-d\",\"https://api.llm-stats.com/leaderboard/benchmarks/charxiv-d\"]", + "raw_score_id": "charxiv-d::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "charxiv-r::charxiv-r-seed-2.1-pro", + "evaluation_name": "llm_stats.charxiv-r", + "source_data": { + "dataset_name": "CharXiv-R", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/charxiv-r", + "https://api.llm-stats.com/leaderboard/benchmarks/charxiv-r" + ], + "additional_details": { + "raw_benchmark_id": "charxiv-r", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CharXiv-R is the reasoning component of the CharXiv benchmark, focusing on complex reasoning questions that require synthesizing information across visual chart elements. It evaluates multimodal large language models on their ability to understand and reason about scientific charts from arXiv papers through various reasoning tasks.", + "metric_id": "llm_stats.charxiv-r.score", + "metric_name": "CharXiv-R score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "charxiv-r", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CharXiv-R", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "43" + } + }, + "score_details": { + "score": 0.864, + "details": { + "raw_score": "0.864", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "charxiv-r", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/charxiv-r\",\"https://api.llm-stats.com/leaderboard/benchmarks/charxiv-r\"]", + "raw_score_id": "charxiv-r::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "claw-eval-mm::claw-eval-mm-seed-2.1-pro", + "evaluation_name": "llm_stats.claw-eval-mm", + "source_data": { + "dataset_name": "ClawEval-MM", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/claw-eval-mm", + "https://api.llm-stats.com/leaderboard/benchmarks/claw-eval-mm" + ], + "additional_details": { + "raw_benchmark_id": "claw-eval-mm", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ClawEval-MM is the multimodal variant of ClawEval, evaluating agentic problem solving with visual inputs.", + "metric_id": "llm_stats.claw-eval-mm.score", + "metric_name": "ClawEval-MM score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "claw-eval-mm", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ClawEval-MM", + "raw_categories": "[\"multimodal\",\"agents\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.51, + "details": { + "raw_score": "0.51", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "claw-eval-mm", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/claw-eval-mm\",\"https://api.llm-stats.com/leaderboard/benchmarks/claw-eval-mm\"]", + "raw_score_id": "claw-eval-mm::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "contphy::contphy-seed-2.1-pro", + "evaluation_name": "llm_stats.contphy", + "source_data": { + "dataset_name": "ContPhy", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/contphy", + "https://api.llm-stats.com/leaderboard/benchmarks/contphy" + ], + "additional_details": { + "raw_benchmark_id": "contphy", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ContPhy is a continuum physical-reasoning benchmark evaluating understanding of physical dynamics in video.", + "metric_id": "llm_stats.contphy.score", + "metric_name": "ContPhy score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "contphy", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ContPhy", + "raw_categories": "[\"multimodal\",\"reasoning\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.636, + "details": { + "raw_score": "0.636", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "contphy", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/contphy\",\"https://api.llm-stats.com/leaderboard/benchmarks/contphy\"]", + "raw_score_id": "contphy::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "creativework::creativework-seed-2.1-pro", + "evaluation_name": "llm_stats.creativework", + "source_data": { + "dataset_name": "CreativeWork", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/creativework", + "https://api.llm-stats.com/leaderboard/benchmarks/creativework" + ], + "additional_details": { + "raw_benchmark_id": "creativework", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CreativeWork evaluates agents on open-ended creative production tasks within realistic tool and application environments, measuring the quality and completeness of generated deliverables.", + "metric_id": "llm_stats.creativework.score", + "metric_name": "CreativeWork score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "creativework", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CreativeWork", + "raw_categories": "[\"reasoning\",\"general\",\"agents\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.425, + "details": { + "raw_score": "0.425", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "creativework", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/creativework\",\"https://api.llm-stats.com/leaderboard/benchmarks/creativework\"]", + "raw_score_id": "creativework::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "crossvid::crossvid-seed-2.1-pro", + "evaluation_name": "llm_stats.crossvid", + "source_data": { + "dataset_name": "CrossVid", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/crossvid", + "https://api.llm-stats.com/leaderboard/benchmarks/crossvid" + ], + "additional_details": { + "raw_benchmark_id": "crossvid", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CrossVid evaluates cross-video reasoning, requiring models to integrate information across multiple videos.", + "metric_id": "llm_stats.crossvid.score", + "metric_name": "CrossVid score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "crossvid", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CrossVid", + "raw_categories": "[\"multimodal\",\"reasoning\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.65, + "details": { + "raw_score": "0.65", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "crossvid", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/crossvid\",\"https://api.llm-stats.com/leaderboard/benchmarks/crossvid\"]", + "raw_score_id": "crossvid::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "cybergym::cybergym-seed-2.1-pro", + "evaluation_name": "llm_stats.cybergym", + "source_data": { + "dataset_name": "CyberGym", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/cybergym", + "https://api.llm-stats.com/leaderboard/benchmarks/cybergym" + ], + "additional_details": { + "raw_benchmark_id": "cybergym", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CyberGym is a benchmark for evaluating AI agents on cybersecurity tasks, testing their ability to identify vulnerabilities, perform security analysis, and complete security-related challenges in a controlled environment.", + "metric_id": "llm_stats.cybergym.score", + "metric_name": "CyberGym score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "cybergym", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CyberGym", + "raw_categories": "[\"safety\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.702, + "details": { + "raw_score": "0.702", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "cybergym", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/cybergym\",\"https://api.llm-stats.com/leaderboard/benchmarks/cybergym\"]", + "raw_score_id": "cybergym::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "deepswe::deepswe-seed-2.1-pro", + "evaluation_name": "llm_stats.deepswe", + "source_data": { + "dataset_name": "DeepSWE", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/deepswe", + "https://api.llm-stats.com/leaderboard/benchmarks/deepswe" + ], + "additional_details": { + "raw_benchmark_id": "deepswe", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "DeepSWE is a software engineering agent benchmark evaluated with the mini-swe-agent harness, where each task is solved in an isolated container with no internet access. It measures an agent's ability to autonomously resolve real-world coding issues end to end.", + "metric_id": "llm_stats.deepswe.score", + "metric_name": "DeepSWE score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "deepswe", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DeepSWE", + "raw_categories": "[\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.327, + "details": { + "raw_score": "0.327", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "deepswe", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/deepswe\",\"https://api.llm-stats.com/leaderboard/benchmarks/deepswe\"]", + "raw_score_id": "deepswe::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "doubao-multi-turn-bench::doubao-multi-turn-bench-seed-2.1-pro", + "evaluation_name": "llm_stats.doubao-multi-turn-bench", + "source_data": { + "dataset_name": "Doubao Multi-Turn Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/doubao-multi-turn-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/doubao-multi-turn-bench" + ], + "additional_details": { + "raw_benchmark_id": "doubao-multi-turn-bench", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Doubao Multi-Turn Bench evaluates models on multi-turn conversational tasks, measuring context retention, instruction following, and coherent reasoning across extended dialogues.", + "metric_id": "llm_stats.doubao-multi-turn-bench.score", + "metric_name": "Doubao Multi-Turn Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "doubao-multi-turn-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Doubao Multi-Turn Bench", + "raw_categories": "[\"instruction_following\",\"general\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.525, + "details": { + "raw_score": "0.525", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "doubao-multi-turn-bench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/doubao-multi-turn-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/doubao-multi-turn-bench\"]", + "raw_score_id": "doubao-multi-turn-bench::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "dude::dude-seed-2.1-pro", + "evaluation_name": "llm_stats.dude", + "source_data": { + "dataset_name": "DUDE", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/dude", + "https://api.llm-stats.com/leaderboard/benchmarks/dude" + ], + "additional_details": { + "raw_benchmark_id": "dude", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "DUDE (Document Understanding Dataset and Evaluation) tests multi-page, multi-domain document understanding and reasoning.", + "metric_id": "llm_stats.dude.score", + "metric_name": "DUDE score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "dude", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DUDE", + "raw_categories": "[\"long_context\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.828, + "details": { + "raw_score": "0.828", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "dude", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/dude\",\"https://api.llm-stats.com/leaderboard/benchmarks/dude\"]", + "raw_score_id": "dude::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "dynamath::dynamath-seed-2.1-pro", + "evaluation_name": "llm_stats.dynamath", + "source_data": { + "dataset_name": "DynaMath", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/dynamath", + "https://api.llm-stats.com/leaderboard/benchmarks/dynamath" + ], + "additional_details": { + "raw_benchmark_id": "dynamath", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multimodal mathematics and reasoning benchmark focused on dynamic visual problem solving.", + "metric_id": "llm_stats.dynamath.score", + "metric_name": "DynaMath score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "dynamath", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DynaMath", + "raw_categories": "[\"math\",\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.731, + "details": { + "raw_score": "0.731", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "dynamath", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/dynamath\",\"https://api.llm-stats.com/leaderboard/benchmarks/dynamath\"]", + "raw_score_id": "dynamath::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "embspatialbench::embspatialbench-seed-2.1-pro", + "evaluation_name": "llm_stats.embspatialbench", + "source_data": { + "dataset_name": "EmbSpatialBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/embspatialbench", + "https://api.llm-stats.com/leaderboard/benchmarks/embspatialbench" + ], + "additional_details": { + "raw_benchmark_id": "embspatialbench", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "EmbSpatialBench evaluates embodied spatial understanding and reasoning capabilities.", + "metric_id": "llm_stats.embspatialbench.score", + "metric_name": "EmbSpatialBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "embspatialbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "EmbSpatialBench", + "raw_categories": "[\"spatial_reasoning\",\"embodied\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.834, + "details": { + "raw_score": "0.834", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "embspatialbench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/embspatialbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/embspatialbench\"]", + "raw_score_id": "embspatialbench::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "emma::emma-seed-2.1-pro", + "evaluation_name": "llm_stats.emma", + "source_data": { + "dataset_name": "EMMA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/emma", + "https://api.llm-stats.com/leaderboard/benchmarks/emma" + ], + "additional_details": { + "raw_benchmark_id": "emma", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "EMMA (Enhanced MultiModal reAsoning) is a benchmark for organic multimodal reasoning across mathematics, physics, chemistry, and coding.", + "metric_id": "llm_stats.emma.score", + "metric_name": "EMMA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "emma", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "EMMA", + "raw_categories": "[\"math\",\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.793, + "details": { + "raw_score": "0.793", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "emma", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/emma\",\"https://api.llm-stats.com/leaderboard/benchmarks/emma\"]", + "raw_score_id": "emma::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "erqa::erqa-seed-2.1-pro", + "evaluation_name": "llm_stats.erqa", + "source_data": { + "dataset_name": "ERQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/erqa", + "https://api.llm-stats.com/leaderboard/benchmarks/erqa" + ], + "additional_details": { + "raw_benchmark_id": "erqa", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Embodied Reasoning Question Answering benchmark consisting of 400 multiple-choice visual questions across spatial reasoning, trajectory reasoning, action reasoning, state estimation, and multi-view reasoning for evaluating AI capabilities in physical world interactions", + "metric_id": "llm_stats.erqa.score", + "metric_name": "ERQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "erqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ERQA", + "raw_categories": "[\"reasoning\",\"spatial_reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "22" + } + }, + "score_details": { + "score": 0.72, + "details": { + "raw_score": "0.72", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "erqa", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/erqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/erqa\"]", + "raw_score_id": "erqa::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "finance-agent-v1.1::finance-agent-v1.1-seed-2.1-pro", + "evaluation_name": "llm_stats.finance-agent-v1.1", + "source_data": { + "dataset_name": "Finance Agent v1.1", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/finance-agent-v1.1", + "https://api.llm-stats.com/leaderboard/benchmarks/finance-agent-v1.1" + ], + "additional_details": { + "raw_benchmark_id": "finance-agent-v1.1", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Finance Agent v1.1 is an agentic financial-analysis benchmark that evaluates models on real-world finance workflows, including retrieving and reasoning over financial documents and performing multi-step calculations.", + "metric_id": "llm_stats.finance-agent-v1.1.score", + "metric_name": "Finance Agent v1.1 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "finance-agent-v1.1", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Finance Agent v1.1", + "raw_categories": "[\"reasoning\",\"finance\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.607, + "details": { + "raw_score": "0.607", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "finance-agent-v1.1", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/finance-agent-v1.1\",\"https://api.llm-stats.com/leaderboard/benchmarks/finance-agent-v1.1\"]", + "raw_score_id": "finance-agent-v1.1::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "frontiercs::frontiercs-seed-2.1-pro", + "evaluation_name": "llm_stats.frontiercs", + "source_data": { + "dataset_name": "FrontierCS", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/frontiercs", + "https://api.llm-stats.com/leaderboard/benchmarks/frontiercs" + ], + "additional_details": { + "raw_benchmark_id": "frontiercs", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "FrontierCS is a benchmark of frontier computer-science problems requiring deep theoretical understanding and rigorous multi-step reasoning at the edge of the field.", + "metric_id": "llm_stats.frontiercs.score", + "metric_name": "FrontierCS score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "frontiercs", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FrontierCS", + "raw_categories": "[\"reasoning\",\"science\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.463, + "details": { + "raw_score": "0.463", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "frontiercs", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/frontiercs\",\"https://api.llm-stats.com/leaderboard/benchmarks/frontiercs\"]", + "raw_score_id": "frontiercs::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "frontierscience-olympiad::frontierscience-olympiad-seed-2.1-pro", + "evaluation_name": "llm_stats.frontierscience-olympiad", + "source_data": { + "dataset_name": "FrontierScience Olympiad", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/frontierscience-olympiad", + "https://api.llm-stats.com/leaderboard/benchmarks/frontierscience-olympiad" + ], + "additional_details": { + "raw_benchmark_id": "frontierscience-olympiad", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "FrontierScience Olympiad is a benchmark of olympiad-level scientific reasoning problems, testing expert understanding and multi-step reasoning across advanced natural-science domains.", + "metric_id": "llm_stats.frontierscience-olympiad.score", + "metric_name": "FrontierScience Olympiad score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "frontierscience-olympiad", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FrontierScience Olympiad", + "raw_categories": "[\"math\",\"reasoning\",\"science\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.75, + "details": { + "raw_score": "0.75", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "frontierscience-olympiad", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/frontierscience-olympiad\",\"https://api.llm-stats.com/leaderboard/benchmarks/frontierscience-olympiad\"]", + "raw_score_id": "frontierscience-olympiad::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "frontierscience-research::frontierscience-research-seed-2.1-pro", + "evaluation_name": "llm_stats.frontierscience-research", + "source_data": { + "dataset_name": "FrontierScience Research", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/frontierscience-research", + "https://api.llm-stats.com/leaderboard/benchmarks/frontierscience-research" + ], + "additional_details": { + "raw_benchmark_id": "frontierscience-research", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "FrontierScience Research is a benchmark evaluating AI models on cutting-edge scientific research questions requiring deep domain expertise, multi-step reasoning, and synthesis of complex scientific concepts across disciplines.", + "metric_id": "llm_stats.frontierscience-research.score", + "metric_name": "FrontierScience Research score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "frontierscience-research", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FrontierScience Research", + "raw_categories": "[\"reasoning\",\"science\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.283, + "details": { + "raw_score": "0.283", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "frontierscience-research", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/frontierscience-research\",\"https://api.llm-stats.com/leaderboard/benchmarks/frontierscience-research\"]", + "raw_score_id": "frontierscience-research::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gameworld::gameworld-seed-2.1-pro", + "evaluation_name": "llm_stats.gameworld", + "source_data": { + "dataset_name": "GameWorld", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/gameworld", + "https://api.llm-stats.com/leaderboard/benchmarks/gameworld" + ], + "additional_details": { + "raw_benchmark_id": "gameworld", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "GameWorld evaluates agents on interactive game environments, testing perception, planning, and sequential decision-making to accomplish in-game objectives.", + "metric_id": "llm_stats.gameworld.score", + "metric_name": "GameWorld score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gameworld", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GameWorld", + "raw_categories": "[\"multimodal\",\"reasoning\",\"agents\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.312, + "details": { + "raw_score": "0.312", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "gameworld", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/gameworld\",\"https://api.llm-stats.com/leaderboard/benchmarks/gameworld\"]", + "raw_score_id": "gameworld::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gdpval::gdpval-seed-2.1-pro", + "evaluation_name": "llm_stats.gdpval", + "source_data": { + "dataset_name": "GDPval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/gdpval", + "https://api.llm-stats.com/leaderboard/benchmarks/gdpval" + ], + "additional_details": { + "raw_benchmark_id": "gdpval", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "GDPval is an OpenAI benchmark evaluating AI models on economically valuable, real-world knowledge-work tasks spanning many professional occupations and industries.", + "metric_id": "llm_stats.gdpval.score", + "metric_name": "GDPval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gdpval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GDPval", + "raw_categories": "[\"legal\",\"reasoning\",\"finance\",\"general\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.879, + "details": { + "raw_score": "0.879", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "gdpval", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/gdpval\",\"https://api.llm-stats.com/leaderboard/benchmarks/gdpval\"]", + "raw_score_id": "gdpval::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "horizonmath::horizonmath-seed-2.1-pro", + "evaluation_name": "llm_stats.horizonmath", + "source_data": { + "dataset_name": "HorizonMath", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/horizonmath", + "https://api.llm-stats.com/leaderboard/benchmarks/horizonmath" + ], + "additional_details": { + "raw_benchmark_id": "horizonmath", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "HorizonMath is an extremely difficult frontier mathematics benchmark designed to test the limits of mathematical reasoning on research-level and competition-beyond problems.", + "metric_id": "llm_stats.horizonmath.score", + "metric_name": "HorizonMath score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "horizonmath", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HorizonMath", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.02, + "details": { + "raw_score": "0.02", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "horizonmath", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/horizonmath\",\"https://api.llm-stats.com/leaderboard/benchmarks/horizonmath\"]", + "raw_score_id": "horizonmath::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "humanity-s-last-exam::humanity-s-last-exam-seed-2.1-pro", + "evaluation_name": "llm_stats.humanity-s-last-exam", + "source_data": { + "dataset_name": "Humanity's Last Exam", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/humanity-s-last-exam", + "https://api.llm-stats.com/leaderboard/benchmarks/humanity's-last-exam" + ], + "additional_details": { + "raw_benchmark_id": "humanity's-last-exam", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Humanity's Last Exam (HLE) is a multi-modal academic benchmark with 2,500 questions across mathematics, humanities, and natural sciences, designed to test LLM capabilities at the frontier of human knowledge with unambiguous, verifiable solutions", + "metric_id": "llm_stats.humanity-s-last-exam.score", + "metric_name": "Humanity's Last Exam score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "humanity's-last-exam", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Humanity's Last Exam", + "raw_categories": "[\"math\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "88" + } + }, + "score_details": { + "score": 0.557, + "details": { + "raw_score": "0.557", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "humanity's-last-exam", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/humanity-s-last-exam\",\"https://api.llm-stats.com/leaderboard/benchmarks/humanity's-last-exam\"]", + "raw_score_id": "humanity's-last-exam::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "image2floorplan::image2floorplan-seed-2.1-pro", + "evaluation_name": "llm_stats.image2floorplan", + "source_data": { + "dataset_name": "Image2FloorPlan", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/image2floorplan", + "https://api.llm-stats.com/leaderboard/benchmarks/image2floorplan" + ], + "additional_details": { + "raw_benchmark_id": "image2floorplan", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Image2FloorPlan is an in-house benchmark evaluating multimodal models on generating structured floor plans and interactive frontends directly from images such as design mockups and room photos.", + "metric_id": "llm_stats.image2floorplan.score", + "metric_name": "Image2FloorPlan score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "image2floorplan", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Image2FloorPlan", + "raw_categories": "[\"multimodal\",\"coding\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.48, + "details": { + "raw_score": "0.48", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "image2floorplan", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/image2floorplan\",\"https://api.llm-stats.com/leaderboard/benchmarks/image2floorplan\"]", + "raw_score_id": "image2floorplan::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "imo-2025::imo-2025-seed-2.1-pro", + "evaluation_name": "llm_stats.imo-2025", + "source_data": { + "dataset_name": "IMO 2025", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/imo-2025", + "https://api.llm-stats.com/leaderboard/benchmarks/imo-2025" + ], + "additional_details": { + "raw_benchmark_id": "imo-2025", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "IMO 2025 evaluates models on the six problems from the 2025 International Mathematical Olympiad, requiring rigorous proof-based reasoning. Following the MathArena methodology, proofs are graded against human expert rubrics by dual strong judge models, with the minimum of the two scores taken as final. Maximum score is 42 points.", + "metric_id": "llm_stats.imo-2025.score", + "metric_name": "IMO 2025 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "imo-2025", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IMO 2025", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.652, + "details": { + "raw_score": "0.652", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "imo-2025", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/imo-2025\",\"https://api.llm-stats.com/leaderboard/benchmarks/imo-2025\"]", + "raw_score_id": "imo-2025::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "imoproof-adv::imoproof-adv-seed-2.1-pro", + "evaluation_name": "llm_stats.imoproof-adv", + "source_data": { + "dataset_name": "IMOProof-Adv", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/imoproof-adv", + "https://api.llm-stats.com/leaderboard/benchmarks/imoproof-adv" + ], + "additional_details": { + "raw_benchmark_id": "imoproof-adv", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "IMOProof-Adv is an advanced benchmark of International Mathematical Olympiad-style proof problems requiring rigorous multi-step mathematical proofs.", + "metric_id": "llm_stats.imoproof-adv.score", + "metric_name": "IMOProof-Adv score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "imoproof-adv", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IMOProof-Adv", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.543, + "details": { + "raw_score": "0.543", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "imoproof-adv", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/imoproof-adv\",\"https://api.llm-stats.com/leaderboard/benchmarks/imoproof-adv\"]", + "raw_score_id": "imoproof-adv::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ipho-2025::ipho-2025-seed-2.1-pro", + "evaluation_name": "llm_stats.ipho-2025", + "source_data": { + "dataset_name": "IPhO 2025", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/ipho-2025", + "https://api.llm-stats.com/leaderboard/benchmarks/ipho-2025" + ], + "additional_details": { + "raw_benchmark_id": "ipho-2025", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "International Physics Olympiad 2025 (theory) comprises all 3 theory problems from the official 2025 IPhO competition. Results are based on blinded human evaluation with guidelines based on the official competition scoring, validated by domain experts.", + "metric_id": "llm_stats.ipho-2025.score", + "metric_name": "IPhO 2025 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ipho-2025", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IPhO 2025", + "raw_categories": "[\"physics\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.793, + "details": { + "raw_score": "0.793", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "ipho-2025", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/ipho-2025\",\"https://api.llm-stats.com/leaderboard/benchmarks/ipho-2025\"]", + "raw_score_id": "ipho-2025::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "kina::kina-seed-2.1-pro", + "evaluation_name": "llm_stats.kina", + "source_data": { + "dataset_name": "KINA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/kina", + "https://api.llm-stats.com/leaderboard/benchmarks/kina" + ], + "additional_details": { + "raw_benchmark_id": "kina", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "KINA is a knowledge-intensive evaluation that measures a model's breadth and depth of factual knowledge across academic and professional domains.", + "metric_id": "llm_stats.kina.score", + "metric_name": "KINA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "kina", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "KINA", + "raw_categories": "[\"knowledge\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.483, + "details": { + "raw_score": "0.483", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "kina", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/kina\",\"https://api.llm-stats.com/leaderboard/benchmarks/kina\"]", + "raw_score_id": "kina::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "livemathematicianbench::livemathematicianbench-seed-2.1-pro", + "evaluation_name": "llm_stats.livemathematicianbench", + "source_data": { + "dataset_name": "LiveMathematicianBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/livemathematicianbench", + "https://api.llm-stats.com/leaderboard/benchmarks/livemathematicianbench" + ], + "additional_details": { + "raw_benchmark_id": "livemathematicianbench", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LiveMathematicianBench evaluates research-level mathematical reasoning on continuously refreshed, contamination-resistant problems.", + "metric_id": "llm_stats.livemathematicianbench.score", + "metric_name": "LiveMathematicianBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "livemathematicianbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LiveMathematicianBench", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.209, + "details": { + "raw_score": "0.209", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "livemathematicianbench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/livemathematicianbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/livemathematicianbench\"]", + "raw_score_id": "livemathematicianbench::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "livesports-3k::livesports-3k-seed-2.1-pro", + "evaluation_name": "llm_stats.livesports-3k", + "source_data": { + "dataset_name": "LiveSports-3K", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/livesports-3k", + "https://api.llm-stats.com/leaderboard/benchmarks/livesports-3k" + ], + "additional_details": { + "raw_benchmark_id": "livesports-3k", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LiveSports-3K evaluates fine-grained understanding and commentary of live sports video.", + "metric_id": "llm_stats.livesports-3k.score", + "metric_name": "LiveSports-3K score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "livesports-3k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LiveSports-3K", + "raw_categories": "[\"multimodal\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.768, + "details": { + "raw_score": "0.768", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "livesports-3k", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/livesports-3k\",\"https://api.llm-stats.com/leaderboard/benchmarks/livesports-3k\"]", + "raw_score_id": "livesports-3k::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "longvideobench::longvideobench-seed-2.1-pro", + "evaluation_name": "llm_stats.longvideobench", + "source_data": { + "dataset_name": "LongVideoBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/longvideobench", + "https://api.llm-stats.com/leaderboard/benchmarks/longvideobench" + ], + "additional_details": { + "raw_benchmark_id": "longvideobench", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LongVideoBench is a question-answering benchmark featuring video-language interleaved inputs up to an hour long. It includes 3,763 varying-length web-collected videos with subtitles across diverse themes and 6,678 human-annotated multiple-choice questions in 17 fine-grained categories for comprehensive evaluation of long-term multimodal understanding.", + "metric_id": "llm_stats.longvideobench.score", + "metric_name": "LongVideoBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "longvideobench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LongVideoBench", + "raw_categories": "[\"long_context\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.806, + "details": { + "raw_score": "0.806", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "longvideobench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/longvideobench\",\"https://api.llm-stats.com/leaderboard/benchmarks/longvideobench\"]", + "raw_score_id": "longvideobench::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "lvbench::lvbench-seed-2.1-pro", + "evaluation_name": "llm_stats.lvbench", + "source_data": { + "dataset_name": "LVBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/lvbench", + "https://api.llm-stats.com/leaderboard/benchmarks/lvbench" + ], + "additional_details": { + "raw_benchmark_id": "lvbench", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LVBench is an extreme long video understanding benchmark designed to evaluate multimodal models on videos up to two hours in duration. It contains 6 major categories and 21 subcategories, with videos averaging five times longer than existing datasets. The benchmark addresses applications requiring comprehension of extremely long videos.", + "metric_id": "llm_stats.lvbench.score", + "metric_name": "LVBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "lvbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LVBench", + "raw_categories": "[\"long_context\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.78, + "details": { + "raw_score": "0.78", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "lvbench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/lvbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/lvbench\"]", + "raw_score_id": "lvbench::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "matharena-apex::matharena-apex-seed-2.1-pro", + "evaluation_name": "llm_stats.matharena-apex", + "source_data": { + "dataset_name": "MathArena Apex", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/matharena-apex", + "https://api.llm-stats.com/leaderboard/benchmarks/matharena-apex" + ], + "additional_details": { + "raw_benchmark_id": "matharena-apex", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MathArena Apex is a challenging math contest benchmark featuring the most difficult mathematical problems designed to test advanced reasoning and problem-solving abilities of AI models. It focuses on olympiad-level mathematics and complex multi-step mathematical reasoning.", + "metric_id": "llm_stats.matharena-apex.score", + "metric_name": "MathArena Apex score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "matharena-apex", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MathArena Apex", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.313, + "details": { + "raw_score": "0.313", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "matharena-apex", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/matharena-apex\",\"https://api.llm-stats.com/leaderboard/benchmarks/matharena-apex\"]", + "raw_score_id": "matharena-apex::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mathverse::mathverse-seed-2.1-pro", + "evaluation_name": "llm_stats.mathverse", + "source_data": { + "dataset_name": "MathVerse", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/mathverse", + "https://api.llm-stats.com/leaderboard/benchmarks/mathverse" + ], + "additional_details": { + "raw_benchmark_id": "mathverse", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MathVerse evaluates multimodal mathematical reasoning, testing whether models genuinely interpret visual math diagrams rather than relying on text.", + "metric_id": "llm_stats.mathverse.score", + "metric_name": "MathVerse score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mathverse", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MathVerse", + "raw_categories": "[\"math\",\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.897, + "details": { + "raw_score": "0.897", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "mathverse", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/mathverse\",\"https://api.llm-stats.com/leaderboard/benchmarks/mathverse\"]", + "raw_score_id": "mathverse::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mathvision::mathvision-seed-2.1-pro", + "evaluation_name": "llm_stats.mathvision", + "source_data": { + "dataset_name": "MathVision", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/mathvision", + "https://api.llm-stats.com/leaderboard/benchmarks/mathvision" + ], + "additional_details": { + "raw_benchmark_id": "mathvision", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MATH-Vision is a dataset designed to measure multimodal mathematical reasoning capabilities. It focuses on evaluating how well models can solve mathematical problems that require both visual understanding and mathematical reasoning, bridging the gap between visual and mathematical domains.", + "metric_id": "llm_stats.mathvision.score", + "metric_name": "MathVision score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mathvision", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MathVision", + "raw_categories": "[\"math\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "31" + } + }, + "score_details": { + "score": 0.945, + "details": { + "raw_score": "0.945", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "mathvision", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/mathvision\",\"https://api.llm-stats.com/leaderboard/benchmarks/mathvision\"]", + "raw_score_id": "mathvision::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mathvista::mathvista-seed-2.1-pro", + "evaluation_name": "llm_stats.mathvista", + "source_data": { + "dataset_name": "MathVista", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/mathvista", + "https://api.llm-stats.com/leaderboard/benchmarks/mathvista" + ], + "additional_details": { + "raw_benchmark_id": "mathvista", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MathVista evaluates mathematical reasoning of foundation models in visual contexts. It consists of 6,141 examples derived from 28 existing multimodal datasets and 3 newly created datasets (IQTest, FunctionQA, and PaperQA), combining challenges from diverse mathematical and visual tasks to assess models' ability to understand complex figures and perform rigorous reasoning.", + "metric_id": "llm_stats.mathvista.score", + "metric_name": "MathVista score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mathvista", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MathVista", + "raw_categories": "[\"math\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "39" + } + }, + "score_details": { + "score": 0.907, + "details": { + "raw_score": "0.907", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "mathvista", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/mathvista\",\"https://api.llm-stats.com/leaderboard/benchmarks/mathvista\"]", + "raw_score_id": "mathvista::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mcp-atlas::mcp-atlas-seed-2.1-pro", + "evaluation_name": "llm_stats.mcp-atlas", + "source_data": { + "dataset_name": "MCP Atlas", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/mcp-atlas", + "https://api.llm-stats.com/leaderboard/benchmarks/mcp-atlas" + ], + "additional_details": { + "raw_benchmark_id": "mcp-atlas", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MCP Atlas is a benchmark for evaluating AI models on scaled tool use capabilities, measuring how well models can coordinate and utilize multiple tools across complex multi-step tasks.", + "metric_id": "llm_stats.mcp-atlas.score", + "metric_name": "MCP Atlas score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mcp-atlas", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MCP Atlas", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.838, + "details": { + "raw_score": "0.838", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "mcp-atlas", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/mcp-atlas\",\"https://api.llm-stats.com/leaderboard/benchmarks/mcp-atlas\"]", + "raw_score_id": "mcp-atlas::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "measurebench::measurebench-seed-2.1-pro", + "evaluation_name": "llm_stats.measurebench", + "source_data": { + "dataset_name": "MeasureBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/measurebench", + "https://api.llm-stats.com/leaderboard/benchmarks/measurebench" + ], + "additional_details": { + "raw_benchmark_id": "measurebench", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MeasureBench evaluates multimodal models on visual measurement and quantitative perception tasks across both real and synthetic imagery, reported as the average over the two settings.", + "metric_id": "llm_stats.measurebench.score", + "metric_name": "MeasureBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "measurebench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MeasureBench", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.629, + "details": { + "raw_score": "0.629", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "measurebench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/measurebench\",\"https://api.llm-stats.com/leaderboard/benchmarks/measurebench\"]", + "raw_score_id": "measurebench::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "minerva::minerva-seed-2.1-pro", + "evaluation_name": "llm_stats.minerva", + "source_data": { + "dataset_name": "Minerva", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/minerva", + "https://api.llm-stats.com/leaderboard/benchmarks/minerva" + ], + "additional_details": { + "raw_benchmark_id": "minerva", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Minerva is a benchmark for complex video reasoning, evaluating models on multi-step reasoning over long and information-dense video content.", + "metric_id": "llm_stats.minerva.score", + "metric_name": "Minerva score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "minerva", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Minerva", + "raw_categories": "[\"multimodal\",\"reasoning\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.707, + "details": { + "raw_score": "0.707", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "minerva", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/minerva\",\"https://api.llm-stats.com/leaderboard/benchmarks/minerva\"]", + "raw_score_id": "minerva::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlongbench-128k::mmlongbench-128k-seed-2.1-pro", + "evaluation_name": "llm_stats.mmlongbench-128k", + "source_data": { + "dataset_name": "MMLongBench-128K", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/mmlongbench-128k", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlongbench-128k" + ], + "additional_details": { + "raw_benchmark_id": "mmlongbench-128k", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MMLongBench-128K evaluates multimodal long-context understanding at a 128K token context length, testing how well vision-language models reason over very long mixed text and image inputs.", + "metric_id": "llm_stats.mmlongbench-128k.score", + "metric_name": "MMLongBench-128K score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlongbench-128k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLongBench-128K", + "raw_categories": "[\"long_context\",\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.783, + "details": { + "raw_score": "0.783", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "mmlongbench-128k", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/mmlongbench-128k\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlongbench-128k\"]", + "raw_score_id": "mmlongbench-128k::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmmu-pro::mmmu-pro-seed-2.1-pro", + "evaluation_name": "llm_stats.mmmu-pro", + "source_data": { + "dataset_name": "MMMU-Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/mmmu-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/mmmu-pro" + ], + "additional_details": { + "raw_benchmark_id": "mmmu-pro", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A more robust multi-discipline multimodal understanding benchmark that enhances MMMU through a three-step process: filtering text-only answerable questions, augmenting candidate options, and introducing vision-only input settings. Achieves significantly lower model performance (16.8-26.9%) compared to original MMMU, providing more rigorous evaluation that closely mimics real-world scenarios.", + "metric_id": "llm_stats.mmmu-pro.score", + "metric_name": "MMMU-Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmmu-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMMU-Pro", + "raw_categories": "[\"multimodal\",\"reasoning\",\"general\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "60" + } + }, + "score_details": { + "score": 0.827, + "details": { + "raw_score": "0.827", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "mmmu-pro", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/mmmu-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmmu-pro\"]", + "raw_score_id": "mmmu-pro::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmsibench::mmsibench-seed-2.1-pro", + "evaluation_name": "llm_stats.mmsibench", + "source_data": { + "dataset_name": "MMSIBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/mmsibench", + "https://api.llm-stats.com/leaderboard/benchmarks/mmsibench" + ], + "additional_details": { + "raw_benchmark_id": "mmsibench", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MMSIBench is a multimodal spatial-intelligence benchmark evaluating spatial reasoning over images.", + "metric_id": "llm_stats.mmsibench.score", + "metric_name": "MMSIBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmsibench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMSIBench", + "raw_categories": "[\"multimodal\",\"spatial_reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.359, + "details": { + "raw_score": "0.359", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "mmsibench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/mmsibench\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmsibench\"]", + "raw_score_id": "mmsibench::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mobileworld::mobileworld-seed-2.1-pro", + "evaluation_name": "llm_stats.mobileworld", + "source_data": { + "dataset_name": "MobileWorld", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/mobileworld", + "https://api.llm-stats.com/leaderboard/benchmarks/mobileworld" + ], + "additional_details": { + "raw_benchmark_id": "mobileworld", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MobileWorld is a benchmark for evaluating multimodal agents on real mobile-device tasks, testing GUI grounding, navigation, and multi-step task completion in mobile environments.", + "metric_id": "llm_stats.mobileworld.score", + "metric_name": "MobileWorld score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mobileworld", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MobileWorld", + "raw_categories": "[\"multimodal\",\"agents\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.731, + "details": { + "raw_score": "0.731", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "mobileworld", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/mobileworld\",\"https://api.llm-stats.com/leaderboard/benchmarks/mobileworld\"]", + "raw_score_id": "mobileworld::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "motionbench::motionbench-seed-2.1-pro", + "evaluation_name": "llm_stats.motionbench", + "source_data": { + "dataset_name": "MotionBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/motionbench", + "https://api.llm-stats.com/leaderboard/benchmarks/motionbench" + ], + "additional_details": { + "raw_benchmark_id": "motionbench", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MotionBench is a benchmark for evaluating multimodal models on motion understanding in videos, testing the ability to comprehend temporal dynamics, movement patterns, and action sequences.", + "metric_id": "llm_stats.motionbench.score", + "metric_name": "MotionBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "motionbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MotionBench", + "raw_categories": "[\"multimodal\",\"reasoning\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.749, + "details": { + "raw_score": "0.749", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "motionbench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/motionbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/motionbench\"]", + "raw_score_id": "motionbench::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "msqa::msqa-seed-2.1-pro", + "evaluation_name": "llm_stats.msqa", + "source_data": { + "dataset_name": "MSQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/msqa", + "https://api.llm-stats.com/leaderboard/benchmarks/msqa" + ], + "additional_details": { + "raw_benchmark_id": "msqa", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MSQA is a multilingual question-answering benchmark that measures knowledge and reasoning across a diverse set of languages.", + "metric_id": "llm_stats.msqa.score", + "metric_name": "MSQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "msqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MSQA", + "raw_categories": "[\"knowledge\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.502, + "details": { + "raw_score": "0.502", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "msqa", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/msqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/msqa\"]", + "raw_score_id": "msqa::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "nl2repo::nl2repo-seed-2.1-pro", + "evaluation_name": "llm_stats.nl2repo", + "source_data": { + "dataset_name": "NL2Repo", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/nl2repo", + "https://api.llm-stats.com/leaderboard/benchmarks/nl2repo" + ], + "additional_details": { + "raw_benchmark_id": "nl2repo", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "NL2Repo evaluates long-horizon coding capabilities including repository-level understanding, where models must generate or modify code across entire repositories from natural language specifications.", + "metric_id": "llm_stats.nl2repo.score", + "metric_name": "NL2Repo score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "nl2repo", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "NL2Repo", + "raw_categories": "[\"agents\",\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.47, + "details": { + "raw_score": "0.47", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "nl2repo", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/nl2repo\",\"https://api.llm-stats.com/leaderboard/benchmarks/nl2repo\"]", + "raw_score_id": "nl2repo::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ocrbench-v2::ocrbench-v2-seed-2.1-pro", + "evaluation_name": "llm_stats.ocrbench-v2", + "source_data": { + "dataset_name": "OCRBench_V2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/ocrbench-v2", + "https://api.llm-stats.com/leaderboard/benchmarks/ocrbench-v2" + ], + "additional_details": { + "raw_benchmark_id": "ocrbench-v2", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OCRBench v2: Enhanced large-scale bilingual benchmark for evaluating Large Multimodal Models on visual text localization and reasoning with 10,000 human-verified question-answering pairs across 8 core OCR capabilities", + "metric_id": "llm_stats.ocrbench-v2.score", + "metric_name": "OCRBench_V2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ocrbench-v2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OCRBench_V2", + "raw_categories": "[\"image_to_text\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.632, + "details": { + "raw_score": "0.632", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "ocrbench-v2", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/ocrbench-v2\",\"https://api.llm-stats.com/leaderboard/benchmarks/ocrbench-v2\"]", + "raw_score_id": "ocrbench-v2::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "officeqa-pro::officeqa-pro-seed-2.1-pro", + "evaluation_name": "llm_stats.officeqa-pro", + "source_data": { + "dataset_name": "OfficeQA Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/officeqa-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/officeqa-pro" + ], + "additional_details": { + "raw_benchmark_id": "officeqa-pro", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OfficeQA Pro evaluates AI models on professional knowledge-work questions and tasks drawn from real office workflows, including document analysis, spreadsheet reasoning, and information synthesis across business domains.", + "metric_id": "llm_stats.officeqa-pro.score", + "metric_name": "OfficeQA Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "officeqa-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OfficeQA Pro", + "raw_categories": "[\"reasoning\",\"general\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.722, + "details": { + "raw_score": "0.722", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "officeqa-pro", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/officeqa-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/officeqa-pro\"]", + "raw_score_id": "officeqa-pro::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "onemillion-bench::onemillion-bench-seed-2.1-pro", + "evaluation_name": "llm_stats.onemillion-bench", + "source_data": { + "dataset_name": "OneMillion Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/onemillion-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/onemillion-bench" + ], + "additional_details": { + "raw_benchmark_id": "onemillion-bench", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OneMillion Bench evaluates AI agents on high-economic-value tasks that require sustained, reliable execution across long-horizon real-world workflows.", + "metric_id": "llm_stats.onemillion-bench.score", + "metric_name": "OneMillion Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "onemillion-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OneMillion Bench", + "raw_categories": "[\"long_context\",\"reasoning\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.688, + "details": { + "raw_score": "0.688", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "onemillion-bench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/onemillion-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/onemillion-bench\"]", + "raw_score_id": "onemillion-bench::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "osworld::osworld-seed-2.1-pro", + "evaluation_name": "llm_stats.osworld", + "source_data": { + "dataset_name": "OSWorld", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/osworld", + "https://api.llm-stats.com/leaderboard/benchmarks/osworld" + ], + "additional_details": { + "raw_benchmark_id": "osworld", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OSWorld: The first-of-its-kind scalable, real computer environment for multimodal agents, supporting task setup, execution-based evaluation, and interactive learning across Ubuntu, Windows, and macOS with 369 computer tasks involving real web and desktop applications, OS file I/O, and multi-application workflows", + "metric_id": "llm_stats.osworld.score", + "metric_name": "OSWorld score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "osworld", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OSWorld", + "raw_categories": "[\"multimodal\",\"general\",\"agents\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "20" + } + }, + "score_details": { + "score": 0.788, + "details": { + "raw_score": "0.788", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "osworld", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/osworld\",\"https://api.llm-stats.com/leaderboard/benchmarks/osworld\"]", + "raw_score_id": "osworld::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ovbench::ovbench-seed-2.1-pro", + "evaluation_name": "llm_stats.ovbench", + "source_data": { + "dataset_name": "OVBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/ovbench", + "https://api.llm-stats.com/leaderboard/benchmarks/ovbench" + ], + "additional_details": { + "raw_benchmark_id": "ovbench", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OVBench is an online video understanding benchmark that evaluates a model's ability to perceive, memorize, and reason about real-time video streams as they unfold.", + "metric_id": "llm_stats.ovbench.score", + "metric_name": "OVBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ovbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OVBench", + "raw_categories": "[\"multimodal\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.7, + "details": { + "raw_score": "0.7", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "ovbench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/ovbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/ovbench\"]", + "raw_score_id": "ovbench::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ovobench::ovobench-seed-2.1-pro", + "evaluation_name": "llm_stats.ovobench", + "source_data": { + "dataset_name": "OVOBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/ovobench", + "https://api.llm-stats.com/leaderboard/benchmarks/ovobench" + ], + "additional_details": { + "raw_benchmark_id": "ovobench", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OVOBench (Online Video Online Benchmark) evaluates streaming video understanding, testing a model's ability to perceive and respond to video content in real time as it unfolds.", + "metric_id": "llm_stats.ovobench.score", + "metric_name": "OVOBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ovobench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OVOBench", + "raw_categories": "[\"multimodal\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.807, + "details": { + "raw_score": "0.807", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "ovobench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/ovobench\",\"https://api.llm-stats.com/leaderboard/benchmarks/ovobench\"]", + "raw_score_id": "ovobench::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "posttrainbench::posttrainbench-seed-2.1-pro", + "evaluation_name": "llm_stats.posttrainbench", + "source_data": { + "dataset_name": "PostTrainBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/posttrainbench", + "https://api.llm-stats.com/leaderboard/benchmarks/posttrainbench" + ], + "additional_details": { + "raw_benchmark_id": "posttrainbench", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "PostTrainBench evaluates a model's ability to autonomously post-train base models. Given pretrain-only base models, the agent must complete the full pipeline of data synthesis, training, evaluation, and iteration within a time budget, scored across downstream benchmarks such as AIME2025, BFCL, GPQA Main, GSM8K, and HumanEval.", + "metric_id": "llm_stats.posttrainbench.score", + "metric_name": "PostTrainBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "posttrainbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PostTrainBench", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"systems\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.165, + "details": { + "raw_score": "0.165", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "posttrainbench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/posttrainbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/posttrainbench\"]", + "raw_score_id": "posttrainbench::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "presentbench::presentbench-seed-2.1-pro", + "evaluation_name": "llm_stats.presentbench", + "source_data": { + "dataset_name": "PresentBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/presentbench", + "https://api.llm-stats.com/leaderboard/benchmarks/presentbench" + ], + "additional_details": { + "raw_benchmark_id": "presentbench", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "PresentBench evaluates AI agents on producing presentation-style deliverables, such as generating lesson-plan slides and structured documents from source materials.", + "metric_id": "llm_stats.presentbench.score", + "metric_name": "PresentBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "presentbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PresentBench", + "raw_categories": "[\"reasoning\",\"general\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.546, + "details": { + "raw_score": "0.546", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "presentbench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/presentbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/presentbench\"]", + "raw_score_id": "presentbench::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "program-bench::program-bench-seed-2.1-pro", + "evaluation_name": "llm_stats.program-bench", + "source_data": { + "dataset_name": "Program Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/program-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/program-bench" + ], + "additional_details": { + "raw_benchmark_id": "program-bench", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Program Bench evaluates code-generation agents by asking them to recreate a program's behavior from only a compiled binary and documentation. It spans 200 tasks from small CLI tools to large systems such as FFmpeg and SQLite, with submissions judged against more than 248,000 fuzz-generated behavioral tests.", + "metric_id": "llm_stats.program-bench.score", + "metric_name": "Program Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "program-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Program Bench", + "raw_categories": "[\"agents\",\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.503, + "details": { + "raw_score": "0.503", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "program-bench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/program-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/program-bench\"]", + "raw_score_id": "program-bench::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "realworldqa::realworldqa-seed-2.1-pro", + "evaluation_name": "llm_stats.realworldqa", + "source_data": { + "dataset_name": "RealWorldQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/realworldqa", + "https://api.llm-stats.com/leaderboard/benchmarks/realworldqa" + ], + "additional_details": { + "raw_benchmark_id": "realworldqa", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "RealWorldQA is a benchmark designed to evaluate basic real-world spatial understanding capabilities of multimodal models. The initial release consists of over 700 anonymized images taken from vehicles and other real-world scenarios, each accompanied by a question and easily verifiable answer. Released by xAI as part of their Grok-1.5 Vision preview to test models' ability to understand natural scenes and spatial relationships in everyday visual contexts.", + "metric_id": "llm_stats.realworldqa.score", + "metric_name": "RealWorldQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "realworldqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "RealWorldQA", + "raw_categories": "[\"spatial_reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "25" + } + }, + "score_details": { + "score": 0.867, + "details": { + "raw_score": "0.867", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "realworldqa", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/realworldqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/realworldqa\"]", + "raw_score_id": "realworldqa::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "repo-env::repo-env-seed-2.1-pro", + "evaluation_name": "llm_stats.repo-env", + "source_data": { + "dataset_name": "Repo Env", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/repo-env", + "https://api.llm-stats.com/leaderboard/benchmarks/repo-env" + ], + "additional_details": { + "raw_benchmark_id": "repo-env", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Repo Env evaluates an agent's ability to set up, configure, and run real repositories, including dependency resolution and environment management.", + "metric_id": "llm_stats.repo-env.score", + "metric_name": "Repo Env score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "repo-env", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Repo Env", + "raw_categories": "[\"agents\",\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.55, + "details": { + "raw_score": "0.55", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "repo-env", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/repo-env\",\"https://api.llm-stats.com/leaderboard/benchmarks/repo-env\"]", + "raw_score_id": "repo-env::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "scicode::scicode-seed-2.1-pro", + "evaluation_name": "llm_stats.scicode", + "source_data": { + "dataset_name": "SciCode", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/scicode", + "https://api.llm-stats.com/leaderboard/benchmarks/scicode" + ], + "additional_details": { + "raw_benchmark_id": "scicode", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SciCode is a research coding benchmark curated by scientists that challenges language models to code solutions for scientific problems. It contains 338 subproblems decomposed from 80 challenging main problems across 16 natural science sub-fields including mathematics, physics, chemistry, biology, and materials science. Problems require knowledge recall, reasoning, and code synthesis skills.", + "metric_id": "llm_stats.scicode.score", + "metric_name": "SciCode score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "scicode", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SciCode", + "raw_categories": "[\"math\",\"physics\",\"reasoning\",\"biology\",\"chemistry\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.598, + "details": { + "raw_score": "0.598", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "scicode", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/scicode\",\"https://api.llm-stats.com/leaderboard/benchmarks/scicode\"]", + "raw_score_id": "scicode::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "seedclawbench::seedclawbench-seed-2.1-pro", + "evaluation_name": "llm_stats.seedclawbench", + "source_data": { + "dataset_name": "SeedClawBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/seedclawbench", + "https://api.llm-stats.com/leaderboard/benchmarks/seedclawbench" + ], + "additional_details": { + "raw_benchmark_id": "seedclawbench", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SeedClawBench is an agentic coding benchmark measuring overall model performance on real-world, tool-using software development tasks.", + "metric_id": "llm_stats.seedclawbench.score", + "metric_name": "SeedClawBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "seedclawbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SeedClawBench", + "raw_categories": "[\"agents\",\"coding\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.666, + "details": { + "raw_score": "0.666", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "seedclawbench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/seedclawbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/seedclawbench\"]", + "raw_score_id": "seedclawbench::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "simplevqa::simplevqa-seed-2.1-pro", + "evaluation_name": "llm_stats.simplevqa", + "source_data": { + "dataset_name": "SimpleVQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/simplevqa", + "https://api.llm-stats.com/leaderboard/benchmarks/simplevqa" + ], + "additional_details": { + "raw_benchmark_id": "simplevqa", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SimpleVQA is a visual question answering benchmark focused on simple queries.", + "metric_id": "llm_stats.simplevqa.score", + "metric_name": "SimpleVQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "simplevqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SimpleVQA", + "raw_categories": "[\"image_to_text\",\"multimodal\",\"general\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.745, + "details": { + "raw_score": "0.745", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "simplevqa", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/simplevqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/simplevqa\"]", + "raw_score_id": "simplevqa::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "superchem::superchem-seed-2.1-pro", + "evaluation_name": "llm_stats.superchem", + "source_data": { + "dataset_name": "SuperChem", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/superchem", + "https://api.llm-stats.com/leaderboard/benchmarks/superchem" + ], + "additional_details": { + "raw_benchmark_id": "superchem", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SuperChem is a benchmark of advanced chemistry problems requiring expert-level domain knowledge and reasoning.", + "metric_id": "llm_stats.superchem.score", + "metric_name": "SuperChem score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "superchem", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SuperChem", + "raw_categories": "[\"knowledge\",\"reasoning\",\"chemistry\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.598, + "details": { + "raw_score": "0.598", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "superchem", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/superchem\",\"https://api.llm-stats.com/leaderboard/benchmarks/superchem\"]", + "raw_score_id": "superchem::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "supergpqa::supergpqa-seed-2.1-pro", + "evaluation_name": "llm_stats.supergpqa", + "source_data": { + "dataset_name": "SuperGPQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/supergpqa", + "https://api.llm-stats.com/leaderboard/benchmarks/supergpqa" + ], + "additional_details": { + "raw_benchmark_id": "supergpqa", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SuperGPQA is a comprehensive benchmark that evaluates large language models across 285 graduate-level academic disciplines. The benchmark contains 25,957 questions covering 13 broad disciplinary areas including Engineering, Medicine, Science, and Law, with specialized fields in light industry, agriculture, and service-oriented domains. It employs a Human-LLM collaborative filtering mechanism with over 80 expert annotators to create challenging questions that assess graduate-level knowledge and reasoning capabilities.", + "metric_id": "llm_stats.supergpqa.score", + "metric_name": "SuperGPQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "supergpqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SuperGPQA", + "raw_categories": "[\"legal\",\"math\",\"physics\",\"reasoning\",\"finance\",\"general\",\"healthcare\",\"chemistry\",\"economics\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "34" + } + }, + "score_details": { + "score": 0.708, + "details": { + "raw_score": "0.708", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "supergpqa", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/supergpqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/supergpqa\"]", + "raw_score_id": "supergpqa::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-atlas::swe-atlas-seed-2.1-pro", + "evaluation_name": "llm_stats.swe-atlas", + "source_data": { + "dataset_name": "SWE-Atlas", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/swe-atlas", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-atlas" + ], + "additional_details": { + "raw_benchmark_id": "swe-atlas", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SWE-Atlas is a software engineering benchmark focused on debugging, evaluating a model's ability to localize and fix bugs in real-world codebases.", + "metric_id": "llm_stats.swe-atlas.score", + "metric_name": "SWE-Atlas score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-atlas", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Atlas", + "raw_categories": "[\"agents\",\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.352, + "details": { + "raw_score": "0.352", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "swe-atlas", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/swe-atlas\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-atlas\"]", + "raw_score_id": "swe-atlas::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-pro::swe-bench-pro-seed-2.1-pro", + "evaluation_name": "llm_stats.swe-bench-pro", + "source_data": { + "dataset_name": "SWE-Bench Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/swe-bench-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-pro" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-pro", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SWE-Bench Pro is an advanced version of SWE-Bench that evaluates language models on complex, real-world software engineering tasks requiring extended reasoning and multi-step problem solving.", + "metric_id": "llm_stats.swe-bench-pro.score", + "metric_name": "SWE-Bench Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Bench Pro", + "raw_categories": "[\"reasoning\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "35" + } + }, + "score_details": { + "score": 0.575, + "details": { + "raw_score": "0.575", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "swe-bench-pro", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/swe-bench-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-pro\"]", + "raw_score_id": "swe-bench-pro::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2.1::terminal-bench-2.1-seed-2.1-pro", + "evaluation_name": "llm_stats.terminal-bench-2.1", + "source_data": { + "dataset_name": "Terminal-Bench 2.1", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/terminal-bench-2.1", + "https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2.1" + ], + "additional_details": { + "raw_benchmark_id": "terminal-bench-2.1", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Terminal-Bench 2.1 is an updated release of the Terminal-Bench benchmark that tests AI agents' ability to operate a computer via the terminal. It evaluates how well models handle real-world, end-to-end tasks autonomously, including compiling code, training models, setting up servers, system administration, data science workflows, and security tasks.", + "metric_id": "llm_stats.terminal-bench-2.1.score", + "metric_name": "Terminal-Bench 2.1 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "terminal-bench-2.1", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Terminal-Bench 2.1", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.71, + "details": { + "raw_score": "0.71", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "terminal-bench-2.1", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/terminal-bench-2.1\",\"https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2.1\"]", + "raw_score_id": "terminal-bench-2.1::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "tomato::tomato-seed-2.1-pro", + "evaluation_name": "llm_stats.tomato", + "source_data": { + "dataset_name": "TOMATO", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/tomato", + "https://api.llm-stats.com/leaderboard/benchmarks/tomato" + ], + "additional_details": { + "raw_benchmark_id": "tomato", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "TOMATO (Temporal Reasoning Multimodal Evaluation) assesses multimodal models on motion and temporal perception in video, testing understanding of actions, motion, and changes over time.", + "metric_id": "llm_stats.tomato.score", + "metric_name": "TOMATO score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "tomato", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TOMATO", + "raw_categories": "[\"multimodal\",\"reasoning\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.795, + "details": { + "raw_score": "0.795", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "tomato", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/tomato\",\"https://api.llm-stats.com/leaderboard/benchmarks/tomato\"]", + "raw_score_id": "tomato::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "toolathlon::toolathlon-seed-2.1-pro", + "evaluation_name": "llm_stats.toolathlon", + "source_data": { + "dataset_name": "Toolathlon", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/toolathlon", + "https://api.llm-stats.com/leaderboard/benchmarks/toolathlon" + ], + "additional_details": { + "raw_benchmark_id": "toolathlon", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Tool Decathlon is a comprehensive benchmark for evaluating AI agents' ability to use multiple tools across diverse task categories. It measures proficiency in tool selection, sequencing, and execution across ten different tool-use scenarios.", + "metric_id": "llm_stats.toolathlon.score", + "metric_name": "Toolathlon score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "toolathlon", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Toolathlon", + "raw_categories": "[\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "24" + } + }, + "score_details": { + "score": 0.506, + "details": { + "raw_score": "0.506", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "toolathlon", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/toolathlon\",\"https://api.llm-stats.com/leaderboard/benchmarks/toolathlon\"]", + "raw_score_id": "toolathlon::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "trae-code-gen::trae-code-gen-seed-2.1-pro", + "evaluation_name": "llm_stats.trae-code-gen", + "source_data": { + "dataset_name": "Trae Code Gen", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/trae-code-gen", + "https://api.llm-stats.com/leaderboard/benchmarks/trae-code-gen" + ], + "additional_details": { + "raw_benchmark_id": "trae-code-gen", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Trae Code Gen is a component of Trae Agent Bench that evaluates implementing new functionality across multiple programming languages in containerized, runnable repositories.", + "metric_id": "llm_stats.trae-code-gen.score", + "metric_name": "Trae Code Gen score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "trae-code-gen", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Trae Code Gen", + "raw_categories": "[\"agents\",\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.624, + "details": { + "raw_score": "0.624", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "trae-code-gen", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/trae-code-gen\",\"https://api.llm-stats.com/leaderboard/benchmarks/trae-code-gen\"]", + "raw_score_id": "trae-code-gen::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "trae-error-fix::trae-error-fix-seed-2.1-pro", + "evaluation_name": "llm_stats.trae-error-fix", + "source_data": { + "dataset_name": "Trae Error Fix", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/trae-error-fix", + "https://api.llm-stats.com/leaderboard/benchmarks/trae-error-fix" + ], + "additional_details": { + "raw_benchmark_id": "trae-error-fix", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Trae Error Fix is a component of Trae Agent Bench that evaluates fixing existing code across multiple programming languages in containerized, runnable repositories.", + "metric_id": "llm_stats.trae-error-fix.score", + "metric_name": "Trae Error Fix score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "trae-error-fix", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Trae Error Fix", + "raw_categories": "[\"agents\",\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.633, + "details": { + "raw_score": "0.633", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "trae-error-fix", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/trae-error-fix\",\"https://api.llm-stats.com/leaderboard/benchmarks/trae-error-fix\"]", + "raw_score_id": "trae-error-fix::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "treebench::treebench-seed-2.1-pro", + "evaluation_name": "llm_stats.treebench", + "source_data": { + "dataset_name": "TreeBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/treebench", + "https://api.llm-stats.com/leaderboard/benchmarks/treebench" + ], + "additional_details": { + "raw_benchmark_id": "treebench", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "TreeBench evaluates visual grounded reasoning, requiring models to localize and reason about fine-grained visual details.", + "metric_id": "llm_stats.treebench.score", + "metric_name": "TreeBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "treebench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TreeBench", + "raw_categories": "[\"multimodal\",\"reasoning\",\"spatial_reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.711, + "details": { + "raw_score": "0.711", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "treebench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/treebench\",\"https://api.llm-stats.com/leaderboard/benchmarks/treebench\"]", + "raw_score_id": "treebench::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "tvbench::tvbench-seed-2.1-pro", + "evaluation_name": "llm_stats.tvbench", + "source_data": { + "dataset_name": "TVBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/tvbench", + "https://api.llm-stats.com/leaderboard/benchmarks/tvbench" + ], + "additional_details": { + "raw_benchmark_id": "tvbench", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "TVBench is a temporal video understanding benchmark evaluating reasoning over actions, events, and temporal dynamics in videos.", + "metric_id": "llm_stats.tvbench.score", + "metric_name": "TVBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "tvbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TVBench", + "raw_categories": "[\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.805, + "details": { + "raw_score": "0.805", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "tvbench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/tvbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/tvbench\"]", + "raw_score_id": "tvbench::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "video-mme::video-mme-seed-2.1-pro", + "evaluation_name": "llm_stats.video-mme", + "source_data": { + "dataset_name": "Video-MME", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/video-mme", + "https://api.llm-stats.com/leaderboard/benchmarks/video-mme" + ], + "additional_details": { + "raw_benchmark_id": "video-mme", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Video-MME is the first-ever comprehensive evaluation benchmark of Multi-modal Large Language Models (MLLMs) in video analysis. It features 900 videos totaling 254 hours with 2,700 human-annotated question-answer pairs across 6 primary visual domains (Knowledge, Film & Television, Sports Competition, Life Record, Multilingual, and others) and 30 subfields. The benchmark evaluates models across diverse temporal dimensions (11 seconds to 1 hour), integrates multi-modal inputs including video frames, subtitles, and audio, and uses rigorous manual labeling by expert annotators for precise assessment.", + "metric_id": "llm_stats.video-mme.score", + "metric_name": "Video-MME score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "video-mme", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Video-MME", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "17" + } + }, + "score_details": { + "score": 0.892, + "details": { + "raw_score": "0.892", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "video-mme", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/video-mme\",\"https://api.llm-stats.com/leaderboard/benchmarks/video-mme\"]", + "raw_score_id": "video-mme::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "videoholmes::videoholmes-seed-2.1-pro", + "evaluation_name": "llm_stats.videoholmes", + "source_data": { + "dataset_name": "VideoHolmes", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/videoholmes", + "https://api.llm-stats.com/leaderboard/benchmarks/videoholmes" + ], + "additional_details": { + "raw_benchmark_id": "videoholmes", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VideoHolmes evaluates video understanding and reasoning capabilities in multimodal models.", + "metric_id": "llm_stats.videoholmes.score", + "metric_name": "VideoHolmes score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "videoholmes", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VideoHolmes", + "raw_categories": "[\"multimodal\",\"reasoning\",\"video\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.682, + "details": { + "raw_score": "0.682", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "videoholmes", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/videoholmes\",\"https://api.llm-stats.com/leaderboard/benchmarks/videoholmes\"]", + "raw_score_id": "videoholmes::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "videosimpleqa::videosimpleqa-seed-2.1-pro", + "evaluation_name": "llm_stats.videosimpleqa", + "source_data": { + "dataset_name": "VideoSimpleQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/videosimpleqa", + "https://api.llm-stats.com/leaderboard/benchmarks/videosimpleqa" + ], + "additional_details": { + "raw_benchmark_id": "videosimpleqa", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VideoSimpleQA evaluates factual knowledge grounded in video content, measuring how accurately models answer short, fact-seeking questions about videos.", + "metric_id": "llm_stats.videosimpleqa.score", + "metric_name": "VideoSimpleQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "videosimpleqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VideoSimpleQA", + "raw_categories": "[\"knowledge\",\"multimodal\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.764, + "details": { + "raw_score": "0.764", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "videosimpleqa", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/videosimpleqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/videosimpleqa\"]", + "raw_score_id": "videosimpleqa::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "visfactor::visfactor-seed-2.1-pro", + "evaluation_name": "llm_stats.visfactor", + "source_data": { + "dataset_name": "VisFactor", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/visfactor", + "https://api.llm-stats.com/leaderboard/benchmarks/visfactor" + ], + "additional_details": { + "raw_benchmark_id": "visfactor", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VisFactor is a benchmark evaluating fine-grained visual factor perception and reasoning over images.", + "metric_id": "llm_stats.visfactor.score", + "metric_name": "VisFactor score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "visfactor", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VisFactor", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.514, + "details": { + "raw_score": "0.514", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "visfactor", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/visfactor\",\"https://api.llm-stats.com/leaderboard/benchmarks/visfactor\"]", + "raw_score_id": "visfactor::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "visulogic::visulogic-seed-2.1-pro", + "evaluation_name": "llm_stats.visulogic", + "source_data": { + "dataset_name": "VisuLogic", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/visulogic", + "https://api.llm-stats.com/leaderboard/benchmarks/visulogic" + ], + "additional_details": { + "raw_benchmark_id": "visulogic", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VisuLogic evaluates logical reasoning capabilities in visual contexts.", + "metric_id": "llm_stats.visulogic.score", + "metric_name": "VisuLogic score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "visulogic", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VisuLogic", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.543, + "details": { + "raw_score": "0.543", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "visulogic", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/visulogic\",\"https://api.llm-stats.com/leaderboard/benchmarks/visulogic\"]", + "raw_score_id": "visulogic::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "vlmsarebiased::vlmsarebiased-seed-2.1-pro", + "evaluation_name": "llm_stats.vlmsarebiased", + "source_data": { + "dataset_name": "VLMsAreBiased", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/vlmsarebiased", + "https://api.llm-stats.com/leaderboard/benchmarks/vlmsarebiased" + ], + "additional_details": { + "raw_benchmark_id": "vlmsarebiased", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VLMsAreBiased evaluates whether vision-language models rely on visual evidence or fall back on language priors when answering.", + "metric_id": "llm_stats.vlmsarebiased.score", + "metric_name": "VLMsAreBiased score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "vlmsarebiased", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VLMsAreBiased", + "raw_categories": "[\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.836, + "details": { + "raw_score": "0.836", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "vlmsarebiased", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/vlmsarebiased\",\"https://api.llm-stats.com/leaderboard/benchmarks/vlmsarebiased\"]", + "raw_score_id": "vlmsarebiased::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "web-bench::web-bench-seed-2.1-pro", + "evaluation_name": "llm_stats.web-bench", + "source_data": { + "dataset_name": "Web Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/web-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/web-bench" + ], + "additional_details": { + "raw_benchmark_id": "web-bench", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Web Bench evaluates agents on realistic web-development engineering tasks, measuring end-to-end implementation in browser-based workflows.", + "metric_id": "llm_stats.web-bench.score", + "metric_name": "Web Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "web-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Web Bench", + "raw_categories": "[\"agents\",\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.784, + "details": { + "raw_score": "0.784", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "web-bench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/web-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/web-bench\"]", + "raw_score_id": "web-bench::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "wildclawbench::wildclawbench-seed-2.1-pro", + "evaluation_name": "llm_stats.wildclawbench", + "source_data": { + "dataset_name": "WildClawBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/wildclawbench", + "https://api.llm-stats.com/leaderboard/benchmarks/wildclawbench" + ], + "additional_details": { + "raw_benchmark_id": "wildclawbench", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WildClawBench is an agentic coding benchmark from InternLM/Claw-Eval that reports overall model performance on real-world tool-using development tasks.", + "metric_id": "llm_stats.wildclawbench.score", + "metric_name": "WildClawBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "wildclawbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WildClawBench", + "raw_categories": "[\"agents\",\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.617, + "details": { + "raw_score": "0.617", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "wildclawbench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/wildclawbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/wildclawbench\"]", + "raw_score_id": "wildclawbench::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "workspace-bench::workspace-bench-seed-2.1-pro", + "evaluation_name": "llm_stats.workspace-bench", + "source_data": { + "dataset_name": "Workspace Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/workspace-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/workspace-bench" + ], + "additional_details": { + "raw_benchmark_id": "workspace-bench", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Workspace Bench evaluates AI agents on high-economic-value workplace tasks that span multi-step planning, file processing, and tool use across realistic office and productivity workflows.", + "metric_id": "llm_stats.workspace-bench.score", + "metric_name": "Workspace Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "workspace-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Workspace Bench", + "raw_categories": "[\"reasoning\",\"general\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.53, + "details": { + "raw_score": "0.53", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "workspace-bench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/workspace-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/workspace-bench\"]", + "raw_score_id": "workspace-bench::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "worldbench::worldbench-seed-2.1-pro", + "evaluation_name": "llm_stats.worldbench", + "source_data": { + "dataset_name": "WorldBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/worldbench", + "https://api.llm-stats.com/leaderboard/benchmarks/worldbench" + ], + "additional_details": { + "raw_benchmark_id": "worldbench", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WorldBench evaluates real-world visual knowledge and understanding across diverse everyday scenes.", + "metric_id": "llm_stats.worldbench.score", + "metric_name": "WorldBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "worldbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WorldBench", + "raw_categories": "[\"knowledge\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.676, + "details": { + "raw_score": "0.676", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "worldbench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/worldbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/worldbench\"]", + "raw_score_id": "worldbench::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "worldvqa::worldvqa-seed-2.1-pro", + "evaluation_name": "llm_stats.worldvqa", + "source_data": { + "dataset_name": "WorldVQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/worldvqa", + "https://api.llm-stats.com/leaderboard/benchmarks/worldvqa" + ], + "additional_details": { + "raw_benchmark_id": "worldvqa", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WorldVQA is a benchmark designed to evaluate atomic vision-centric world knowledge. It assesses models' ability to understand and reason about visual elements representing real-world knowledge.", + "metric_id": "llm_stats.worldvqa.score", + "metric_name": "WorldVQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "worldvqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WorldVQA", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.53, + "details": { + "raw_score": "0.53", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "worldvqa", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/worldvqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/worldvqa\"]", + "raw_score_id": "worldvqa::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "xdailybench::xdailybench-seed-2.1-pro", + "evaluation_name": "llm_stats.xdailybench", + "source_data": { + "dataset_name": "xDailyBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/xdailybench", + "https://api.llm-stats.com/leaderboard/benchmarks/xdailybench" + ], + "additional_details": { + "raw_benchmark_id": "xdailybench", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "xDailyBench evaluates AI agents on white-collar office work, covering everyday professional tasks such as document handling, consultation, and multi-step productivity workflows.", + "metric_id": "llm_stats.xdailybench.score", + "metric_name": "xDailyBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "xdailybench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "xDailyBench", + "raw_categories": "[\"reasoning\",\"general\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.61, + "details": { + "raw_score": "0.61", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "xdailybench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/xdailybench\",\"https://api.llm-stats.com/leaderboard/benchmarks/xdailybench\"]", + "raw_score_id": "xdailybench::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "zerobench::zerobench-seed-2.1-pro", + "evaluation_name": "llm_stats.zerobench", + "source_data": { + "dataset_name": "ZEROBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-pro", + "https://llm-stats.com/benchmarks/zerobench", + "https://api.llm-stats.com/leaderboard/benchmarks/zerobench" + ], + "additional_details": { + "raw_benchmark_id": "zerobench", + "raw_model_id": "seed-2.1-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ZEROBench is a challenging vision benchmark designed to test models on zero-shot visual understanding tasks.", + "metric_id": "llm_stats.zerobench.score", + "metric_name": "ZEROBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "zerobench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ZEROBench", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.563, + "details": { + "raw_score": "0.563", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-pro", + "raw_benchmark_id": "zerobench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-pro\",\"https://llm-stats.com/benchmarks/zerobench\",\"https://api.llm-stats.com/leaderboard/benchmarks/zerobench\"]", + "raw_score_id": "zerobench::seed-2.1-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/bytedance/seed-2.1-turbo/91ebb4fa-76ec-4909-8177-c01dfbe27dba.json b/data/llm_stats/bytedance/seed-2.1-turbo/91ebb4fa-76ec-4909-8177-c01dfbe27dba.json new file mode 100644 index 0000000000000000000000000000000000000000..9c0d0d01b3b9585f18ce15798888672fe1298b0a --- /dev/null +++ b/data/llm_stats/bytedance/seed-2.1-turbo/91ebb4fa-76ec-4909-8177-c01dfbe27dba.json @@ -0,0 +1,4726 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/bytedance_seed-2.1-turbo/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Seed 2.1 Turbo", + "id": "bytedance/seed-2.1-turbo", + "developer": "bytedance", + "additional_details": { + "raw_id": "seed-2.1-turbo", + "raw_name": "Seed 2.1 Turbo", + "raw_model_id": "seed-2.1-turbo", + "raw_model_name": "Seed 2.1 Turbo", + "raw_organization_id": "bytedance", + "raw_organization_name": "ByteDance", + "raw_release_date": "2026-06-24", + "raw_announcement_date": "2026-06-24", + "raw_multimodal": "true", + "raw_provider_slug": "bytedance", + "raw_provider_name": "ByteDance" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aethercode::aethercode-seed-2.1-turbo", + "evaluation_name": "llm_stats.aethercode", + "source_data": { + "dataset_name": "AetherCode", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/aethercode", + "https://api.llm-stats.com/leaderboard/benchmarks/aethercode" + ], + "additional_details": { + "raw_benchmark_id": "aethercode", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "AetherCode is a competitive-programming benchmark of olympiad-level algorithmic coding problems.", + "metric_id": "llm_stats.aethercode.score", + "metric_name": "AetherCode score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aethercode", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AetherCode", + "raw_categories": "[\"reasoning\",\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.679, + "details": { + "raw_score": "0.679", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "aethercode", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/aethercode\",\"https://api.llm-stats.com/leaderboard/benchmarks/aethercode\"]", + "raw_score_id": "aethercode::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "agent-startup-bench::agent-startup-bench-seed-2.1-turbo", + "evaluation_name": "llm_stats.agent-startup-bench", + "source_data": { + "dataset_name": "Agent Startup Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/agent-startup-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/agent-startup-bench" + ], + "additional_details": { + "raw_benchmark_id": "agent-startup-bench", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Agent Startup Bench measures AI agents on high-economic-value, startup-style tasks that require autonomous planning and execution to deliver practical, verifiable results.", + "metric_id": "llm_stats.agent-startup-bench.score", + "metric_name": "Agent Startup Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "agent-startup-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Agent Startup Bench", + "raw_categories": "[\"reasoning\",\"general\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.54, + "details": { + "raw_score": "0.54", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "agent-startup-bench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/agent-startup-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/agent-startup-bench\"]", + "raw_score_id": "agent-startup-bench::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "apex-agents::apex-agents-seed-2.1-turbo", + "evaluation_name": "llm_stats.apex-agents", + "source_data": { + "dataset_name": "APEX-Agents", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/apex-agents", + "https://api.llm-stats.com/leaderboard/benchmarks/apex-agents" + ], + "additional_details": { + "raw_benchmark_id": "apex-agents", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "APEX-Agents is a benchmark evaluating AI agents on long horizon professional tasks that require sustained reasoning, planning, and execution across complex multi-step workflows.", + "metric_id": "llm_stats.apex-agents.score", + "metric_name": "APEX-Agents score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "apex-agents", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "APEX-Agents", + "raw_categories": "[\"reasoning\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.292, + "details": { + "raw_score": "0.292", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "apex-agents", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/apex-agents\",\"https://api.llm-stats.com/leaderboard/benchmarks/apex-agents\"]", + "raw_score_id": "apex-agents::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "arcagi2::arcagi2-seed-2.1-turbo", + "evaluation_name": "llm_stats.arcagi2", + "source_data": { + "dataset_name": "ArcAGI2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/arcagi2", + "https://api.llm-stats.com/leaderboard/benchmarks/arcagi2" + ], + "additional_details": { + "raw_benchmark_id": "arcagi2", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ARC-AGI-2 is the second-generation Abstraction and Reasoning Corpus benchmark measuring fluid, general reasoning and abstraction.", + "metric_id": "llm_stats.arcagi2.score", + "metric_name": "ArcAGI2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arcagi2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ArcAGI2", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.613, + "details": { + "raw_score": "0.613", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "arcagi2", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/arcagi2\",\"https://api.llm-stats.com/leaderboard/benchmarks/arcagi2\"]", + "raw_score_id": "arcagi2::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "artifacts-bench::artifacts-bench-seed-2.1-turbo", + "evaluation_name": "llm_stats.artifacts-bench", + "source_data": { + "dataset_name": "Artifacts Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/artifacts-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/artifacts-bench" + ], + "additional_details": { + "raw_benchmark_id": "artifacts-bench", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Artifacts Bench evaluates a model's ability to generate visual code artifacts, measuring the quality of generated interactive and visual front-end outputs from natural-language requests.", + "metric_id": "llm_stats.artifacts-bench.score", + "metric_name": "Artifacts Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "artifacts-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Artifacts Bench", + "raw_categories": "[\"frontend_development\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.47, + "details": { + "raw_score": "0.47", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "artifacts-bench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/artifacts-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/artifacts-bench\"]", + "raw_score_id": "artifacts-bench::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "babyvision::babyvision-seed-2.1-turbo", + "evaluation_name": "llm_stats.babyvision", + "source_data": { + "dataset_name": "BabyVision", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/babyvision", + "https://api.llm-stats.com/leaderboard/benchmarks/babyvision" + ], + "additional_details": { + "raw_benchmark_id": "babyvision", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A benchmark for early-stage visual reasoning and perception on child-like vision tasks.", + "metric_id": "llm_stats.babyvision.score", + "metric_name": "BabyVision score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "babyvision", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BabyVision", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.629, + "details": { + "raw_score": "0.629", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "babyvision", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/babyvision\",\"https://api.llm-stats.com/leaderboard/benchmarks/babyvision\"]", + "raw_score_id": "babyvision::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "beyond-aime::beyond-aime-seed-2.1-turbo", + "evaluation_name": "llm_stats.beyond-aime", + "source_data": { + "dataset_name": "Beyond AIME", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/beyond-aime", + "https://api.llm-stats.com/leaderboard/benchmarks/beyond-aime" + ], + "additional_details": { + "raw_benchmark_id": "beyond-aime", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Beyond AIME is a difficult mathematical reasoning benchmark designed to test deeper reasoning chains and harder decomposition than standard AIME-style problem sets.", + "metric_id": "llm_stats.beyond-aime.score", + "metric_name": "Beyond AIME score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "beyond-aime", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Beyond AIME", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.88, + "details": { + "raw_score": "0.88", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "beyond-aime", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/beyond-aime\",\"https://api.llm-stats.com/leaderboard/benchmarks/beyond-aime\"]", + "raw_score_id": "beyond-aime::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "blink::blink-seed-2.1-turbo", + "evaluation_name": "llm_stats.blink", + "source_data": { + "dataset_name": "BLINK", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/blink", + "https://api.llm-stats.com/leaderboard/benchmarks/blink" + ], + "additional_details": { + "raw_benchmark_id": "blink", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BLINK: Multimodal Large Language Models Can See but Not Perceive. A benchmark for multimodal language models focusing on core visual perception abilities. Reformats 14 classic computer vision tasks into 3,807 multiple-choice questions paired with single or multiple images and visual prompting. Tasks include relative depth estimation, visual correspondence, forensics detection, multi-view reasoning, counting, object localization, and spatial reasoning that humans can solve 'within a blink'.", + "metric_id": "llm_stats.blink.score", + "metric_name": "BLINK score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "blink", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BLINK", + "raw_categories": "[\"multimodal\",\"reasoning\",\"spatial_reasoning\",\"3d\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.794, + "details": { + "raw_score": "0.794", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "blink", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/blink\",\"https://api.llm-stats.com/leaderboard/benchmarks/blink\"]", + "raw_score_id": "blink::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "browsecomp::browsecomp-seed-2.1-turbo", + "evaluation_name": "llm_stats.browsecomp", + "source_data": { + "dataset_name": "BrowseComp", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/browsecomp", + "https://api.llm-stats.com/leaderboard/benchmarks/browsecomp" + ], + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BrowseComp is a benchmark comprising 1,266 questions that challenge AI agents to persistently navigate the internet in search of hard-to-find, entangled information. The benchmark measures agents' ability to exercise persistence in information gathering, demonstrate creativity in web navigation, and find concise, verifiable answers. Despite the difficulty of the questions, BrowseComp is simple and easy-to-use, as predicted answers are short and easily verifiable against reference answers.", + "metric_id": "llm_stats.browsecomp.score", + "metric_name": "BrowseComp score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BrowseComp", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "52" + } + }, + "score_details": { + "score": 0.849, + "details": { + "raw_score": "0.849", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "browsecomp", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/browsecomp\",\"https://api.llm-stats.com/leaderboard/benchmarks/browsecomp\"]", + "raw_score_id": "browsecomp::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "chartqapro::chartqapro-seed-2.1-turbo", + "evaluation_name": "llm_stats.chartqapro", + "source_data": { + "dataset_name": "ChartQAPro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/chartqapro", + "https://api.llm-stats.com/leaderboard/benchmarks/chartqapro" + ], + "additional_details": { + "raw_benchmark_id": "chartqapro", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ChartQAPro is a challenging benchmark for question answering over diverse, real-world charts and infographics.", + "metric_id": "llm_stats.chartqapro.score", + "metric_name": "ChartQAPro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "chartqapro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ChartQAPro", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.709, + "details": { + "raw_score": "0.709", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "chartqapro", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/chartqapro\",\"https://api.llm-stats.com/leaderboard/benchmarks/chartqapro\"]", + "raw_score_id": "chartqapro::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "charxiv-d::charxiv-d-seed-2.1-turbo", + "evaluation_name": "llm_stats.charxiv-d", + "source_data": { + "dataset_name": "CharXiv-D", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/charxiv-d", + "https://api.llm-stats.com/leaderboard/benchmarks/charxiv-d" + ], + "additional_details": { + "raw_benchmark_id": "charxiv-d", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CharXiv-D is the descriptive questions subset of the CharXiv benchmark, designed to assess multimodal large language models' ability to extract basic information from scientific charts. It contains descriptive questions covering information extraction, enumeration, pattern recognition, and counting across 2,323 diverse charts from arXiv papers, all curated and verified by human experts.", + "metric_id": "llm_stats.charxiv-d.score", + "metric_name": "CharXiv-D score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "charxiv-d", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CharXiv-D", + "raw_categories": "[\"multimodal\",\"reasoning\",\"structured_output\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.946, + "details": { + "raw_score": "0.946", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "charxiv-d", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/charxiv-d\",\"https://api.llm-stats.com/leaderboard/benchmarks/charxiv-d\"]", + "raw_score_id": "charxiv-d::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "charxiv-r::charxiv-r-seed-2.1-turbo", + "evaluation_name": "llm_stats.charxiv-r", + "source_data": { + "dataset_name": "CharXiv-R", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/charxiv-r", + "https://api.llm-stats.com/leaderboard/benchmarks/charxiv-r" + ], + "additional_details": { + "raw_benchmark_id": "charxiv-r", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CharXiv-R is the reasoning component of the CharXiv benchmark, focusing on complex reasoning questions that require synthesizing information across visual chart elements. It evaluates multimodal large language models on their ability to understand and reason about scientific charts from arXiv papers through various reasoning tasks.", + "metric_id": "llm_stats.charxiv-r.score", + "metric_name": "CharXiv-R score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "charxiv-r", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CharXiv-R", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "43" + } + }, + "score_details": { + "score": 0.836, + "details": { + "raw_score": "0.836", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "charxiv-r", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/charxiv-r\",\"https://api.llm-stats.com/leaderboard/benchmarks/charxiv-r\"]", + "raw_score_id": "charxiv-r::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "claw-eval-mm::claw-eval-mm-seed-2.1-turbo", + "evaluation_name": "llm_stats.claw-eval-mm", + "source_data": { + "dataset_name": "ClawEval-MM", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/claw-eval-mm", + "https://api.llm-stats.com/leaderboard/benchmarks/claw-eval-mm" + ], + "additional_details": { + "raw_benchmark_id": "claw-eval-mm", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ClawEval-MM is the multimodal variant of ClawEval, evaluating agentic problem solving with visual inputs.", + "metric_id": "llm_stats.claw-eval-mm.score", + "metric_name": "ClawEval-MM score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "claw-eval-mm", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ClawEval-MM", + "raw_categories": "[\"multimodal\",\"agents\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.46, + "details": { + "raw_score": "0.46", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "claw-eval-mm", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/claw-eval-mm\",\"https://api.llm-stats.com/leaderboard/benchmarks/claw-eval-mm\"]", + "raw_score_id": "claw-eval-mm::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "contphy::contphy-seed-2.1-turbo", + "evaluation_name": "llm_stats.contphy", + "source_data": { + "dataset_name": "ContPhy", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/contphy", + "https://api.llm-stats.com/leaderboard/benchmarks/contphy" + ], + "additional_details": { + "raw_benchmark_id": "contphy", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ContPhy is a continuum physical-reasoning benchmark evaluating understanding of physical dynamics in video.", + "metric_id": "llm_stats.contphy.score", + "metric_name": "ContPhy score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "contphy", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ContPhy", + "raw_categories": "[\"multimodal\",\"reasoning\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.611, + "details": { + "raw_score": "0.611", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "contphy", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/contphy\",\"https://api.llm-stats.com/leaderboard/benchmarks/contphy\"]", + "raw_score_id": "contphy::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "creativework::creativework-seed-2.1-turbo", + "evaluation_name": "llm_stats.creativework", + "source_data": { + "dataset_name": "CreativeWork", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/creativework", + "https://api.llm-stats.com/leaderboard/benchmarks/creativework" + ], + "additional_details": { + "raw_benchmark_id": "creativework", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CreativeWork evaluates agents on open-ended creative production tasks within realistic tool and application environments, measuring the quality and completeness of generated deliverables.", + "metric_id": "llm_stats.creativework.score", + "metric_name": "CreativeWork score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "creativework", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CreativeWork", + "raw_categories": "[\"reasoning\",\"general\",\"agents\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.345, + "details": { + "raw_score": "0.345", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "creativework", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/creativework\",\"https://api.llm-stats.com/leaderboard/benchmarks/creativework\"]", + "raw_score_id": "creativework::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "crossvid::crossvid-seed-2.1-turbo", + "evaluation_name": "llm_stats.crossvid", + "source_data": { + "dataset_name": "CrossVid", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/crossvid", + "https://api.llm-stats.com/leaderboard/benchmarks/crossvid" + ], + "additional_details": { + "raw_benchmark_id": "crossvid", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CrossVid evaluates cross-video reasoning, requiring models to integrate information across multiple videos.", + "metric_id": "llm_stats.crossvid.score", + "metric_name": "CrossVid score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "crossvid", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CrossVid", + "raw_categories": "[\"multimodal\",\"reasoning\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.632, + "details": { + "raw_score": "0.632", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "crossvid", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/crossvid\",\"https://api.llm-stats.com/leaderboard/benchmarks/crossvid\"]", + "raw_score_id": "crossvid::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "cybergym::cybergym-seed-2.1-turbo", + "evaluation_name": "llm_stats.cybergym", + "source_data": { + "dataset_name": "CyberGym", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/cybergym", + "https://api.llm-stats.com/leaderboard/benchmarks/cybergym" + ], + "additional_details": { + "raw_benchmark_id": "cybergym", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CyberGym is a benchmark for evaluating AI agents on cybersecurity tasks, testing their ability to identify vulnerabilities, perform security analysis, and complete security-related challenges in a controlled environment.", + "metric_id": "llm_stats.cybergym.score", + "metric_name": "CyberGym score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "cybergym", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CyberGym", + "raw_categories": "[\"safety\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.67, + "details": { + "raw_score": "0.67", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "cybergym", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/cybergym\",\"https://api.llm-stats.com/leaderboard/benchmarks/cybergym\"]", + "raw_score_id": "cybergym::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "deepswe::deepswe-seed-2.1-turbo", + "evaluation_name": "llm_stats.deepswe", + "source_data": { + "dataset_name": "DeepSWE", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/deepswe", + "https://api.llm-stats.com/leaderboard/benchmarks/deepswe" + ], + "additional_details": { + "raw_benchmark_id": "deepswe", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "DeepSWE is a software engineering agent benchmark evaluated with the mini-swe-agent harness, where each task is solved in an isolated container with no internet access. It measures an agent's ability to autonomously resolve real-world coding issues end to end.", + "metric_id": "llm_stats.deepswe.score", + "metric_name": "DeepSWE score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "deepswe", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DeepSWE", + "raw_categories": "[\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.23, + "details": { + "raw_score": "0.23", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "deepswe", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/deepswe\",\"https://api.llm-stats.com/leaderboard/benchmarks/deepswe\"]", + "raw_score_id": "deepswe::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "doubao-multi-turn-bench::doubao-multi-turn-bench-seed-2.1-turbo", + "evaluation_name": "llm_stats.doubao-multi-turn-bench", + "source_data": { + "dataset_name": "Doubao Multi-Turn Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/doubao-multi-turn-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/doubao-multi-turn-bench" + ], + "additional_details": { + "raw_benchmark_id": "doubao-multi-turn-bench", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Doubao Multi-Turn Bench evaluates models on multi-turn conversational tasks, measuring context retention, instruction following, and coherent reasoning across extended dialogues.", + "metric_id": "llm_stats.doubao-multi-turn-bench.score", + "metric_name": "Doubao Multi-Turn Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "doubao-multi-turn-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Doubao Multi-Turn Bench", + "raw_categories": "[\"instruction_following\",\"general\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.49, + "details": { + "raw_score": "0.49", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "doubao-multi-turn-bench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/doubao-multi-turn-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/doubao-multi-turn-bench\"]", + "raw_score_id": "doubao-multi-turn-bench::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "dude::dude-seed-2.1-turbo", + "evaluation_name": "llm_stats.dude", + "source_data": { + "dataset_name": "DUDE", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/dude", + "https://api.llm-stats.com/leaderboard/benchmarks/dude" + ], + "additional_details": { + "raw_benchmark_id": "dude", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "DUDE (Document Understanding Dataset and Evaluation) tests multi-page, multi-domain document understanding and reasoning.", + "metric_id": "llm_stats.dude.score", + "metric_name": "DUDE score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "dude", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DUDE", + "raw_categories": "[\"long_context\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.831, + "details": { + "raw_score": "0.831", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "dude", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/dude\",\"https://api.llm-stats.com/leaderboard/benchmarks/dude\"]", + "raw_score_id": "dude::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "dynamath::dynamath-seed-2.1-turbo", + "evaluation_name": "llm_stats.dynamath", + "source_data": { + "dataset_name": "DynaMath", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/dynamath", + "https://api.llm-stats.com/leaderboard/benchmarks/dynamath" + ], + "additional_details": { + "raw_benchmark_id": "dynamath", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multimodal mathematics and reasoning benchmark focused on dynamic visual problem solving.", + "metric_id": "llm_stats.dynamath.score", + "metric_name": "DynaMath score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "dynamath", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DynaMath", + "raw_categories": "[\"math\",\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.681, + "details": { + "raw_score": "0.681", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "dynamath", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/dynamath\",\"https://api.llm-stats.com/leaderboard/benchmarks/dynamath\"]", + "raw_score_id": "dynamath::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "embspatialbench::embspatialbench-seed-2.1-turbo", + "evaluation_name": "llm_stats.embspatialbench", + "source_data": { + "dataset_name": "EmbSpatialBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/embspatialbench", + "https://api.llm-stats.com/leaderboard/benchmarks/embspatialbench" + ], + "additional_details": { + "raw_benchmark_id": "embspatialbench", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "EmbSpatialBench evaluates embodied spatial understanding and reasoning capabilities.", + "metric_id": "llm_stats.embspatialbench.score", + "metric_name": "EmbSpatialBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "embspatialbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "EmbSpatialBench", + "raw_categories": "[\"spatial_reasoning\",\"embodied\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.825, + "details": { + "raw_score": "0.825", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "embspatialbench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/embspatialbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/embspatialbench\"]", + "raw_score_id": "embspatialbench::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "emma::emma-seed-2.1-turbo", + "evaluation_name": "llm_stats.emma", + "source_data": { + "dataset_name": "EMMA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/emma", + "https://api.llm-stats.com/leaderboard/benchmarks/emma" + ], + "additional_details": { + "raw_benchmark_id": "emma", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "EMMA (Enhanced MultiModal reAsoning) is a benchmark for organic multimodal reasoning across mathematics, physics, chemistry, and coding.", + "metric_id": "llm_stats.emma.score", + "metric_name": "EMMA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "emma", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "EMMA", + "raw_categories": "[\"math\",\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.784, + "details": { + "raw_score": "0.784", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "emma", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/emma\",\"https://api.llm-stats.com/leaderboard/benchmarks/emma\"]", + "raw_score_id": "emma::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "erqa::erqa-seed-2.1-turbo", + "evaluation_name": "llm_stats.erqa", + "source_data": { + "dataset_name": "ERQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/erqa", + "https://api.llm-stats.com/leaderboard/benchmarks/erqa" + ], + "additional_details": { + "raw_benchmark_id": "erqa", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Embodied Reasoning Question Answering benchmark consisting of 400 multiple-choice visual questions across spatial reasoning, trajectory reasoning, action reasoning, state estimation, and multi-view reasoning for evaluating AI capabilities in physical world interactions", + "metric_id": "llm_stats.erqa.score", + "metric_name": "ERQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "erqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ERQA", + "raw_categories": "[\"reasoning\",\"spatial_reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "22" + } + }, + "score_details": { + "score": 0.713, + "details": { + "raw_score": "0.713", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "erqa", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/erqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/erqa\"]", + "raw_score_id": "erqa::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "finance-agent-v1.1::finance-agent-v1.1-seed-2.1-turbo", + "evaluation_name": "llm_stats.finance-agent-v1.1", + "source_data": { + "dataset_name": "Finance Agent v1.1", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/finance-agent-v1.1", + "https://api.llm-stats.com/leaderboard/benchmarks/finance-agent-v1.1" + ], + "additional_details": { + "raw_benchmark_id": "finance-agent-v1.1", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Finance Agent v1.1 is an agentic financial-analysis benchmark that evaluates models on real-world finance workflows, including retrieving and reasoning over financial documents and performing multi-step calculations.", + "metric_id": "llm_stats.finance-agent-v1.1.score", + "metric_name": "Finance Agent v1.1 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "finance-agent-v1.1", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Finance Agent v1.1", + "raw_categories": "[\"reasoning\",\"finance\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.56, + "details": { + "raw_score": "0.56", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "finance-agent-v1.1", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/finance-agent-v1.1\",\"https://api.llm-stats.com/leaderboard/benchmarks/finance-agent-v1.1\"]", + "raw_score_id": "finance-agent-v1.1::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "frontiercs::frontiercs-seed-2.1-turbo", + "evaluation_name": "llm_stats.frontiercs", + "source_data": { + "dataset_name": "FrontierCS", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/frontiercs", + "https://api.llm-stats.com/leaderboard/benchmarks/frontiercs" + ], + "additional_details": { + "raw_benchmark_id": "frontiercs", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "FrontierCS is a benchmark of frontier computer-science problems requiring deep theoretical understanding and rigorous multi-step reasoning at the edge of the field.", + "metric_id": "llm_stats.frontiercs.score", + "metric_name": "FrontierCS score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "frontiercs", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FrontierCS", + "raw_categories": "[\"reasoning\",\"science\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.508, + "details": { + "raw_score": "0.508", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "frontiercs", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/frontiercs\",\"https://api.llm-stats.com/leaderboard/benchmarks/frontiercs\"]", + "raw_score_id": "frontiercs::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "frontierscience-olympiad::frontierscience-olympiad-seed-2.1-turbo", + "evaluation_name": "llm_stats.frontierscience-olympiad", + "source_data": { + "dataset_name": "FrontierScience Olympiad", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/frontierscience-olympiad", + "https://api.llm-stats.com/leaderboard/benchmarks/frontierscience-olympiad" + ], + "additional_details": { + "raw_benchmark_id": "frontierscience-olympiad", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "FrontierScience Olympiad is a benchmark of olympiad-level scientific reasoning problems, testing expert understanding and multi-step reasoning across advanced natural-science domains.", + "metric_id": "llm_stats.frontierscience-olympiad.score", + "metric_name": "FrontierScience Olympiad score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "frontierscience-olympiad", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FrontierScience Olympiad", + "raw_categories": "[\"math\",\"reasoning\",\"science\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.76, + "details": { + "raw_score": "0.76", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "frontierscience-olympiad", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/frontierscience-olympiad\",\"https://api.llm-stats.com/leaderboard/benchmarks/frontierscience-olympiad\"]", + "raw_score_id": "frontierscience-olympiad::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "frontierscience-research::frontierscience-research-seed-2.1-turbo", + "evaluation_name": "llm_stats.frontierscience-research", + "source_data": { + "dataset_name": "FrontierScience Research", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/frontierscience-research", + "https://api.llm-stats.com/leaderboard/benchmarks/frontierscience-research" + ], + "additional_details": { + "raw_benchmark_id": "frontierscience-research", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "FrontierScience Research is a benchmark evaluating AI models on cutting-edge scientific research questions requiring deep domain expertise, multi-step reasoning, and synthesis of complex scientific concepts across disciplines.", + "metric_id": "llm_stats.frontierscience-research.score", + "metric_name": "FrontierScience Research score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "frontierscience-research", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FrontierScience Research", + "raw_categories": "[\"reasoning\",\"science\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.333, + "details": { + "raw_score": "0.333", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "frontierscience-research", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/frontierscience-research\",\"https://api.llm-stats.com/leaderboard/benchmarks/frontierscience-research\"]", + "raw_score_id": "frontierscience-research::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gameworld::gameworld-seed-2.1-turbo", + "evaluation_name": "llm_stats.gameworld", + "source_data": { + "dataset_name": "GameWorld", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/gameworld", + "https://api.llm-stats.com/leaderboard/benchmarks/gameworld" + ], + "additional_details": { + "raw_benchmark_id": "gameworld", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "GameWorld evaluates agents on interactive game environments, testing perception, planning, and sequential decision-making to accomplish in-game objectives.", + "metric_id": "llm_stats.gameworld.score", + "metric_name": "GameWorld score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gameworld", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GameWorld", + "raw_categories": "[\"multimodal\",\"reasoning\",\"agents\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.259, + "details": { + "raw_score": "0.259", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "gameworld", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/gameworld\",\"https://api.llm-stats.com/leaderboard/benchmarks/gameworld\"]", + "raw_score_id": "gameworld::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gdpval::gdpval-seed-2.1-turbo", + "evaluation_name": "llm_stats.gdpval", + "source_data": { + "dataset_name": "GDPval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/gdpval", + "https://api.llm-stats.com/leaderboard/benchmarks/gdpval" + ], + "additional_details": { + "raw_benchmark_id": "gdpval", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "GDPval is an OpenAI benchmark evaluating AI models on economically valuable, real-world knowledge-work tasks spanning many professional occupations and industries.", + "metric_id": "llm_stats.gdpval.score", + "metric_name": "GDPval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gdpval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GDPval", + "raw_categories": "[\"legal\",\"reasoning\",\"finance\",\"general\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.827, + "details": { + "raw_score": "0.827", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "gdpval", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/gdpval\",\"https://api.llm-stats.com/leaderboard/benchmarks/gdpval\"]", + "raw_score_id": "gdpval::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "horizonmath::horizonmath-seed-2.1-turbo", + "evaluation_name": "llm_stats.horizonmath", + "source_data": { + "dataset_name": "HorizonMath", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/horizonmath", + "https://api.llm-stats.com/leaderboard/benchmarks/horizonmath" + ], + "additional_details": { + "raw_benchmark_id": "horizonmath", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "HorizonMath is an extremely difficult frontier mathematics benchmark designed to test the limits of mathematical reasoning on research-level and competition-beyond problems.", + "metric_id": "llm_stats.horizonmath.score", + "metric_name": "HorizonMath score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "horizonmath", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HorizonMath", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.02, + "details": { + "raw_score": "0.02", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "horizonmath", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/horizonmath\",\"https://api.llm-stats.com/leaderboard/benchmarks/horizonmath\"]", + "raw_score_id": "horizonmath::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "humanity-s-last-exam::humanity-s-last-exam-seed-2.1-turbo", + "evaluation_name": "llm_stats.humanity-s-last-exam", + "source_data": { + "dataset_name": "Humanity's Last Exam", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/humanity-s-last-exam", + "https://api.llm-stats.com/leaderboard/benchmarks/humanity's-last-exam" + ], + "additional_details": { + "raw_benchmark_id": "humanity's-last-exam", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Humanity's Last Exam (HLE) is a multi-modal academic benchmark with 2,500 questions across mathematics, humanities, and natural sciences, designed to test LLM capabilities at the frontier of human knowledge with unambiguous, verifiable solutions", + "metric_id": "llm_stats.humanity-s-last-exam.score", + "metric_name": "Humanity's Last Exam score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "humanity's-last-exam", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Humanity's Last Exam", + "raw_categories": "[\"math\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "88" + } + }, + "score_details": { + "score": 0.546, + "details": { + "raw_score": "0.546", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "humanity's-last-exam", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/humanity-s-last-exam\",\"https://api.llm-stats.com/leaderboard/benchmarks/humanity's-last-exam\"]", + "raw_score_id": "humanity's-last-exam::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "image2floorplan::image2floorplan-seed-2.1-turbo", + "evaluation_name": "llm_stats.image2floorplan", + "source_data": { + "dataset_name": "Image2FloorPlan", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/image2floorplan", + "https://api.llm-stats.com/leaderboard/benchmarks/image2floorplan" + ], + "additional_details": { + "raw_benchmark_id": "image2floorplan", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Image2FloorPlan is an in-house benchmark evaluating multimodal models on generating structured floor plans and interactive frontends directly from images such as design mockups and room photos.", + "metric_id": "llm_stats.image2floorplan.score", + "metric_name": "Image2FloorPlan score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "image2floorplan", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Image2FloorPlan", + "raw_categories": "[\"multimodal\",\"coding\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.359, + "details": { + "raw_score": "0.359", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "image2floorplan", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/image2floorplan\",\"https://api.llm-stats.com/leaderboard/benchmarks/image2floorplan\"]", + "raw_score_id": "image2floorplan::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "kina::kina-seed-2.1-turbo", + "evaluation_name": "llm_stats.kina", + "source_data": { + "dataset_name": "KINA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/kina", + "https://api.llm-stats.com/leaderboard/benchmarks/kina" + ], + "additional_details": { + "raw_benchmark_id": "kina", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "KINA is a knowledge-intensive evaluation that measures a model's breadth and depth of factual knowledge across academic and professional domains.", + "metric_id": "llm_stats.kina.score", + "metric_name": "KINA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "kina", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "KINA", + "raw_categories": "[\"knowledge\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.466, + "details": { + "raw_score": "0.466", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "kina", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/kina\",\"https://api.llm-stats.com/leaderboard/benchmarks/kina\"]", + "raw_score_id": "kina::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "livemathematicianbench::livemathematicianbench-seed-2.1-turbo", + "evaluation_name": "llm_stats.livemathematicianbench", + "source_data": { + "dataset_name": "LiveMathematicianBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/livemathematicianbench", + "https://api.llm-stats.com/leaderboard/benchmarks/livemathematicianbench" + ], + "additional_details": { + "raw_benchmark_id": "livemathematicianbench", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LiveMathematicianBench evaluates research-level mathematical reasoning on continuously refreshed, contamination-resistant problems.", + "metric_id": "llm_stats.livemathematicianbench.score", + "metric_name": "LiveMathematicianBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "livemathematicianbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LiveMathematicianBench", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.277, + "details": { + "raw_score": "0.277", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "livemathematicianbench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/livemathematicianbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/livemathematicianbench\"]", + "raw_score_id": "livemathematicianbench::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "livesports-3k::livesports-3k-seed-2.1-turbo", + "evaluation_name": "llm_stats.livesports-3k", + "source_data": { + "dataset_name": "LiveSports-3K", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/livesports-3k", + "https://api.llm-stats.com/leaderboard/benchmarks/livesports-3k" + ], + "additional_details": { + "raw_benchmark_id": "livesports-3k", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LiveSports-3K evaluates fine-grained understanding and commentary of live sports video.", + "metric_id": "llm_stats.livesports-3k.score", + "metric_name": "LiveSports-3K score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "livesports-3k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LiveSports-3K", + "raw_categories": "[\"multimodal\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.771, + "details": { + "raw_score": "0.771", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "livesports-3k", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/livesports-3k\",\"https://api.llm-stats.com/leaderboard/benchmarks/livesports-3k\"]", + "raw_score_id": "livesports-3k::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "longvideobench::longvideobench-seed-2.1-turbo", + "evaluation_name": "llm_stats.longvideobench", + "source_data": { + "dataset_name": "LongVideoBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/longvideobench", + "https://api.llm-stats.com/leaderboard/benchmarks/longvideobench" + ], + "additional_details": { + "raw_benchmark_id": "longvideobench", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LongVideoBench is a question-answering benchmark featuring video-language interleaved inputs up to an hour long. It includes 3,763 varying-length web-collected videos with subtitles across diverse themes and 6,678 human-annotated multiple-choice questions in 17 fine-grained categories for comprehensive evaluation of long-term multimodal understanding.", + "metric_id": "llm_stats.longvideobench.score", + "metric_name": "LongVideoBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "longvideobench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LongVideoBench", + "raw_categories": "[\"long_context\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.806, + "details": { + "raw_score": "0.806", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "longvideobench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/longvideobench\",\"https://api.llm-stats.com/leaderboard/benchmarks/longvideobench\"]", + "raw_score_id": "longvideobench::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "lvbench::lvbench-seed-2.1-turbo", + "evaluation_name": "llm_stats.lvbench", + "source_data": { + "dataset_name": "LVBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/lvbench", + "https://api.llm-stats.com/leaderboard/benchmarks/lvbench" + ], + "additional_details": { + "raw_benchmark_id": "lvbench", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LVBench is an extreme long video understanding benchmark designed to evaluate multimodal models on videos up to two hours in duration. It contains 6 major categories and 21 subcategories, with videos averaging five times longer than existing datasets. The benchmark addresses applications requiring comprehension of extremely long videos.", + "metric_id": "llm_stats.lvbench.score", + "metric_name": "LVBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "lvbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LVBench", + "raw_categories": "[\"long_context\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.768, + "details": { + "raw_score": "0.768", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "lvbench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/lvbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/lvbench\"]", + "raw_score_id": "lvbench::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "matharena-apex::matharena-apex-seed-2.1-turbo", + "evaluation_name": "llm_stats.matharena-apex", + "source_data": { + "dataset_name": "MathArena Apex", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/matharena-apex", + "https://api.llm-stats.com/leaderboard/benchmarks/matharena-apex" + ], + "additional_details": { + "raw_benchmark_id": "matharena-apex", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MathArena Apex is a challenging math contest benchmark featuring the most difficult mathematical problems designed to test advanced reasoning and problem-solving abilities of AI models. It focuses on olympiad-level mathematics and complex multi-step mathematical reasoning.", + "metric_id": "llm_stats.matharena-apex.score", + "metric_name": "MathArena Apex score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "matharena-apex", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MathArena Apex", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.354, + "details": { + "raw_score": "0.354", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "matharena-apex", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/matharena-apex\",\"https://api.llm-stats.com/leaderboard/benchmarks/matharena-apex\"]", + "raw_score_id": "matharena-apex::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mathverse::mathverse-seed-2.1-turbo", + "evaluation_name": "llm_stats.mathverse", + "source_data": { + "dataset_name": "MathVerse", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/mathverse", + "https://api.llm-stats.com/leaderboard/benchmarks/mathverse" + ], + "additional_details": { + "raw_benchmark_id": "mathverse", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MathVerse evaluates multimodal mathematical reasoning, testing whether models genuinely interpret visual math diagrams rather than relying on text.", + "metric_id": "llm_stats.mathverse.score", + "metric_name": "MathVerse score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mathverse", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MathVerse", + "raw_categories": "[\"math\",\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.892, + "details": { + "raw_score": "0.892", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "mathverse", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/mathverse\",\"https://api.llm-stats.com/leaderboard/benchmarks/mathverse\"]", + "raw_score_id": "mathverse::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mathvision::mathvision-seed-2.1-turbo", + "evaluation_name": "llm_stats.mathvision", + "source_data": { + "dataset_name": "MathVision", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/mathvision", + "https://api.llm-stats.com/leaderboard/benchmarks/mathvision" + ], + "additional_details": { + "raw_benchmark_id": "mathvision", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MATH-Vision is a dataset designed to measure multimodal mathematical reasoning capabilities. It focuses on evaluating how well models can solve mathematical problems that require both visual understanding and mathematical reasoning, bridging the gap between visual and mathematical domains.", + "metric_id": "llm_stats.mathvision.score", + "metric_name": "MathVision score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mathvision", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MathVision", + "raw_categories": "[\"math\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "31" + } + }, + "score_details": { + "score": 0.927, + "details": { + "raw_score": "0.927", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "mathvision", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/mathvision\",\"https://api.llm-stats.com/leaderboard/benchmarks/mathvision\"]", + "raw_score_id": "mathvision::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mathvista::mathvista-seed-2.1-turbo", + "evaluation_name": "llm_stats.mathvista", + "source_data": { + "dataset_name": "MathVista", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/mathvista", + "https://api.llm-stats.com/leaderboard/benchmarks/mathvista" + ], + "additional_details": { + "raw_benchmark_id": "mathvista", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MathVista evaluates mathematical reasoning of foundation models in visual contexts. It consists of 6,141 examples derived from 28 existing multimodal datasets and 3 newly created datasets (IQTest, FunctionQA, and PaperQA), combining challenges from diverse mathematical and visual tasks to assess models' ability to understand complex figures and perform rigorous reasoning.", + "metric_id": "llm_stats.mathvista.score", + "metric_name": "MathVista score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mathvista", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MathVista", + "raw_categories": "[\"math\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "39" + } + }, + "score_details": { + "score": 0.905, + "details": { + "raw_score": "0.905", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "mathvista", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/mathvista\",\"https://api.llm-stats.com/leaderboard/benchmarks/mathvista\"]", + "raw_score_id": "mathvista::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mcp-atlas::mcp-atlas-seed-2.1-turbo", + "evaluation_name": "llm_stats.mcp-atlas", + "source_data": { + "dataset_name": "MCP Atlas", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/mcp-atlas", + "https://api.llm-stats.com/leaderboard/benchmarks/mcp-atlas" + ], + "additional_details": { + "raw_benchmark_id": "mcp-atlas", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MCP Atlas is a benchmark for evaluating AI models on scaled tool use capabilities, measuring how well models can coordinate and utilize multiple tools across complex multi-step tasks.", + "metric_id": "llm_stats.mcp-atlas.score", + "metric_name": "MCP Atlas score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mcp-atlas", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MCP Atlas", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.803, + "details": { + "raw_score": "0.803", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "mcp-atlas", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/mcp-atlas\",\"https://api.llm-stats.com/leaderboard/benchmarks/mcp-atlas\"]", + "raw_score_id": "mcp-atlas::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "measurebench::measurebench-seed-2.1-turbo", + "evaluation_name": "llm_stats.measurebench", + "source_data": { + "dataset_name": "MeasureBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/measurebench", + "https://api.llm-stats.com/leaderboard/benchmarks/measurebench" + ], + "additional_details": { + "raw_benchmark_id": "measurebench", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MeasureBench evaluates multimodal models on visual measurement and quantitative perception tasks across both real and synthetic imagery, reported as the average over the two settings.", + "metric_id": "llm_stats.measurebench.score", + "metric_name": "MeasureBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "measurebench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MeasureBench", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.589, + "details": { + "raw_score": "0.589", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "measurebench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/measurebench\",\"https://api.llm-stats.com/leaderboard/benchmarks/measurebench\"]", + "raw_score_id": "measurebench::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "minerva::minerva-seed-2.1-turbo", + "evaluation_name": "llm_stats.minerva", + "source_data": { + "dataset_name": "Minerva", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/minerva", + "https://api.llm-stats.com/leaderboard/benchmarks/minerva" + ], + "additional_details": { + "raw_benchmark_id": "minerva", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Minerva is a benchmark for complex video reasoning, evaluating models on multi-step reasoning over long and information-dense video content.", + "metric_id": "llm_stats.minerva.score", + "metric_name": "Minerva score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "minerva", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Minerva", + "raw_categories": "[\"multimodal\",\"reasoning\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.659, + "details": { + "raw_score": "0.659", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "minerva", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/minerva\",\"https://api.llm-stats.com/leaderboard/benchmarks/minerva\"]", + "raw_score_id": "minerva::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlongbench-128k::mmlongbench-128k-seed-2.1-turbo", + "evaluation_name": "llm_stats.mmlongbench-128k", + "source_data": { + "dataset_name": "MMLongBench-128K", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/mmlongbench-128k", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlongbench-128k" + ], + "additional_details": { + "raw_benchmark_id": "mmlongbench-128k", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MMLongBench-128K evaluates multimodal long-context understanding at a 128K token context length, testing how well vision-language models reason over very long mixed text and image inputs.", + "metric_id": "llm_stats.mmlongbench-128k.score", + "metric_name": "MMLongBench-128K score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlongbench-128k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLongBench-128K", + "raw_categories": "[\"long_context\",\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.769, + "details": { + "raw_score": "0.769", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "mmlongbench-128k", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/mmlongbench-128k\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlongbench-128k\"]", + "raw_score_id": "mmlongbench-128k::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmmu-pro::mmmu-pro-seed-2.1-turbo", + "evaluation_name": "llm_stats.mmmu-pro", + "source_data": { + "dataset_name": "MMMU-Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/mmmu-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/mmmu-pro" + ], + "additional_details": { + "raw_benchmark_id": "mmmu-pro", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A more robust multi-discipline multimodal understanding benchmark that enhances MMMU through a three-step process: filtering text-only answerable questions, augmenting candidate options, and introducing vision-only input settings. Achieves significantly lower model performance (16.8-26.9%) compared to original MMMU, providing more rigorous evaluation that closely mimics real-world scenarios.", + "metric_id": "llm_stats.mmmu-pro.score", + "metric_name": "MMMU-Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmmu-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMMU-Pro", + "raw_categories": "[\"multimodal\",\"reasoning\",\"general\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "60" + } + }, + "score_details": { + "score": 0.822, + "details": { + "raw_score": "0.822", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "mmmu-pro", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/mmmu-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmmu-pro\"]", + "raw_score_id": "mmmu-pro::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmsibench::mmsibench-seed-2.1-turbo", + "evaluation_name": "llm_stats.mmsibench", + "source_data": { + "dataset_name": "MMSIBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/mmsibench", + "https://api.llm-stats.com/leaderboard/benchmarks/mmsibench" + ], + "additional_details": { + "raw_benchmark_id": "mmsibench", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MMSIBench is a multimodal spatial-intelligence benchmark evaluating spatial reasoning over images.", + "metric_id": "llm_stats.mmsibench.score", + "metric_name": "MMSIBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmsibench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMSIBench", + "raw_categories": "[\"multimodal\",\"spatial_reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.314, + "details": { + "raw_score": "0.314", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "mmsibench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/mmsibench\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmsibench\"]", + "raw_score_id": "mmsibench::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mobileworld::mobileworld-seed-2.1-turbo", + "evaluation_name": "llm_stats.mobileworld", + "source_data": { + "dataset_name": "MobileWorld", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/mobileworld", + "https://api.llm-stats.com/leaderboard/benchmarks/mobileworld" + ], + "additional_details": { + "raw_benchmark_id": "mobileworld", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MobileWorld is a benchmark for evaluating multimodal agents on real mobile-device tasks, testing GUI grounding, navigation, and multi-step task completion in mobile environments.", + "metric_id": "llm_stats.mobileworld.score", + "metric_name": "MobileWorld score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mobileworld", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MobileWorld", + "raw_categories": "[\"multimodal\",\"agents\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.7, + "details": { + "raw_score": "0.7", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "mobileworld", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/mobileworld\",\"https://api.llm-stats.com/leaderboard/benchmarks/mobileworld\"]", + "raw_score_id": "mobileworld::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "motionbench::motionbench-seed-2.1-turbo", + "evaluation_name": "llm_stats.motionbench", + "source_data": { + "dataset_name": "MotionBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/motionbench", + "https://api.llm-stats.com/leaderboard/benchmarks/motionbench" + ], + "additional_details": { + "raw_benchmark_id": "motionbench", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MotionBench is a benchmark for evaluating multimodal models on motion understanding in videos, testing the ability to comprehend temporal dynamics, movement patterns, and action sequences.", + "metric_id": "llm_stats.motionbench.score", + "metric_name": "MotionBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "motionbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MotionBench", + "raw_categories": "[\"multimodal\",\"reasoning\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.748, + "details": { + "raw_score": "0.748", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "motionbench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/motionbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/motionbench\"]", + "raw_score_id": "motionbench::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "msqa::msqa-seed-2.1-turbo", + "evaluation_name": "llm_stats.msqa", + "source_data": { + "dataset_name": "MSQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/msqa", + "https://api.llm-stats.com/leaderboard/benchmarks/msqa" + ], + "additional_details": { + "raw_benchmark_id": "msqa", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MSQA is a multilingual question-answering benchmark that measures knowledge and reasoning across a diverse set of languages.", + "metric_id": "llm_stats.msqa.score", + "metric_name": "MSQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "msqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MSQA", + "raw_categories": "[\"knowledge\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.42, + "details": { + "raw_score": "0.42", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "msqa", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/msqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/msqa\"]", + "raw_score_id": "msqa::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "nl2repo::nl2repo-seed-2.1-turbo", + "evaluation_name": "llm_stats.nl2repo", + "source_data": { + "dataset_name": "NL2Repo", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/nl2repo", + "https://api.llm-stats.com/leaderboard/benchmarks/nl2repo" + ], + "additional_details": { + "raw_benchmark_id": "nl2repo", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "NL2Repo evaluates long-horizon coding capabilities including repository-level understanding, where models must generate or modify code across entire repositories from natural language specifications.", + "metric_id": "llm_stats.nl2repo.score", + "metric_name": "NL2Repo score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "nl2repo", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "NL2Repo", + "raw_categories": "[\"agents\",\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.437, + "details": { + "raw_score": "0.437", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "nl2repo", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/nl2repo\",\"https://api.llm-stats.com/leaderboard/benchmarks/nl2repo\"]", + "raw_score_id": "nl2repo::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ocrbench-v2::ocrbench-v2-seed-2.1-turbo", + "evaluation_name": "llm_stats.ocrbench-v2", + "source_data": { + "dataset_name": "OCRBench_V2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/ocrbench-v2", + "https://api.llm-stats.com/leaderboard/benchmarks/ocrbench-v2" + ], + "additional_details": { + "raw_benchmark_id": "ocrbench-v2", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OCRBench v2: Enhanced large-scale bilingual benchmark for evaluating Large Multimodal Models on visual text localization and reasoning with 10,000 human-verified question-answering pairs across 8 core OCR capabilities", + "metric_id": "llm_stats.ocrbench-v2.score", + "metric_name": "OCRBench_V2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ocrbench-v2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OCRBench_V2", + "raw_categories": "[\"image_to_text\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.628, + "details": { + "raw_score": "0.628", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "ocrbench-v2", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/ocrbench-v2\",\"https://api.llm-stats.com/leaderboard/benchmarks/ocrbench-v2\"]", + "raw_score_id": "ocrbench-v2::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "officeqa-pro::officeqa-pro-seed-2.1-turbo", + "evaluation_name": "llm_stats.officeqa-pro", + "source_data": { + "dataset_name": "OfficeQA Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/officeqa-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/officeqa-pro" + ], + "additional_details": { + "raw_benchmark_id": "officeqa-pro", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OfficeQA Pro evaluates AI models on professional knowledge-work questions and tasks drawn from real office workflows, including document analysis, spreadsheet reasoning, and information synthesis across business domains.", + "metric_id": "llm_stats.officeqa-pro.score", + "metric_name": "OfficeQA Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "officeqa-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OfficeQA Pro", + "raw_categories": "[\"reasoning\",\"general\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.711, + "details": { + "raw_score": "0.711", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "officeqa-pro", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/officeqa-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/officeqa-pro\"]", + "raw_score_id": "officeqa-pro::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "onemillion-bench::onemillion-bench-seed-2.1-turbo", + "evaluation_name": "llm_stats.onemillion-bench", + "source_data": { + "dataset_name": "OneMillion Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/onemillion-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/onemillion-bench" + ], + "additional_details": { + "raw_benchmark_id": "onemillion-bench", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OneMillion Bench evaluates AI agents on high-economic-value tasks that require sustained, reliable execution across long-horizon real-world workflows.", + "metric_id": "llm_stats.onemillion-bench.score", + "metric_name": "OneMillion Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "onemillion-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OneMillion Bench", + "raw_categories": "[\"long_context\",\"reasoning\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.666, + "details": { + "raw_score": "0.666", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "onemillion-bench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/onemillion-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/onemillion-bench\"]", + "raw_score_id": "onemillion-bench::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "osworld::osworld-seed-2.1-turbo", + "evaluation_name": "llm_stats.osworld", + "source_data": { + "dataset_name": "OSWorld", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/osworld", + "https://api.llm-stats.com/leaderboard/benchmarks/osworld" + ], + "additional_details": { + "raw_benchmark_id": "osworld", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OSWorld: The first-of-its-kind scalable, real computer environment for multimodal agents, supporting task setup, execution-based evaluation, and interactive learning across Ubuntu, Windows, and macOS with 369 computer tasks involving real web and desktop applications, OS file I/O, and multi-application workflows", + "metric_id": "llm_stats.osworld.score", + "metric_name": "OSWorld score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "osworld", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OSWorld", + "raw_categories": "[\"multimodal\",\"general\",\"agents\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "20" + } + }, + "score_details": { + "score": 0.764, + "details": { + "raw_score": "0.764", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "osworld", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/osworld\",\"https://api.llm-stats.com/leaderboard/benchmarks/osworld\"]", + "raw_score_id": "osworld::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ovbench::ovbench-seed-2.1-turbo", + "evaluation_name": "llm_stats.ovbench", + "source_data": { + "dataset_name": "OVBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/ovbench", + "https://api.llm-stats.com/leaderboard/benchmarks/ovbench" + ], + "additional_details": { + "raw_benchmark_id": "ovbench", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OVBench is an online video understanding benchmark that evaluates a model's ability to perceive, memorize, and reason about real-time video streams as they unfold.", + "metric_id": "llm_stats.ovbench.score", + "metric_name": "OVBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ovbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OVBench", + "raw_categories": "[\"multimodal\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.697, + "details": { + "raw_score": "0.697", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "ovbench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/ovbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/ovbench\"]", + "raw_score_id": "ovbench::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ovobench::ovobench-seed-2.1-turbo", + "evaluation_name": "llm_stats.ovobench", + "source_data": { + "dataset_name": "OVOBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/ovobench", + "https://api.llm-stats.com/leaderboard/benchmarks/ovobench" + ], + "additional_details": { + "raw_benchmark_id": "ovobench", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OVOBench (Online Video Online Benchmark) evaluates streaming video understanding, testing a model's ability to perceive and respond to video content in real time as it unfolds.", + "metric_id": "llm_stats.ovobench.score", + "metric_name": "OVOBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ovobench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OVOBench", + "raw_categories": "[\"multimodal\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.792, + "details": { + "raw_score": "0.792", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "ovobench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/ovobench\",\"https://api.llm-stats.com/leaderboard/benchmarks/ovobench\"]", + "raw_score_id": "ovobench::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "posttrainbench::posttrainbench-seed-2.1-turbo", + "evaluation_name": "llm_stats.posttrainbench", + "source_data": { + "dataset_name": "PostTrainBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/posttrainbench", + "https://api.llm-stats.com/leaderboard/benchmarks/posttrainbench" + ], + "additional_details": { + "raw_benchmark_id": "posttrainbench", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "PostTrainBench evaluates a model's ability to autonomously post-train base models. Given pretrain-only base models, the agent must complete the full pipeline of data synthesis, training, evaluation, and iteration within a time budget, scored across downstream benchmarks such as AIME2025, BFCL, GPQA Main, GSM8K, and HumanEval.", + "metric_id": "llm_stats.posttrainbench.score", + "metric_name": "PostTrainBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "posttrainbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PostTrainBench", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"systems\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.183, + "details": { + "raw_score": "0.183", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "posttrainbench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/posttrainbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/posttrainbench\"]", + "raw_score_id": "posttrainbench::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "presentbench::presentbench-seed-2.1-turbo", + "evaluation_name": "llm_stats.presentbench", + "source_data": { + "dataset_name": "PresentBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/presentbench", + "https://api.llm-stats.com/leaderboard/benchmarks/presentbench" + ], + "additional_details": { + "raw_benchmark_id": "presentbench", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "PresentBench evaluates AI agents on producing presentation-style deliverables, such as generating lesson-plan slides and structured documents from source materials.", + "metric_id": "llm_stats.presentbench.score", + "metric_name": "PresentBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "presentbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PresentBench", + "raw_categories": "[\"reasoning\",\"general\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.483, + "details": { + "raw_score": "0.483", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "presentbench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/presentbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/presentbench\"]", + "raw_score_id": "presentbench::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "program-bench::program-bench-seed-2.1-turbo", + "evaluation_name": "llm_stats.program-bench", + "source_data": { + "dataset_name": "Program Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/program-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/program-bench" + ], + "additional_details": { + "raw_benchmark_id": "program-bench", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Program Bench evaluates code-generation agents by asking them to recreate a program's behavior from only a compiled binary and documentation. It spans 200 tasks from small CLI tools to large systems such as FFmpeg and SQLite, with submissions judged against more than 248,000 fuzz-generated behavioral tests.", + "metric_id": "llm_stats.program-bench.score", + "metric_name": "Program Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "program-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Program Bench", + "raw_categories": "[\"agents\",\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.494, + "details": { + "raw_score": "0.494", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "program-bench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/program-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/program-bench\"]", + "raw_score_id": "program-bench::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "realworldqa::realworldqa-seed-2.1-turbo", + "evaluation_name": "llm_stats.realworldqa", + "source_data": { + "dataset_name": "RealWorldQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/realworldqa", + "https://api.llm-stats.com/leaderboard/benchmarks/realworldqa" + ], + "additional_details": { + "raw_benchmark_id": "realworldqa", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "RealWorldQA is a benchmark designed to evaluate basic real-world spatial understanding capabilities of multimodal models. The initial release consists of over 700 anonymized images taken from vehicles and other real-world scenarios, each accompanied by a question and easily verifiable answer. Released by xAI as part of their Grok-1.5 Vision preview to test models' ability to understand natural scenes and spatial relationships in everyday visual contexts.", + "metric_id": "llm_stats.realworldqa.score", + "metric_name": "RealWorldQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "realworldqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "RealWorldQA", + "raw_categories": "[\"spatial_reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "25" + } + }, + "score_details": { + "score": 0.863, + "details": { + "raw_score": "0.863", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "realworldqa", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/realworldqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/realworldqa\"]", + "raw_score_id": "realworldqa::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "repo-env::repo-env-seed-2.1-turbo", + "evaluation_name": "llm_stats.repo-env", + "source_data": { + "dataset_name": "Repo Env", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/repo-env", + "https://api.llm-stats.com/leaderboard/benchmarks/repo-env" + ], + "additional_details": { + "raw_benchmark_id": "repo-env", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Repo Env evaluates an agent's ability to set up, configure, and run real repositories, including dependency resolution and environment management.", + "metric_id": "llm_stats.repo-env.score", + "metric_name": "Repo Env score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "repo-env", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Repo Env", + "raw_categories": "[\"agents\",\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.467, + "details": { + "raw_score": "0.467", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "repo-env", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/repo-env\",\"https://api.llm-stats.com/leaderboard/benchmarks/repo-env\"]", + "raw_score_id": "repo-env::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "scicode::scicode-seed-2.1-turbo", + "evaluation_name": "llm_stats.scicode", + "source_data": { + "dataset_name": "SciCode", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/scicode", + "https://api.llm-stats.com/leaderboard/benchmarks/scicode" + ], + "additional_details": { + "raw_benchmark_id": "scicode", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SciCode is a research coding benchmark curated by scientists that challenges language models to code solutions for scientific problems. It contains 338 subproblems decomposed from 80 challenging main problems across 16 natural science sub-fields including mathematics, physics, chemistry, biology, and materials science. Problems require knowledge recall, reasoning, and code synthesis skills.", + "metric_id": "llm_stats.scicode.score", + "metric_name": "SciCode score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "scicode", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SciCode", + "raw_categories": "[\"math\",\"physics\",\"reasoning\",\"biology\",\"chemistry\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.578, + "details": { + "raw_score": "0.578", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "scicode", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/scicode\",\"https://api.llm-stats.com/leaderboard/benchmarks/scicode\"]", + "raw_score_id": "scicode::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "seedclawbench::seedclawbench-seed-2.1-turbo", + "evaluation_name": "llm_stats.seedclawbench", + "source_data": { + "dataset_name": "SeedClawBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/seedclawbench", + "https://api.llm-stats.com/leaderboard/benchmarks/seedclawbench" + ], + "additional_details": { + "raw_benchmark_id": "seedclawbench", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SeedClawBench is an agentic coding benchmark measuring overall model performance on real-world, tool-using software development tasks.", + "metric_id": "llm_stats.seedclawbench.score", + "metric_name": "SeedClawBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "seedclawbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SeedClawBench", + "raw_categories": "[\"agents\",\"coding\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.638, + "details": { + "raw_score": "0.638", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "seedclawbench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/seedclawbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/seedclawbench\"]", + "raw_score_id": "seedclawbench::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "simplevqa::simplevqa-seed-2.1-turbo", + "evaluation_name": "llm_stats.simplevqa", + "source_data": { + "dataset_name": "SimpleVQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/simplevqa", + "https://api.llm-stats.com/leaderboard/benchmarks/simplevqa" + ], + "additional_details": { + "raw_benchmark_id": "simplevqa", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SimpleVQA is a visual question answering benchmark focused on simple queries.", + "metric_id": "llm_stats.simplevqa.score", + "metric_name": "SimpleVQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "simplevqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SimpleVQA", + "raw_categories": "[\"image_to_text\",\"multimodal\",\"general\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.711, + "details": { + "raw_score": "0.711", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "simplevqa", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/simplevqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/simplevqa\"]", + "raw_score_id": "simplevqa::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "superchem::superchem-seed-2.1-turbo", + "evaluation_name": "llm_stats.superchem", + "source_data": { + "dataset_name": "SuperChem", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/superchem", + "https://api.llm-stats.com/leaderboard/benchmarks/superchem" + ], + "additional_details": { + "raw_benchmark_id": "superchem", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SuperChem is a benchmark of advanced chemistry problems requiring expert-level domain knowledge and reasoning.", + "metric_id": "llm_stats.superchem.score", + "metric_name": "SuperChem score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "superchem", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SuperChem", + "raw_categories": "[\"knowledge\",\"reasoning\",\"chemistry\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.566, + "details": { + "raw_score": "0.566", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "superchem", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/superchem\",\"https://api.llm-stats.com/leaderboard/benchmarks/superchem\"]", + "raw_score_id": "superchem::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "supergpqa::supergpqa-seed-2.1-turbo", + "evaluation_name": "llm_stats.supergpqa", + "source_data": { + "dataset_name": "SuperGPQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/supergpqa", + "https://api.llm-stats.com/leaderboard/benchmarks/supergpqa" + ], + "additional_details": { + "raw_benchmark_id": "supergpqa", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SuperGPQA is a comprehensive benchmark that evaluates large language models across 285 graduate-level academic disciplines. The benchmark contains 25,957 questions covering 13 broad disciplinary areas including Engineering, Medicine, Science, and Law, with specialized fields in light industry, agriculture, and service-oriented domains. It employs a Human-LLM collaborative filtering mechanism with over 80 expert annotators to create challenging questions that assess graduate-level knowledge and reasoning capabilities.", + "metric_id": "llm_stats.supergpqa.score", + "metric_name": "SuperGPQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "supergpqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SuperGPQA", + "raw_categories": "[\"legal\",\"math\",\"physics\",\"reasoning\",\"finance\",\"general\",\"healthcare\",\"chemistry\",\"economics\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "34" + } + }, + "score_details": { + "score": 0.674, + "details": { + "raw_score": "0.674", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "supergpqa", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/supergpqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/supergpqa\"]", + "raw_score_id": "supergpqa::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-atlas::swe-atlas-seed-2.1-turbo", + "evaluation_name": "llm_stats.swe-atlas", + "source_data": { + "dataset_name": "SWE-Atlas", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/swe-atlas", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-atlas" + ], + "additional_details": { + "raw_benchmark_id": "swe-atlas", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SWE-Atlas is a software engineering benchmark focused on debugging, evaluating a model's ability to localize and fix bugs in real-world codebases.", + "metric_id": "llm_stats.swe-atlas.score", + "metric_name": "SWE-Atlas score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-atlas", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Atlas", + "raw_categories": "[\"agents\",\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.306, + "details": { + "raw_score": "0.306", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "swe-atlas", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/swe-atlas\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-atlas\"]", + "raw_score_id": "swe-atlas::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-pro::swe-bench-pro-seed-2.1-turbo", + "evaluation_name": "llm_stats.swe-bench-pro", + "source_data": { + "dataset_name": "SWE-Bench Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/swe-bench-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-pro" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-pro", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SWE-Bench Pro is an advanced version of SWE-Bench that evaluates language models on complex, real-world software engineering tasks requiring extended reasoning and multi-step problem solving.", + "metric_id": "llm_stats.swe-bench-pro.score", + "metric_name": "SWE-Bench Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Bench Pro", + "raw_categories": "[\"reasoning\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "35" + } + }, + "score_details": { + "score": 0.57, + "details": { + "raw_score": "0.57", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "swe-bench-pro", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/swe-bench-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-pro\"]", + "raw_score_id": "swe-bench-pro::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2.1::terminal-bench-2.1-seed-2.1-turbo", + "evaluation_name": "llm_stats.terminal-bench-2.1", + "source_data": { + "dataset_name": "Terminal-Bench 2.1", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/terminal-bench-2.1", + "https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2.1" + ], + "additional_details": { + "raw_benchmark_id": "terminal-bench-2.1", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Terminal-Bench 2.1 is an updated release of the Terminal-Bench benchmark that tests AI agents' ability to operate a computer via the terminal. It evaluates how well models handle real-world, end-to-end tasks autonomously, including compiling code, training models, setting up servers, system administration, data science workflows, and security tasks.", + "metric_id": "llm_stats.terminal-bench-2.1.score", + "metric_name": "Terminal-Bench 2.1 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "terminal-bench-2.1", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Terminal-Bench 2.1", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.676, + "details": { + "raw_score": "0.676", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "terminal-bench-2.1", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/terminal-bench-2.1\",\"https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2.1\"]", + "raw_score_id": "terminal-bench-2.1::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "tomato::tomato-seed-2.1-turbo", + "evaluation_name": "llm_stats.tomato", + "source_data": { + "dataset_name": "TOMATO", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/tomato", + "https://api.llm-stats.com/leaderboard/benchmarks/tomato" + ], + "additional_details": { + "raw_benchmark_id": "tomato", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "TOMATO (Temporal Reasoning Multimodal Evaluation) assesses multimodal models on motion and temporal perception in video, testing understanding of actions, motion, and changes over time.", + "metric_id": "llm_stats.tomato.score", + "metric_name": "TOMATO score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "tomato", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TOMATO", + "raw_categories": "[\"multimodal\",\"reasoning\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.568, + "details": { + "raw_score": "0.568", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "tomato", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/tomato\",\"https://api.llm-stats.com/leaderboard/benchmarks/tomato\"]", + "raw_score_id": "tomato::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "toolathlon::toolathlon-seed-2.1-turbo", + "evaluation_name": "llm_stats.toolathlon", + "source_data": { + "dataset_name": "Toolathlon", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/toolathlon", + "https://api.llm-stats.com/leaderboard/benchmarks/toolathlon" + ], + "additional_details": { + "raw_benchmark_id": "toolathlon", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Tool Decathlon is a comprehensive benchmark for evaluating AI agents' ability to use multiple tools across diverse task categories. It measures proficiency in tool selection, sequencing, and execution across ten different tool-use scenarios.", + "metric_id": "llm_stats.toolathlon.score", + "metric_name": "Toolathlon score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "toolathlon", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Toolathlon", + "raw_categories": "[\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "24" + } + }, + "score_details": { + "score": 0.491, + "details": { + "raw_score": "0.491", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "toolathlon", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/toolathlon\",\"https://api.llm-stats.com/leaderboard/benchmarks/toolathlon\"]", + "raw_score_id": "toolathlon::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "trae-code-gen::trae-code-gen-seed-2.1-turbo", + "evaluation_name": "llm_stats.trae-code-gen", + "source_data": { + "dataset_name": "Trae Code Gen", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/trae-code-gen", + "https://api.llm-stats.com/leaderboard/benchmarks/trae-code-gen" + ], + "additional_details": { + "raw_benchmark_id": "trae-code-gen", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Trae Code Gen is a component of Trae Agent Bench that evaluates implementing new functionality across multiple programming languages in containerized, runnable repositories.", + "metric_id": "llm_stats.trae-code-gen.score", + "metric_name": "Trae Code Gen score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "trae-code-gen", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Trae Code Gen", + "raw_categories": "[\"agents\",\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.597, + "details": { + "raw_score": "0.597", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "trae-code-gen", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/trae-code-gen\",\"https://api.llm-stats.com/leaderboard/benchmarks/trae-code-gen\"]", + "raw_score_id": "trae-code-gen::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "trae-error-fix::trae-error-fix-seed-2.1-turbo", + "evaluation_name": "llm_stats.trae-error-fix", + "source_data": { + "dataset_name": "Trae Error Fix", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/trae-error-fix", + "https://api.llm-stats.com/leaderboard/benchmarks/trae-error-fix" + ], + "additional_details": { + "raw_benchmark_id": "trae-error-fix", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Trae Error Fix is a component of Trae Agent Bench that evaluates fixing existing code across multiple programming languages in containerized, runnable repositories.", + "metric_id": "llm_stats.trae-error-fix.score", + "metric_name": "Trae Error Fix score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "trae-error-fix", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Trae Error Fix", + "raw_categories": "[\"agents\",\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.567, + "details": { + "raw_score": "0.567", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "trae-error-fix", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/trae-error-fix\",\"https://api.llm-stats.com/leaderboard/benchmarks/trae-error-fix\"]", + "raw_score_id": "trae-error-fix::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "treebench::treebench-seed-2.1-turbo", + "evaluation_name": "llm_stats.treebench", + "source_data": { + "dataset_name": "TreeBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/treebench", + "https://api.llm-stats.com/leaderboard/benchmarks/treebench" + ], + "additional_details": { + "raw_benchmark_id": "treebench", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "TreeBench evaluates visual grounded reasoning, requiring models to localize and reason about fine-grained visual details.", + "metric_id": "llm_stats.treebench.score", + "metric_name": "TreeBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "treebench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TreeBench", + "raw_categories": "[\"multimodal\",\"reasoning\",\"spatial_reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.711, + "details": { + "raw_score": "0.711", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "treebench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/treebench\",\"https://api.llm-stats.com/leaderboard/benchmarks/treebench\"]", + "raw_score_id": "treebench::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "tvbench::tvbench-seed-2.1-turbo", + "evaluation_name": "llm_stats.tvbench", + "source_data": { + "dataset_name": "TVBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/tvbench", + "https://api.llm-stats.com/leaderboard/benchmarks/tvbench" + ], + "additional_details": { + "raw_benchmark_id": "tvbench", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "TVBench is a temporal video understanding benchmark evaluating reasoning over actions, events, and temporal dynamics in videos.", + "metric_id": "llm_stats.tvbench.score", + "metric_name": "TVBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "tvbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TVBench", + "raw_categories": "[\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.772, + "details": { + "raw_score": "0.772", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "tvbench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/tvbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/tvbench\"]", + "raw_score_id": "tvbench::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "video-mme::video-mme-seed-2.1-turbo", + "evaluation_name": "llm_stats.video-mme", + "source_data": { + "dataset_name": "Video-MME", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/video-mme", + "https://api.llm-stats.com/leaderboard/benchmarks/video-mme" + ], + "additional_details": { + "raw_benchmark_id": "video-mme", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Video-MME is the first-ever comprehensive evaluation benchmark of Multi-modal Large Language Models (MLLMs) in video analysis. It features 900 videos totaling 254 hours with 2,700 human-annotated question-answer pairs across 6 primary visual domains (Knowledge, Film & Television, Sports Competition, Life Record, Multilingual, and others) and 30 subfields. The benchmark evaluates models across diverse temporal dimensions (11 seconds to 1 hour), integrates multi-modal inputs including video frames, subtitles, and audio, and uses rigorous manual labeling by expert annotators for precise assessment.", + "metric_id": "llm_stats.video-mme.score", + "metric_name": "Video-MME score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "video-mme", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Video-MME", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "17" + } + }, + "score_details": { + "score": 0.89, + "details": { + "raw_score": "0.89", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "video-mme", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/video-mme\",\"https://api.llm-stats.com/leaderboard/benchmarks/video-mme\"]", + "raw_score_id": "video-mme::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "videoholmes::videoholmes-seed-2.1-turbo", + "evaluation_name": "llm_stats.videoholmes", + "source_data": { + "dataset_name": "VideoHolmes", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/videoholmes", + "https://api.llm-stats.com/leaderboard/benchmarks/videoholmes" + ], + "additional_details": { + "raw_benchmark_id": "videoholmes", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VideoHolmes evaluates video understanding and reasoning capabilities in multimodal models.", + "metric_id": "llm_stats.videoholmes.score", + "metric_name": "VideoHolmes score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "videoholmes", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VideoHolmes", + "raw_categories": "[\"multimodal\",\"reasoning\",\"video\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.676, + "details": { + "raw_score": "0.676", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "videoholmes", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/videoholmes\",\"https://api.llm-stats.com/leaderboard/benchmarks/videoholmes\"]", + "raw_score_id": "videoholmes::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "videosimpleqa::videosimpleqa-seed-2.1-turbo", + "evaluation_name": "llm_stats.videosimpleqa", + "source_data": { + "dataset_name": "VideoSimpleQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/videosimpleqa", + "https://api.llm-stats.com/leaderboard/benchmarks/videosimpleqa" + ], + "additional_details": { + "raw_benchmark_id": "videosimpleqa", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VideoSimpleQA evaluates factual knowledge grounded in video content, measuring how accurately models answer short, fact-seeking questions about videos.", + "metric_id": "llm_stats.videosimpleqa.score", + "metric_name": "VideoSimpleQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "videosimpleqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VideoSimpleQA", + "raw_categories": "[\"knowledge\",\"multimodal\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.714, + "details": { + "raw_score": "0.714", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "videosimpleqa", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/videosimpleqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/videosimpleqa\"]", + "raw_score_id": "videosimpleqa::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "visfactor::visfactor-seed-2.1-turbo", + "evaluation_name": "llm_stats.visfactor", + "source_data": { + "dataset_name": "VisFactor", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/visfactor", + "https://api.llm-stats.com/leaderboard/benchmarks/visfactor" + ], + "additional_details": { + "raw_benchmark_id": "visfactor", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VisFactor is a benchmark evaluating fine-grained visual factor perception and reasoning over images.", + "metric_id": "llm_stats.visfactor.score", + "metric_name": "VisFactor score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "visfactor", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VisFactor", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.439, + "details": { + "raw_score": "0.439", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "visfactor", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/visfactor\",\"https://api.llm-stats.com/leaderboard/benchmarks/visfactor\"]", + "raw_score_id": "visfactor::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "visulogic::visulogic-seed-2.1-turbo", + "evaluation_name": "llm_stats.visulogic", + "source_data": { + "dataset_name": "VisuLogic", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/visulogic", + "https://api.llm-stats.com/leaderboard/benchmarks/visulogic" + ], + "additional_details": { + "raw_benchmark_id": "visulogic", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VisuLogic evaluates logical reasoning capabilities in visual contexts.", + "metric_id": "llm_stats.visulogic.score", + "metric_name": "VisuLogic score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "visulogic", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VisuLogic", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.529, + "details": { + "raw_score": "0.529", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "visulogic", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/visulogic\",\"https://api.llm-stats.com/leaderboard/benchmarks/visulogic\"]", + "raw_score_id": "visulogic::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "vlmsarebiased::vlmsarebiased-seed-2.1-turbo", + "evaluation_name": "llm_stats.vlmsarebiased", + "source_data": { + "dataset_name": "VLMsAreBiased", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/vlmsarebiased", + "https://api.llm-stats.com/leaderboard/benchmarks/vlmsarebiased" + ], + "additional_details": { + "raw_benchmark_id": "vlmsarebiased", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VLMsAreBiased evaluates whether vision-language models rely on visual evidence or fall back on language priors when answering.", + "metric_id": "llm_stats.vlmsarebiased.score", + "metric_name": "VLMsAreBiased score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "vlmsarebiased", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VLMsAreBiased", + "raw_categories": "[\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.683, + "details": { + "raw_score": "0.683", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "vlmsarebiased", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/vlmsarebiased\",\"https://api.llm-stats.com/leaderboard/benchmarks/vlmsarebiased\"]", + "raw_score_id": "vlmsarebiased::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "web-bench::web-bench-seed-2.1-turbo", + "evaluation_name": "llm_stats.web-bench", + "source_data": { + "dataset_name": "Web Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/web-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/web-bench" + ], + "additional_details": { + "raw_benchmark_id": "web-bench", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Web Bench evaluates agents on realistic web-development engineering tasks, measuring end-to-end implementation in browser-based workflows.", + "metric_id": "llm_stats.web-bench.score", + "metric_name": "Web Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "web-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Web Bench", + "raw_categories": "[\"agents\",\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.736, + "details": { + "raw_score": "0.736", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "web-bench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/web-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/web-bench\"]", + "raw_score_id": "web-bench::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "wildclawbench::wildclawbench-seed-2.1-turbo", + "evaluation_name": "llm_stats.wildclawbench", + "source_data": { + "dataset_name": "WildClawBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/wildclawbench", + "https://api.llm-stats.com/leaderboard/benchmarks/wildclawbench" + ], + "additional_details": { + "raw_benchmark_id": "wildclawbench", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WildClawBench is an agentic coding benchmark from InternLM/Claw-Eval that reports overall model performance on real-world tool-using development tasks.", + "metric_id": "llm_stats.wildclawbench.score", + "metric_name": "WildClawBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "wildclawbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WildClawBench", + "raw_categories": "[\"agents\",\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.628, + "details": { + "raw_score": "0.628", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "wildclawbench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/wildclawbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/wildclawbench\"]", + "raw_score_id": "wildclawbench::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "workspace-bench::workspace-bench-seed-2.1-turbo", + "evaluation_name": "llm_stats.workspace-bench", + "source_data": { + "dataset_name": "Workspace Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/workspace-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/workspace-bench" + ], + "additional_details": { + "raw_benchmark_id": "workspace-bench", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Workspace Bench evaluates AI agents on high-economic-value workplace tasks that span multi-step planning, file processing, and tool use across realistic office and productivity workflows.", + "metric_id": "llm_stats.workspace-bench.score", + "metric_name": "Workspace Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "workspace-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Workspace Bench", + "raw_categories": "[\"reasoning\",\"general\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.547, + "details": { + "raw_score": "0.547", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "workspace-bench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/workspace-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/workspace-bench\"]", + "raw_score_id": "workspace-bench::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "worldbench::worldbench-seed-2.1-turbo", + "evaluation_name": "llm_stats.worldbench", + "source_data": { + "dataset_name": "WorldBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/worldbench", + "https://api.llm-stats.com/leaderboard/benchmarks/worldbench" + ], + "additional_details": { + "raw_benchmark_id": "worldbench", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WorldBench evaluates real-world visual knowledge and understanding across diverse everyday scenes.", + "metric_id": "llm_stats.worldbench.score", + "metric_name": "WorldBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "worldbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WorldBench", + "raw_categories": "[\"knowledge\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.637, + "details": { + "raw_score": "0.637", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "worldbench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/worldbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/worldbench\"]", + "raw_score_id": "worldbench::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "worldvqa::worldvqa-seed-2.1-turbo", + "evaluation_name": "llm_stats.worldvqa", + "source_data": { + "dataset_name": "WorldVQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/worldvqa", + "https://api.llm-stats.com/leaderboard/benchmarks/worldvqa" + ], + "additional_details": { + "raw_benchmark_id": "worldvqa", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WorldVQA is a benchmark designed to evaluate atomic vision-centric world knowledge. It assesses models' ability to understand and reason about visual elements representing real-world knowledge.", + "metric_id": "llm_stats.worldvqa.score", + "metric_name": "WorldVQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "worldvqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WorldVQA", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.486, + "details": { + "raw_score": "0.486", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "worldvqa", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/worldvqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/worldvqa\"]", + "raw_score_id": "worldvqa::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "xdailybench::xdailybench-seed-2.1-turbo", + "evaluation_name": "llm_stats.xdailybench", + "source_data": { + "dataset_name": "xDailyBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/xdailybench", + "https://api.llm-stats.com/leaderboard/benchmarks/xdailybench" + ], + "additional_details": { + "raw_benchmark_id": "xdailybench", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "xDailyBench evaluates AI agents on white-collar office work, covering everyday professional tasks such as document handling, consultation, and multi-step productivity workflows.", + "metric_id": "llm_stats.xdailybench.score", + "metric_name": "xDailyBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "xdailybench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "xDailyBench", + "raw_categories": "[\"reasoning\",\"general\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.564, + "details": { + "raw_score": "0.564", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "xdailybench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/xdailybench\",\"https://api.llm-stats.com/leaderboard/benchmarks/xdailybench\"]", + "raw_score_id": "xdailybench::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "zerobench::zerobench-seed-2.1-turbo", + "evaluation_name": "llm_stats.zerobench", + "source_data": { + "dataset_name": "ZEROBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/seed-2.1-turbo", + "https://llm-stats.com/benchmarks/zerobench", + "https://api.llm-stats.com/leaderboard/benchmarks/zerobench" + ], + "additional_details": { + "raw_benchmark_id": "zerobench", + "raw_model_id": "seed-2.1-turbo", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ZEROBench is a challenging vision benchmark designed to test models on zero-shot visual understanding tasks.", + "metric_id": "llm_stats.zerobench.score", + "metric_name": "ZEROBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "zerobench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ZEROBench", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.572, + "details": { + "raw_score": "0.572", + "raw_score_field": "score", + "raw_model_id": "seed-2.1-turbo", + "raw_benchmark_id": "zerobench", + "source_urls_json": "[\"https://llm-stats.com/models/seed-2.1-turbo\",\"https://llm-stats.com/benchmarks/zerobench\",\"https://api.llm-stats.com/leaderboard/benchmarks/zerobench\"]", + "raw_score_id": "zerobench::seed-2.1-turbo", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/deepseek/deepseek-v2.5/5d724d90-a697-4e6d-aaf0-3550180ea5e5.json b/data/llm_stats/deepseek/deepseek-v2.5/5d724d90-a697-4e6d-aaf0-3550180ea5e5.json new file mode 100644 index 0000000000000000000000000000000000000000..e859e007ce1185d13b92b646ebbd055112b3ca0c --- /dev/null +++ b/data/llm_stats/deepseek/deepseek-v2.5/5d724d90-a697-4e6d-aaf0-3550180ea5e5.json @@ -0,0 +1,671 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/deepseek_deepseek-v2.5/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "DeepSeek-V2.5", + "id": "deepseek/deepseek-v2.5", + "developer": "deepseek", + "additional_details": { + "raw_id": "deepseek-v2.5", + "raw_name": "DeepSeek-V2.5", + "raw_model_id": "deepseek-v2.5", + "raw_model_name": "DeepSeek-V2.5", + "raw_organization_id": "deepseek", + "raw_organization_name": "DeepSeek", + "raw_release_date": "2024-05-08", + "raw_announcement_date": "2024-05-08", + "raw_multimodal": "false", + "raw_param_count": "236000000000", + "raw_provider_slug": "deepseek", + "raw_provider_name": "DeepSeek" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aider::aider-deepseek-v2.5", + "evaluation_name": "llm_stats.aider", + "source_data": { + "dataset_name": "Aider", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/deepseek-v2.5", + "https://llm-stats.com/benchmarks/aider", + "https://api.llm-stats.com/leaderboard/benchmarks/aider" + ], + "additional_details": { + "raw_benchmark_id": "aider", + "raw_model_id": "deepseek-v2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Aider is a comprehensive code editing benchmark based on 133 practice exercises from Exercism's Python repository, designed to evaluate AI models' ability to translate natural language coding requests into executable code that passes unit tests. The benchmark measures end-to-end code editing capabilities, including GPT's ability to edit existing code and format code changes for automated saving to local files. The Aider Polyglot variant extends this evaluation across 225 challenging exercises spanning C++, Go, Java, JavaScript, Python, and Rust, making it a standard benchmark for assessing multilingual code editing performance in AI research.", + "metric_id": "llm_stats.aider.score", + "metric_name": "Aider score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aider", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Aider", + "raw_categories": "[\"reasoning\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.722, + "details": { + "raw_score": "0.722", + "raw_score_field": "score", + "raw_model_id": "deepseek-v2.5", + "raw_benchmark_id": "aider", + "source_urls_json": "[\"https://llm-stats.com/models/deepseek-v2.5\",\"https://llm-stats.com/benchmarks/aider\",\"https://api.llm-stats.com/leaderboard/benchmarks/aider\"]", + "raw_score_id": "aider::deepseek-v2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "alignbench::alignbench-deepseek-v2.5", + "evaluation_name": "llm_stats.alignbench", + "source_data": { + "dataset_name": "AlignBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/deepseek-v2.5", + "https://llm-stats.com/benchmarks/alignbench", + "https://api.llm-stats.com/leaderboard/benchmarks/alignbench" + ], + "additional_details": { + "raw_benchmark_id": "alignbench", + "raw_model_id": "deepseek-v2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "AlignBench is a comprehensive multi-dimensional benchmark for evaluating Chinese alignment of Large Language Models. It contains 8 main categories: Fundamental Language Ability, Advanced Chinese Understanding, Open-ended Questions, Writing Ability, Logical Reasoning, Mathematics, Task-oriented Role Play, and Professional Knowledge. The benchmark includes 683 real-scenario rooted queries with human-verified references and uses a rule-calibrated multi-dimensional LLM-as-Judge approach with Chain-of-Thought for evaluation.", + "metric_id": "llm_stats.alignbench.score", + "metric_name": "AlignBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "alignbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AlignBench", + "raw_categories": "[\"language\",\"math\",\"reasoning\",\"roleplay\",\"general\",\"creativity\",\"writing\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.804, + "details": { + "raw_score": "0.804", + "raw_score_field": "score", + "raw_model_id": "deepseek-v2.5", + "raw_benchmark_id": "alignbench", + "source_urls_json": "[\"https://llm-stats.com/models/deepseek-v2.5\",\"https://llm-stats.com/benchmarks/alignbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/alignbench\"]", + "raw_score_id": "alignbench::deepseek-v2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "alpacaeval-2.0::alpacaeval-2.0-deepseek-v2.5", + "evaluation_name": "llm_stats.alpacaeval-2.0", + "source_data": { + "dataset_name": "AlpacaEval 2.0", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/deepseek-v2.5", + "https://llm-stats.com/benchmarks/alpacaeval-2.0", + "https://api.llm-stats.com/leaderboard/benchmarks/alpacaeval-2.0" + ], + "additional_details": { + "raw_benchmark_id": "alpacaeval-2.0", + "raw_model_id": "deepseek-v2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "AlpacaEval 2.0 is a length-controlled automatic evaluator for instruction-following language models that uses GPT-4 Turbo to assess model responses against a baseline. It evaluates models on 805 diverse instruction-following tasks including creative writing, classification, programming, and general knowledge questions. The benchmark achieves 0.98 Spearman correlation with ChatBot Arena while being fast (< 3 minutes) and affordable (< $10 in OpenAI credits). It addresses length bias in automatic evaluation through length-controlled win-rates and uses weighted scoring based on response quality.", + "metric_id": "llm_stats.alpacaeval-2.0.score", + "metric_name": "AlpacaEval 2.0 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "alpacaeval-2.0", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AlpacaEval 2.0", + "raw_categories": "[\"reasoning\",\"general\",\"creativity\",\"writing\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.505, + "details": { + "raw_score": "0.505", + "raw_score_field": "score", + "raw_model_id": "deepseek-v2.5", + "raw_benchmark_id": "alpacaeval-2.0", + "source_urls_json": "[\"https://llm-stats.com/models/deepseek-v2.5\",\"https://llm-stats.com/benchmarks/alpacaeval-2.0\",\"https://api.llm-stats.com/leaderboard/benchmarks/alpacaeval-2.0\"]", + "raw_score_id": "alpacaeval-2.0::deepseek-v2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "arena-hard::arena-hard-deepseek-v2.5", + "evaluation_name": "llm_stats.arena-hard", + "source_data": { + "dataset_name": "Arena Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/deepseek-v2.5", + "https://llm-stats.com/benchmarks/arena-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/arena-hard" + ], + "additional_details": { + "raw_benchmark_id": "arena-hard", + "raw_model_id": "deepseek-v2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Arena-Hard-Auto is an automatic evaluation benchmark for instruction-tuned LLMs consisting of 500 challenging real-world prompts curated by BenchBuilder. It includes open-ended software engineering problems, mathematical questions, and creative writing tasks. The benchmark uses LLM-as-a-Judge methodology with GPT-4.1 and Gemini-2.5 as automatic judges to approximate human preference. Arena-Hard achieves 98.6% correlation with human preference rankings and provides 3x higher separation of model performances compared to MT-Bench, making it highly effective for distinguishing between models of similar quality.", + "metric_id": "llm_stats.arena-hard.score", + "metric_name": "Arena Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arena-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Arena Hard", + "raw_categories": "[\"reasoning\",\"general\",\"creativity\",\"writing\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "26" + } + }, + "score_details": { + "score": 0.762, + "details": { + "raw_score": "0.762", + "raw_score_field": "score", + "raw_model_id": "deepseek-v2.5", + "raw_benchmark_id": "arena-hard", + "source_urls_json": "[\"https://llm-stats.com/models/deepseek-v2.5\",\"https://llm-stats.com/benchmarks/arena-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/arena-hard\"]", + "raw_score_id": "arena-hard::deepseek-v2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bbh::bbh-deepseek-v2.5", + "evaluation_name": "llm_stats.bbh", + "source_data": { + "dataset_name": "BBH", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/deepseek-v2.5", + "https://llm-stats.com/benchmarks/bbh", + "https://api.llm-stats.com/leaderboard/benchmarks/bbh" + ], + "additional_details": { + "raw_benchmark_id": "bbh", + "raw_model_id": "deepseek-v2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Big-Bench Hard (BBH) is a suite of 23 challenging tasks selected from BIG-Bench for which prior language model evaluations did not outperform the average human-rater. These tasks require multi-step reasoning across diverse domains including arithmetic, logical reasoning, reading comprehension, and commonsense reasoning. The benchmark was designed to test capabilities believed to be beyond current language models and focuses on evaluating complex reasoning skills including temporal understanding, spatial reasoning, causal understanding, and deductive logical reasoning.", + "metric_id": "llm_stats.bbh.score", + "metric_name": "BBH score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bbh", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BBH", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.843, + "details": { + "raw_score": "0.843", + "raw_score_field": "score", + "raw_model_id": "deepseek-v2.5", + "raw_benchmark_id": "bbh", + "source_urls_json": "[\"https://llm-stats.com/models/deepseek-v2.5\",\"https://llm-stats.com/benchmarks/bbh\",\"https://api.llm-stats.com/leaderboard/benchmarks/bbh\"]", + "raw_score_id": "bbh::deepseek-v2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ds-arena-code::ds-arena-code-deepseek-v2.5", + "evaluation_name": "llm_stats.ds-arena-code", + "source_data": { + "dataset_name": "DS-Arena-Code", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/deepseek-v2.5", + "https://llm-stats.com/benchmarks/ds-arena-code", + "https://api.llm-stats.com/leaderboard/benchmarks/ds-arena-code" + ], + "additional_details": { + "raw_benchmark_id": "ds-arena-code", + "raw_model_id": "deepseek-v2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Data Science Arena Code benchmark for evaluating LLMs on realistic data science code generation tasks. Tests capabilities in complex data processing, analysis, and programming across popular Python libraries used in data science workflows.", + "metric_id": "llm_stats.ds-arena-code.score", + "metric_name": "DS-Arena-Code score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ds-arena-code", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DS-Arena-Code", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.631, + "details": { + "raw_score": "0.631", + "raw_score_field": "score", + "raw_model_id": "deepseek-v2.5", + "raw_benchmark_id": "ds-arena-code", + "source_urls_json": "[\"https://llm-stats.com/models/deepseek-v2.5\",\"https://llm-stats.com/benchmarks/ds-arena-code\",\"https://api.llm-stats.com/leaderboard/benchmarks/ds-arena-code\"]", + "raw_score_id": "ds-arena-code::deepseek-v2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ds-fim-eval::ds-fim-eval-deepseek-v2.5", + "evaluation_name": "llm_stats.ds-fim-eval", + "source_data": { + "dataset_name": "DS-FIM-Eval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/deepseek-v2.5", + "https://llm-stats.com/benchmarks/ds-fim-eval", + "https://api.llm-stats.com/leaderboard/benchmarks/ds-fim-eval" + ], + "additional_details": { + "raw_benchmark_id": "ds-fim-eval", + "raw_model_id": "deepseek-v2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "DeepSeek's internal Fill-in-the-Middle evaluation dataset for measuring code completion performance improvements in data science contexts", + "metric_id": "llm_stats.ds-fim-eval.score", + "metric_name": "DS-FIM-Eval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ds-fim-eval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DS-FIM-Eval", + "raw_categories": "[\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.783, + "details": { + "raw_score": "0.783", + "raw_score_field": "score", + "raw_model_id": "deepseek-v2.5", + "raw_benchmark_id": "ds-fim-eval", + "source_urls_json": "[\"https://llm-stats.com/models/deepseek-v2.5\",\"https://llm-stats.com/benchmarks/ds-fim-eval\",\"https://api.llm-stats.com/leaderboard/benchmarks/ds-fim-eval\"]", + "raw_score_id": "ds-fim-eval::deepseek-v2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gsm8k::gsm8k-deepseek-v2.5", + "evaluation_name": "llm_stats.gsm8k", + "source_data": { + "dataset_name": "GSM8k", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/deepseek-v2.5", + "https://llm-stats.com/benchmarks/gsm8k", + "https://api.llm-stats.com/leaderboard/benchmarks/gsm8k" + ], + "additional_details": { + "raw_benchmark_id": "gsm8k", + "raw_model_id": "deepseek-v2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Grade School Math 8K, a dataset of 8.5K high-quality linguistically diverse grade school math word problems requiring multi-step reasoning and elementary arithmetic operations.", + "metric_id": "llm_stats.gsm8k.score", + "metric_name": "GSM8k score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gsm8k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GSM8k", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "48" + } + }, + "score_details": { + "score": 0.951, + "details": { + "raw_score": "0.951", + "raw_score_field": "score", + "raw_model_id": "deepseek-v2.5", + "raw_benchmark_id": "gsm8k", + "source_urls_json": "[\"https://llm-stats.com/models/deepseek-v2.5\",\"https://llm-stats.com/benchmarks/gsm8k\",\"https://api.llm-stats.com/leaderboard/benchmarks/gsm8k\"]", + "raw_score_id": "gsm8k::deepseek-v2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "humaneval-mul::humaneval-mul-deepseek-v2.5", + "evaluation_name": "llm_stats.humaneval-mul", + "source_data": { + "dataset_name": "HumanEval-Mul", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/deepseek-v2.5", + "https://llm-stats.com/benchmarks/humaneval-mul", + "https://api.llm-stats.com/leaderboard/benchmarks/humaneval-mul" + ], + "additional_details": { + "raw_benchmark_id": "humaneval-mul", + "raw_model_id": "deepseek-v2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multilingual variant of the HumanEval benchmark that measures functional correctness for synthesizing programs from docstrings, consisting of 164 original programming problems assessing language comprehension, algorithms, and simple mathematics", + "metric_id": "llm_stats.humaneval-mul.score", + "metric_name": "HumanEval-Mul score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "humaneval-mul", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HumanEval-Mul", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.738, + "details": { + "raw_score": "0.738", + "raw_score_field": "score", + "raw_model_id": "deepseek-v2.5", + "raw_benchmark_id": "humaneval-mul", + "source_urls_json": "[\"https://llm-stats.com/models/deepseek-v2.5\",\"https://llm-stats.com/benchmarks/humaneval-mul\",\"https://api.llm-stats.com/leaderboard/benchmarks/humaneval-mul\"]", + "raw_score_id": "humaneval-mul::deepseek-v2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "humaneval::humaneval-deepseek-v2.5", + "evaluation_name": "llm_stats.humaneval", + "source_data": { + "dataset_name": "HumanEval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/deepseek-v2.5", + "https://llm-stats.com/benchmarks/humaneval", + "https://api.llm-stats.com/leaderboard/benchmarks/humaneval" + ], + "additional_details": { + "raw_benchmark_id": "humaneval", + "raw_model_id": "deepseek-v2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A benchmark that measures functional correctness for synthesizing programs from docstrings, consisting of 164 original programming problems assessing language comprehension, algorithms, and simple mathematics", + "metric_id": "llm_stats.humaneval.score", + "metric_name": "HumanEval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "humaneval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HumanEval", + "raw_categories": "[\"reasoning\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "66" + } + }, + "score_details": { + "score": 0.89, + "details": { + "raw_score": "0.89", + "raw_score_field": "score", + "raw_model_id": "deepseek-v2.5", + "raw_benchmark_id": "humaneval", + "source_urls_json": "[\"https://llm-stats.com/models/deepseek-v2.5\",\"https://llm-stats.com/benchmarks/humaneval\",\"https://api.llm-stats.com/leaderboard/benchmarks/humaneval\"]", + "raw_score_id": "humaneval::deepseek-v2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "livecodebench-01-09::livecodebench-01-09-deepseek-v2.5", + "evaluation_name": "llm_stats.livecodebench-01-09", + "source_data": { + "dataset_name": "LiveCodeBench(01-09)", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/deepseek-v2.5", + "https://llm-stats.com/benchmarks/livecodebench-01-09", + "https://api.llm-stats.com/leaderboard/benchmarks/livecodebench(01-09)" + ], + "additional_details": { + "raw_benchmark_id": "livecodebench(01-09)", + "raw_model_id": "deepseek-v2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LiveCodeBench is a holistic and contamination-free evaluation benchmark for large language models for code. It continuously collects new problems from programming contests (LeetCode, AtCoder, CodeForces) and evaluates four different scenarios: code generation, self-repair, code execution, and test output prediction. Problems are annotated with release dates to enable evaluation on unseen problems released after a model's training cutoff.", + "metric_id": "llm_stats.livecodebench-01-09.score", + "metric_name": "LiveCodeBench(01-09) score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "livecodebench(01-09)", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LiveCodeBench(01-09)", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.418, + "details": { + "raw_score": "0.418", + "raw_score_field": "score", + "raw_model_id": "deepseek-v2.5", + "raw_benchmark_id": "livecodebench(01-09)", + "source_urls_json": "[\"https://llm-stats.com/models/deepseek-v2.5\",\"https://llm-stats.com/benchmarks/livecodebench-01-09\",\"https://api.llm-stats.com/leaderboard/benchmarks/livecodebench(01-09)\"]", + "raw_score_id": "livecodebench(01-09)::deepseek-v2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mt-bench::mt-bench-deepseek-v2.5", + "evaluation_name": "llm_stats.mt-bench", + "source_data": { + "dataset_name": "MT-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/deepseek-v2.5", + "https://llm-stats.com/benchmarks/mt-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/mt-bench" + ], + "additional_details": { + "raw_benchmark_id": "mt-bench", + "raw_model_id": "deepseek-v2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MT-Bench is a challenging multi-turn benchmark that measures the ability of large language models to engage in coherent, informative, and engaging conversations. It uses strong LLMs as judges for scalable and explainable evaluation of multi-turn dialogue capabilities.", + "metric_id": "llm_stats.mt-bench.score", + "metric_name": "MT-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mt-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MT-Bench", + "raw_categories": "[\"reasoning\",\"roleplay\",\"general\",\"communication\",\"creativity\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.902, + "details": { + "raw_score": "0.902", + "raw_score_field": "score", + "raw_model_id": "deepseek-v2.5", + "raw_benchmark_id": "mt-bench", + "source_urls_json": "[\"https://llm-stats.com/models/deepseek-v2.5\",\"https://llm-stats.com/benchmarks/mt-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/mt-bench\"]", + "raw_score_id": "mt-bench::deepseek-v2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/deepseek/deepseek-v4-pro-max/9ec86732-b59c-446b-8207-c7f1ab0662f5.json b/data/llm_stats/deepseek/deepseek-v4-pro-max/9ec86732-b59c-446b-8207-c7f1ab0662f5.json new file mode 100644 index 0000000000000000000000000000000000000000..3f141cb235c7db149eecc4b8e7e64f0bb77c5a48 --- /dev/null +++ b/data/llm_stats/deepseek/deepseek-v4-pro-max/9ec86732-b59c-446b-8207-c7f1ab0662f5.json @@ -0,0 +1,984 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/deepseek_deepseek-v4-pro-max/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v4-pro-max", + "id": "deepseek/deepseek-v4-pro-max", + "developer": "deepseek", + "additional_details": { + "raw_id": "deepseek-v4-pro-max", + "raw_model_name": "deepseek-v4-pro-max", + "raw_display_name": "DeepSeek-V4-Pro-Max", + "raw_organization_id": "deepseek", + "raw_organization_name": "DeepSeek", + "raw_context_length": "1048576", + "raw_input_modalities": "[\"text\"]", + "raw_output_modalities": "[\"text\"]", + "raw_input_price": "1.6", + "raw_output_price": "3.2", + "raw_model_id": "deepseek-v4-pro-max", + "raw_provider_slug": "deepseek", + "raw_provider_name": "ZeroEval" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "browsecomp::browsecomp-deepseek-v4-pro-max", + "evaluation_name": "llm_stats.browsecomp", + "source_data": { + "dataset_name": "BrowseComp", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/deepseek-v4-pro-max", + "https://llm-stats.com/benchmarks/browsecomp", + "https://api.llm-stats.com/leaderboard/benchmarks/browsecomp" + ], + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_model_id": "deepseek-v4-pro-max", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BrowseComp is a benchmark comprising 1,266 questions that challenge AI agents to persistently navigate the internet in search of hard-to-find, entangled information. The benchmark measures agents' ability to exercise persistence in information gathering, demonstrate creativity in web navigation, and find concise, verifiable answers. Despite the difficulty of the questions, BrowseComp is simple and easy-to-use, as predicted answers are short and easily verifiable against reference answers.", + "metric_id": "llm_stats.browsecomp.score", + "metric_name": "BrowseComp score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BrowseComp", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "52" + } + }, + "score_details": { + "score": 0.834, + "details": { + "raw_score": "0.834", + "raw_score_field": "score", + "raw_model_id": "deepseek-v4-pro-max", + "raw_benchmark_id": "browsecomp", + "source_urls_json": "[\"https://llm-stats.com/models/deepseek-v4-pro-max\",\"https://llm-stats.com/benchmarks/browsecomp\",\"https://api.llm-stats.com/leaderboard/benchmarks/browsecomp\"]", + "raw_score_id": "browsecomp::deepseek-v4-pro-max", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "codeforces::codeforces-deepseek-v4-pro-max", + "evaluation_name": "llm_stats.codeforces", + "source_data": { + "dataset_name": "CodeForces", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/deepseek-v4-pro-max", + "https://llm-stats.com/benchmarks/codeforces", + "https://api.llm-stats.com/leaderboard/benchmarks/codeforces" + ], + "additional_details": { + "raw_benchmark_id": "codeforces", + "raw_model_id": "deepseek-v4-pro-max", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A competitive programming benchmark using problems from the CodeForces platform. The benchmark evaluates code generation capabilities of LLMs on algorithmic problems with difficulty ratings ranging from 800 to 2400. Problems cover diverse algorithmic categories including dynamic programming, graph algorithms, data structures, and mathematical problems with standardized evaluation through direct platform submission.", + "metric_id": "llm_stats.codeforces.score", + "metric_name": "CodeForces score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "codeforces", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CodeForces", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 1.0, + "details": { + "raw_score": "1.0", + "raw_score_field": "score", + "raw_model_id": "deepseek-v4-pro-max", + "raw_benchmark_id": "codeforces", + "source_urls_json": "[\"https://llm-stats.com/models/deepseek-v4-pro-max\",\"https://llm-stats.com/benchmarks/codeforces\",\"https://api.llm-stats.com/leaderboard/benchmarks/codeforces\"]", + "raw_score_id": "codeforces::deepseek-v4-pro-max", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "corpusqa-1m::corpusqa-1m-deepseek-v4-pro-max", + "evaluation_name": "llm_stats.corpusqa-1m", + "source_data": { + "dataset_name": "CorpusQA 1M", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/deepseek-v4-pro-max", + "https://llm-stats.com/benchmarks/corpusqa-1m", + "https://api.llm-stats.com/leaderboard/benchmarks/corpusqa-1m" + ], + "additional_details": { + "raw_benchmark_id": "corpusqa-1m", + "raw_model_id": "deepseek-v4-pro-max", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CorpusQA 1M is a long-context question answering benchmark designed to evaluate models at approximately 1 million token contexts. Models are scored on accuracy when retrieving and reasoning over information distributed across an extremely long input corpus.", + "metric_id": "llm_stats.corpusqa-1m.score", + "metric_name": "CorpusQA 1M score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "corpusqa-1m", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CorpusQA 1M", + "raw_categories": "[\"long_context\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.62, + "details": { + "raw_score": "0.62", + "raw_score_field": "score", + "raw_model_id": "deepseek-v4-pro-max", + "raw_benchmark_id": "corpusqa-1m", + "source_urls_json": "[\"https://llm-stats.com/models/deepseek-v4-pro-max\",\"https://llm-stats.com/benchmarks/corpusqa-1m\",\"https://api.llm-stats.com/leaderboard/benchmarks/corpusqa-1m\"]", + "raw_score_id": "corpusqa-1m::deepseek-v4-pro-max", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "csimpleqa::csimpleqa-deepseek-v4-pro-max", + "evaluation_name": "llm_stats.csimpleqa", + "source_data": { + "dataset_name": "CSimpleQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/deepseek-v4-pro-max", + "https://llm-stats.com/benchmarks/csimpleqa", + "https://api.llm-stats.com/leaderboard/benchmarks/csimpleqa" + ], + "additional_details": { + "raw_benchmark_id": "csimpleqa", + "raw_model_id": "deepseek-v4-pro-max", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Chinese SimpleQA is the first comprehensive Chinese benchmark to evaluate the factuality ability of language models to answer short questions. It contains 3,000 high-quality questions spanning 6 major topics with 99 diverse subtopics, designed to assess Chinese factual knowledge across humanities, science, engineering, culture, and society.", + "metric_id": "llm_stats.csimpleqa.score", + "metric_name": "CSimpleQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "csimpleqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CSimpleQA", + "raw_categories": "[\"language\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.844, + "details": { + "raw_score": "0.844", + "raw_score_field": "score", + "raw_model_id": "deepseek-v4-pro-max", + "raw_benchmark_id": "csimpleqa", + "source_urls_json": "[\"https://llm-stats.com/models/deepseek-v4-pro-max\",\"https://llm-stats.com/benchmarks/csimpleqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/csimpleqa\"]", + "raw_score_id": "csimpleqa::deepseek-v4-pro-max", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gpqa::gpqa-deepseek-v4-pro-max", + "evaluation_name": "llm_stats.gpqa", + "source_data": { + "dataset_name": "GPQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/deepseek-v4-pro-max", + "https://llm-stats.com/benchmarks/gpqa", + "https://api.llm-stats.com/leaderboard/benchmarks/gpqa" + ], + "additional_details": { + "raw_benchmark_id": "gpqa", + "raw_model_id": "deepseek-v4-pro-max", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A challenging dataset of 448 multiple-choice questions written by domain experts in biology, physics, and chemistry. Questions are Google-proof and extremely difficult, with PhD experts reaching 65% accuracy.", + "metric_id": "llm_stats.gpqa.score", + "metric_name": "GPQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gpqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GPQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\",\"biology\",\"chemistry\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "226" + } + }, + "score_details": { + "score": 0.901, + "details": { + "raw_score": "0.901", + "raw_score_field": "score", + "raw_model_id": "deepseek-v4-pro-max", + "raw_benchmark_id": "gpqa", + "source_urls_json": "[\"https://llm-stats.com/models/deepseek-v4-pro-max\",\"https://llm-stats.com/benchmarks/gpqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/gpqa\"]", + "raw_score_id": "gpqa::deepseek-v4-pro-max", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hmmt-feb-26::hmmt-feb-26-deepseek-v4-pro-max", + "evaluation_name": "llm_stats.hmmt-feb-26", + "source_data": { + "dataset_name": "HMMT Feb 26", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/deepseek-v4-pro-max", + "https://llm-stats.com/benchmarks/hmmt-feb-26", + "https://api.llm-stats.com/leaderboard/benchmarks/hmmt-feb-26" + ], + "additional_details": { + "raw_benchmark_id": "hmmt-feb-26", + "raw_model_id": "deepseek-v4-pro-max", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "HMMT February 2026 is a math competition benchmark based on problems from the Harvard-MIT Mathematics Tournament, testing advanced mathematical problem-solving and reasoning.", + "metric_id": "llm_stats.hmmt-feb-26.score", + "metric_name": "HMMT Feb 26 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hmmt-feb-26", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HMMT Feb 26", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.952, + "details": { + "raw_score": "0.952", + "raw_score_field": "score", + "raw_model_id": "deepseek-v4-pro-max", + "raw_benchmark_id": "hmmt-feb-26", + "source_urls_json": "[\"https://llm-stats.com/models/deepseek-v4-pro-max\",\"https://llm-stats.com/benchmarks/hmmt-feb-26\",\"https://api.llm-stats.com/leaderboard/benchmarks/hmmt-feb-26\"]", + "raw_score_id": "hmmt-feb-26::deepseek-v4-pro-max", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "humanity-s-last-exam::humanity-s-last-exam-deepseek-v4-pro-max", + "evaluation_name": "llm_stats.humanity-s-last-exam", + "source_data": { + "dataset_name": "Humanity's Last Exam", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/deepseek-v4-pro-max", + "https://llm-stats.com/benchmarks/humanity-s-last-exam", + "https://api.llm-stats.com/leaderboard/benchmarks/humanity's-last-exam" + ], + "additional_details": { + "raw_benchmark_id": "humanity's-last-exam", + "raw_model_id": "deepseek-v4-pro-max", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Humanity's Last Exam (HLE) is a multi-modal academic benchmark with 2,500 questions across mathematics, humanities, and natural sciences, designed to test LLM capabilities at the frontier of human knowledge with unambiguous, verifiable solutions", + "metric_id": "llm_stats.humanity-s-last-exam.score", + "metric_name": "Humanity's Last Exam score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "humanity's-last-exam", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Humanity's Last Exam", + "raw_categories": "[\"math\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "88" + } + }, + "score_details": { + "score": 0.482, + "details": { + "raw_score": "0.482", + "raw_score_field": "score", + "raw_model_id": "deepseek-v4-pro-max", + "raw_benchmark_id": "humanity's-last-exam", + "source_urls_json": "[\"https://llm-stats.com/models/deepseek-v4-pro-max\",\"https://llm-stats.com/benchmarks/humanity-s-last-exam\",\"https://api.llm-stats.com/leaderboard/benchmarks/humanity's-last-exam\"]", + "raw_score_id": "humanity's-last-exam::deepseek-v4-pro-max", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "imo-answerbench::imo-answerbench-deepseek-v4-pro-max", + "evaluation_name": "llm_stats.imo-answerbench", + "source_data": { + "dataset_name": "IMO-AnswerBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/deepseek-v4-pro-max", + "https://llm-stats.com/benchmarks/imo-answerbench", + "https://api.llm-stats.com/leaderboard/benchmarks/imo-answerbench" + ], + "additional_details": { + "raw_benchmark_id": "imo-answerbench", + "raw_model_id": "deepseek-v4-pro-max", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "IMO-AnswerBench is a benchmark for evaluating mathematical reasoning capabilities on International Mathematical Olympiad (IMO) problems, focusing on answer generation and verification.", + "metric_id": "llm_stats.imo-answerbench.score", + "metric_name": "IMO-AnswerBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "imo-answerbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IMO-AnswerBench", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.898, + "details": { + "raw_score": "0.898", + "raw_score_field": "score", + "raw_model_id": "deepseek-v4-pro-max", + "raw_benchmark_id": "imo-answerbench", + "source_urls_json": "[\"https://llm-stats.com/models/deepseek-v4-pro-max\",\"https://llm-stats.com/benchmarks/imo-answerbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/imo-answerbench\"]", + "raw_score_id": "imo-answerbench::deepseek-v4-pro-max", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "livecodebench::livecodebench-deepseek-v4-pro-max", + "evaluation_name": "llm_stats.livecodebench", + "source_data": { + "dataset_name": "LiveCodeBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/deepseek-v4-pro-max", + "https://llm-stats.com/benchmarks/livecodebench", + "https://api.llm-stats.com/leaderboard/benchmarks/livecodebench" + ], + "additional_details": { + "raw_benchmark_id": "livecodebench", + "raw_model_id": "deepseek-v4-pro-max", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LiveCodeBench is a holistic and contamination-free evaluation benchmark for large language models for code. It continuously collects new problems from programming contests (LeetCode, AtCoder, CodeForces) and evaluates four different scenarios: code generation, self-repair, code execution, and test output prediction. Problems are annotated with release dates to enable evaluation on unseen problems released after a model's training cutoff.", + "metric_id": "llm_stats.livecodebench.score", + "metric_name": "LiveCodeBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "livecodebench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LiveCodeBench", + "raw_categories": "[\"reasoning\",\"general\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "73" + } + }, + "score_details": { + "score": 0.935, + "details": { + "raw_score": "0.935", + "raw_score_field": "score", + "raw_model_id": "deepseek-v4-pro-max", + "raw_benchmark_id": "livecodebench", + "source_urls_json": "[\"https://llm-stats.com/models/deepseek-v4-pro-max\",\"https://llm-stats.com/benchmarks/livecodebench\",\"https://api.llm-stats.com/leaderboard/benchmarks/livecodebench\"]", + "raw_score_id": "livecodebench::deepseek-v4-pro-max", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "matharena-apex::matharena-apex-deepseek-v4-pro-max", + "evaluation_name": "llm_stats.matharena-apex", + "source_data": { + "dataset_name": "MathArena Apex", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/deepseek-v4-pro-max", + "https://llm-stats.com/benchmarks/matharena-apex", + "https://api.llm-stats.com/leaderboard/benchmarks/matharena-apex" + ], + "additional_details": { + "raw_benchmark_id": "matharena-apex", + "raw_model_id": "deepseek-v4-pro-max", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MathArena Apex is a challenging math contest benchmark featuring the most difficult mathematical problems designed to test advanced reasoning and problem-solving abilities of AI models. It focuses on olympiad-level mathematics and complex multi-step mathematical reasoning.", + "metric_id": "llm_stats.matharena-apex.score", + "metric_name": "MathArena Apex score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "matharena-apex", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MathArena Apex", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.902, + "details": { + "raw_score": "0.902", + "raw_score_field": "score", + "raw_model_id": "deepseek-v4-pro-max", + "raw_benchmark_id": "matharena-apex", + "source_urls_json": "[\"https://llm-stats.com/models/deepseek-v4-pro-max\",\"https://llm-stats.com/benchmarks/matharena-apex\",\"https://api.llm-stats.com/leaderboard/benchmarks/matharena-apex\"]", + "raw_score_id": "matharena-apex::deepseek-v4-pro-max", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mcp-atlas::mcp-atlas-deepseek-v4-pro-max", + "evaluation_name": "llm_stats.mcp-atlas", + "source_data": { + "dataset_name": "MCP Atlas", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/deepseek-v4-pro-max", + "https://llm-stats.com/benchmarks/mcp-atlas", + "https://api.llm-stats.com/leaderboard/benchmarks/mcp-atlas" + ], + "additional_details": { + "raw_benchmark_id": "mcp-atlas", + "raw_model_id": "deepseek-v4-pro-max", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MCP Atlas is a benchmark for evaluating AI models on scaled tool use capabilities, measuring how well models can coordinate and utilize multiple tools across complex multi-step tasks.", + "metric_id": "llm_stats.mcp-atlas.score", + "metric_name": "MCP Atlas score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mcp-atlas", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MCP Atlas", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.736, + "details": { + "raw_score": "0.736", + "raw_score_field": "score", + "raw_model_id": "deepseek-v4-pro-max", + "raw_benchmark_id": "mcp-atlas", + "source_urls_json": "[\"https://llm-stats.com/models/deepseek-v4-pro-max\",\"https://llm-stats.com/benchmarks/mcp-atlas\",\"https://api.llm-stats.com/leaderboard/benchmarks/mcp-atlas\"]", + "raw_score_id": "mcp-atlas::deepseek-v4-pro-max", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu-pro::mmlu-pro-deepseek-v4-pro-max", + "evaluation_name": "llm_stats.mmlu-pro", + "source_data": { + "dataset_name": "MMLU-Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/deepseek-v4-pro-max", + "https://llm-stats.com/benchmarks/mmlu-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu-pro" + ], + "additional_details": { + "raw_benchmark_id": "mmlu-pro", + "raw_model_id": "deepseek-v4-pro-max", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A more robust and challenging multi-task language understanding benchmark that extends MMLU by expanding multiple-choice options from 4 to 10, eliminating trivial questions, and focusing on reasoning-intensive tasks. Features over 12,000 curated questions across 14 domains and causes a 16-33% accuracy drop compared to original MMLU.", + "metric_id": "llm_stats.mmlu-pro.score", + "metric_name": "MMLU-Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU-Pro", + "raw_categories": "[\"language\",\"legal\",\"math\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "129" + } + }, + "score_details": { + "score": 0.875, + "details": { + "raw_score": "0.875", + "raw_score_field": "score", + "raw_model_id": "deepseek-v4-pro-max", + "raw_benchmark_id": "mmlu-pro", + "source_urls_json": "[\"https://llm-stats.com/models/deepseek-v4-pro-max\",\"https://llm-stats.com/benchmarks/mmlu-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu-pro\"]", + "raw_score_id": "mmlu-pro::deepseek-v4-pro-max", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mrcr-1m::mrcr-1m-deepseek-v4-pro-max", + "evaluation_name": "llm_stats.mrcr-1m", + "source_data": { + "dataset_name": "MRCR 1M", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/deepseek-v4-pro-max", + "https://llm-stats.com/benchmarks/mrcr-1m", + "https://api.llm-stats.com/leaderboard/benchmarks/mrcr-1m" + ], + "additional_details": { + "raw_benchmark_id": "mrcr-1m", + "raw_model_id": "deepseek-v4-pro-max", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MRCR 1M is a variant of the Multi-Round Coreference Resolution benchmark designed for testing extremely long context capabilities with approximately 1 million tokens. It evaluates models' ability to maintain reasoning and attention across ultra-long conversations.", + "metric_id": "llm_stats.mrcr-1m.score", + "metric_name": "MRCR 1M score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mrcr-1m", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MRCR 1M", + "raw_categories": "[\"long_context\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.835, + "details": { + "raw_score": "0.835", + "raw_score_field": "score", + "raw_model_id": "deepseek-v4-pro-max", + "raw_benchmark_id": "mrcr-1m", + "source_urls_json": "[\"https://llm-stats.com/models/deepseek-v4-pro-max\",\"https://llm-stats.com/benchmarks/mrcr-1m\",\"https://api.llm-stats.com/leaderboard/benchmarks/mrcr-1m\"]", + "raw_score_id": "mrcr-1m::deepseek-v4-pro-max", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "simpleqa::simpleqa-deepseek-v4-pro-max", + "evaluation_name": "llm_stats.simpleqa", + "source_data": { + "dataset_name": "SimpleQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/deepseek-v4-pro-max", + "https://llm-stats.com/benchmarks/simpleqa", + "https://api.llm-stats.com/leaderboard/benchmarks/simpleqa" + ], + "additional_details": { + "raw_benchmark_id": "simpleqa", + "raw_model_id": "deepseek-v4-pro-max", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SimpleQA is a factuality benchmark developed by OpenAI that measures the short-form factual accuracy of large language models. The benchmark contains 4,326 short, fact-seeking questions that are adversarially collected and designed to have single, indisputable answers. Questions cover diverse topics from science and technology to entertainment, and the benchmark also measures model calibration by evaluating whether models know what they know.", + "metric_id": "llm_stats.simpleqa.score", + "metric_name": "SimpleQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "simpleqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SimpleQA", + "raw_categories": "[\"reasoning\",\"factuality\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "46" + } + }, + "score_details": { + "score": 0.579, + "details": { + "raw_score": "0.579", + "raw_score_field": "score", + "raw_model_id": "deepseek-v4-pro-max", + "raw_benchmark_id": "simpleqa", + "source_urls_json": "[\"https://llm-stats.com/models/deepseek-v4-pro-max\",\"https://llm-stats.com/benchmarks/simpleqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/simpleqa\"]", + "raw_score_id": "simpleqa::deepseek-v4-pro-max", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-multilingual::swe-bench-multilingual-deepseek-v4-pro-max", + "evaluation_name": "llm_stats.swe-bench-multilingual", + "source_data": { + "dataset_name": "SWE-bench Multilingual", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/deepseek-v4-pro-max", + "https://llm-stats.com/benchmarks/swe-bench-multilingual", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-multilingual" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-multilingual", + "raw_model_id": "deepseek-v4-pro-max", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multilingual benchmark for issue resolving in software engineering that covers Java, TypeScript, JavaScript, Go, Rust, C, and C++. Contains 1,632 high-quality instances carefully annotated from 2,456 candidates by 68 expert annotators, designed to evaluate Large Language Models across diverse software ecosystems beyond Python.", + "metric_id": "llm_stats.swe-bench-multilingual.score", + "metric_name": "SWE-bench Multilingual score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-multilingual", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-bench Multilingual", + "raw_categories": "[\"reasoning\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "33" + } + }, + "score_details": { + "score": 0.762, + "details": { + "raw_score": "0.762", + "raw_score_field": "score", + "raw_model_id": "deepseek-v4-pro-max", + "raw_benchmark_id": "swe-bench-multilingual", + "source_urls_json": "[\"https://llm-stats.com/models/deepseek-v4-pro-max\",\"https://llm-stats.com/benchmarks/swe-bench-multilingual\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-multilingual\"]", + "raw_score_id": "swe-bench-multilingual::deepseek-v4-pro-max", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-verified::swe-bench-verified-deepseek-v4-pro-max", + "evaluation_name": "llm_stats.swe-bench-verified", + "source_data": { + "dataset_name": "SWE-Bench Verified", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/deepseek-v4-pro-max", + "https://llm-stats.com/benchmarks/swe-bench-verified", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-verified" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-verified", + "raw_model_id": "deepseek-v4-pro-max", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A verified subset of 500 software engineering problems from real GitHub issues, validated by human annotators for evaluating language models' ability to resolve real-world coding issues by generating patches for Python codebases.", + "metric_id": "llm_stats.swe-bench-verified.score", + "metric_name": "SWE-Bench Verified score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-verified", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Bench Verified", + "raw_categories": "[\"reasoning\",\"frontend_development\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "103" + } + }, + "score_details": { + "score": 0.806, + "details": { + "raw_score": "0.806", + "raw_score_field": "score", + "raw_model_id": "deepseek-v4-pro-max", + "raw_benchmark_id": "swe-bench-verified", + "source_urls_json": "[\"https://llm-stats.com/models/deepseek-v4-pro-max\",\"https://llm-stats.com/benchmarks/swe-bench-verified\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-verified\"]", + "raw_score_id": "swe-bench-verified::deepseek-v4-pro-max", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2::terminal-bench-2-deepseek-v4-pro-max", + "evaluation_name": "llm_stats.terminal-bench-2", + "source_data": { + "dataset_name": "Terminal-Bench 2.0", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/deepseek-v4-pro-max", + "https://llm-stats.com/benchmarks/terminal-bench-2", + "https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2" + ], + "additional_details": { + "raw_benchmark_id": "terminal-bench-2", + "raw_model_id": "deepseek-v4-pro-max", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Terminal-Bench 2.0 is an updated benchmark for testing AI agents' tool use ability to operate a computer via terminal. It evaluates how well models can handle real-world, end-to-end tasks autonomously, including compiling code, training models, setting up servers, system administration, security tasks, data science workflows, and cybersecurity vulnerabilities.", + "metric_id": "llm_stats.terminal-bench-2.score", + "metric_name": "Terminal-Bench 2.0 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "terminal-bench-2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Terminal-Bench 2.0", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "49" + } + }, + "score_details": { + "score": 0.679, + "details": { + "raw_score": "0.679", + "raw_score_field": "score", + "raw_model_id": "deepseek-v4-pro-max", + "raw_benchmark_id": "terminal-bench-2", + "source_urls_json": "[\"https://llm-stats.com/models/deepseek-v4-pro-max\",\"https://llm-stats.com/benchmarks/terminal-bench-2\",\"https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2\"]", + "raw_score_id": "terminal-bench-2::deepseek-v4-pro-max", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "toolathlon::toolathlon-deepseek-v4-pro-max", + "evaluation_name": "llm_stats.toolathlon", + "source_data": { + "dataset_name": "Toolathlon", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/deepseek-v4-pro-max", + "https://llm-stats.com/benchmarks/toolathlon", + "https://api.llm-stats.com/leaderboard/benchmarks/toolathlon" + ], + "additional_details": { + "raw_benchmark_id": "toolathlon", + "raw_model_id": "deepseek-v4-pro-max", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Tool Decathlon is a comprehensive benchmark for evaluating AI agents' ability to use multiple tools across diverse task categories. It measures proficiency in tool selection, sequencing, and execution across ten different tool-use scenarios.", + "metric_id": "llm_stats.toolathlon.score", + "metric_name": "Toolathlon score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "toolathlon", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Toolathlon", + "raw_categories": "[\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "24" + } + }, + "score_details": { + "score": 0.518, + "details": { + "raw_score": "0.518", + "raw_score_field": "score", + "raw_model_id": "deepseek-v4-pro-max", + "raw_benchmark_id": "toolathlon", + "source_urls_json": "[\"https://llm-stats.com/models/deepseek-v4-pro-max\",\"https://llm-stats.com/benchmarks/toolathlon\",\"https://api.llm-stats.com/leaderboard/benchmarks/toolathlon\"]", + "raw_score_id": "toolathlon::deepseek-v4-pro-max", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/google/diffusiongemma-26b-a4b-it/34e7228d-cc4c-4db3-b56d-07a13f2abb7e.json b/data/llm_stats/google/diffusiongemma-26b-a4b-it/34e7228d-cc4c-4db3-b56d-07a13f2abb7e.json new file mode 100644 index 0000000000000000000000000000000000000000..95be227a219f1fe48a27ed780be2b032296a6dbf --- /dev/null +++ b/data/llm_stats/google/diffusiongemma-26b-a4b-it/34e7228d-cc4c-4db3-b56d-07a13f2abb7e.json @@ -0,0 +1,463 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/google_diffusiongemma-26b-a4b-it/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "DiffusionGemma 26B-A4B", + "id": "google/diffusiongemma-26b-a4b-it", + "developer": "google", + "additional_details": { + "raw_id": "diffusiongemma-26b-a4b-it", + "raw_name": "DiffusionGemma 26B-A4B", + "raw_model_id": "diffusiongemma-26b-a4b-it", + "raw_model_name": "DiffusionGemma 26B-A4B", + "raw_organization_id": "google", + "raw_organization_name": "Google", + "raw_release_date": "2026-06-10", + "raw_announcement_date": "2026-06-10", + "raw_multimodal": "true", + "raw_param_count": "25200000000", + "raw_provider_slug": "google", + "raw_provider_name": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aime-2026::aime-2026-diffusiongemma-26b-a4b-it", + "evaluation_name": "llm_stats.aime-2026", + "source_data": { + "dataset_name": "AIME 2026", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/diffusiongemma-26b-a4b-it", + "https://llm-stats.com/benchmarks/aime-2026", + "https://api.llm-stats.com/leaderboard/benchmarks/aime-2026" + ], + "additional_details": { + "raw_benchmark_id": "aime-2026", + "raw_model_id": "diffusiongemma-26b-a4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "All 30 problems from the 2026 American Invitational Mathematics Examination (AIME I and AIME II), testing olympiad-level mathematical reasoning with integer answers from 000-999. Used as an AI benchmark to evaluate large language models' ability to solve complex mathematical problems requiring multi-step logical deductions and structured symbolic reasoning.", + "metric_id": "llm_stats.aime-2026.score", + "metric_name": "AIME 2026 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aime-2026", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AIME 2026", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "17" + } + }, + "score_details": { + "score": 0.691, + "details": { + "raw_score": "0.691", + "raw_score_field": "score", + "raw_model_id": "diffusiongemma-26b-a4b-it", + "raw_benchmark_id": "aime-2026", + "source_urls_json": "[\"https://llm-stats.com/models/diffusiongemma-26b-a4b-it\",\"https://llm-stats.com/benchmarks/aime-2026\",\"https://api.llm-stats.com/leaderboard/benchmarks/aime-2026\"]", + "raw_score_id": "aime-2026::diffusiongemma-26b-a4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "big-bench-extra-hard::big-bench-extra-hard-diffusiongemma-26b-a4b-it", + "evaluation_name": "llm_stats.big-bench-extra-hard", + "source_data": { + "dataset_name": "BIG-Bench Extra Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/diffusiongemma-26b-a4b-it", + "https://llm-stats.com/benchmarks/big-bench-extra-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench-extra-hard" + ], + "additional_details": { + "raw_benchmark_id": "big-bench-extra-hard", + "raw_model_id": "diffusiongemma-26b-a4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIG-Bench Extra Hard (BBEH) is a challenging benchmark that replaces each task in BIG-Bench Hard with a novel task that probes similar reasoning capabilities but exhibits significantly increased difficulty. The benchmark contains 23 tasks testing diverse reasoning skills including many-hop reasoning, causal understanding, spatial reasoning, temporal arithmetic, geometric reasoning, linguistic reasoning, logic puzzles, and humor understanding. Designed to address saturation on existing benchmarks where state-of-the-art models achieve near-perfect scores, BBEH shows substantial room for improvement with best models achieving only 9.8-44.8% average accuracy.", + "metric_id": "llm_stats.big-bench-extra-hard.score", + "metric_name": "BIG-Bench Extra Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench-extra-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench Extra Hard", + "raw_categories": "[\"language\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.476, + "details": { + "raw_score": "0.476", + "raw_score_field": "score", + "raw_model_id": "diffusiongemma-26b-a4b-it", + "raw_benchmark_id": "big-bench-extra-hard", + "source_urls_json": "[\"https://llm-stats.com/models/diffusiongemma-26b-a4b-it\",\"https://llm-stats.com/benchmarks/big-bench-extra-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench-extra-hard\"]", + "raw_score_id": "big-bench-extra-hard::diffusiongemma-26b-a4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "codeforces::codeforces-diffusiongemma-26b-a4b-it", + "evaluation_name": "llm_stats.codeforces", + "source_data": { + "dataset_name": "CodeForces", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/diffusiongemma-26b-a4b-it", + "https://llm-stats.com/benchmarks/codeforces", + "https://api.llm-stats.com/leaderboard/benchmarks/codeforces" + ], + "additional_details": { + "raw_benchmark_id": "codeforces", + "raw_model_id": "diffusiongemma-26b-a4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A competitive programming benchmark using problems from the CodeForces platform. The benchmark evaluates code generation capabilities of LLMs on algorithmic problems with difficulty ratings ranging from 800 to 2400. Problems cover diverse algorithmic categories including dynamic programming, graph algorithms, data structures, and mathematical problems with standardized evaluation through direct platform submission.", + "metric_id": "llm_stats.codeforces.score", + "metric_name": "CodeForces score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "codeforces", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CodeForces", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.4763, + "details": { + "raw_score": "0.4763", + "raw_score_field": "score", + "raw_model_id": "diffusiongemma-26b-a4b-it", + "raw_benchmark_id": "codeforces", + "source_urls_json": "[\"https://llm-stats.com/models/diffusiongemma-26b-a4b-it\",\"https://llm-stats.com/benchmarks/codeforces\",\"https://api.llm-stats.com/leaderboard/benchmarks/codeforces\"]", + "raw_score_id": "codeforces::diffusiongemma-26b-a4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mathvision::mathvision-diffusiongemma-26b-a4b-it", + "evaluation_name": "llm_stats.mathvision", + "source_data": { + "dataset_name": "MathVision", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/diffusiongemma-26b-a4b-it", + "https://llm-stats.com/benchmarks/mathvision", + "https://api.llm-stats.com/leaderboard/benchmarks/mathvision" + ], + "additional_details": { + "raw_benchmark_id": "mathvision", + "raw_model_id": "diffusiongemma-26b-a4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MATH-Vision is a dataset designed to measure multimodal mathematical reasoning capabilities. It focuses on evaluating how well models can solve mathematical problems that require both visual understanding and mathematical reasoning, bridging the gap between visual and mathematical domains.", + "metric_id": "llm_stats.mathvision.score", + "metric_name": "MathVision score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mathvision", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MathVision", + "raw_categories": "[\"math\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "31" + } + }, + "score_details": { + "score": 0.705, + "details": { + "raw_score": "0.705", + "raw_score_field": "score", + "raw_model_id": "diffusiongemma-26b-a4b-it", + "raw_benchmark_id": "mathvision", + "source_urls_json": "[\"https://llm-stats.com/models/diffusiongemma-26b-a4b-it\",\"https://llm-stats.com/benchmarks/mathvision\",\"https://api.llm-stats.com/leaderboard/benchmarks/mathvision\"]", + "raw_score_id": "mathvision::diffusiongemma-26b-a4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "medxpertqa::medxpertqa-diffusiongemma-26b-a4b-it", + "evaluation_name": "llm_stats.medxpertqa", + "source_data": { + "dataset_name": "MedXpertQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/diffusiongemma-26b-a4b-it", + "https://llm-stats.com/benchmarks/medxpertqa", + "https://api.llm-stats.com/leaderboard/benchmarks/medxpertqa" + ], + "additional_details": { + "raw_benchmark_id": "medxpertqa", + "raw_model_id": "diffusiongemma-26b-a4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A comprehensive benchmark to evaluate expert-level medical knowledge and advanced reasoning, featuring 4,460 questions spanning 17 specialties and 11 body systems. Includes both text-only and multimodal subsets with expert-level exam questions incorporating diverse medical images and rich clinical information.", + "metric_id": "llm_stats.medxpertqa.score", + "metric_name": "MedXpertQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "medxpertqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MedXpertQA", + "raw_categories": "[\"multimodal\",\"reasoning\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.49, + "details": { + "raw_score": "0.49", + "raw_score_field": "score", + "raw_model_id": "diffusiongemma-26b-a4b-it", + "raw_benchmark_id": "medxpertqa", + "source_urls_json": "[\"https://llm-stats.com/models/diffusiongemma-26b-a4b-it\",\"https://llm-stats.com/benchmarks/medxpertqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/medxpertqa\"]", + "raw_score_id": "medxpertqa::diffusiongemma-26b-a4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mrcr-v2::mrcr-v2-diffusiongemma-26b-a4b-it", + "evaluation_name": "llm_stats.mrcr-v2", + "source_data": { + "dataset_name": "MRCR v2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/diffusiongemma-26b-a4b-it", + "https://llm-stats.com/benchmarks/mrcr-v2", + "https://api.llm-stats.com/leaderboard/benchmarks/mrcr-v2" + ], + "additional_details": { + "raw_benchmark_id": "mrcr-v2", + "raw_model_id": "diffusiongemma-26b-a4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MRCR v2 (Multi-Round Coreference Resolution version 2) is an enhanced version of the synthetic long-context reasoning task. It extends the original MRCR framework with improved evaluation criteria and additional complexity for testing models' ability to maintain attention and reasoning across extended contexts.", + "metric_id": "llm_stats.mrcr-v2.score", + "metric_name": "MRCR v2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mrcr-v2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MRCR v2", + "raw_categories": "[\"long_context\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.32, + "details": { + "raw_score": "0.32", + "raw_score_field": "score", + "raw_model_id": "diffusiongemma-26b-a4b-it", + "raw_benchmark_id": "mrcr-v2", + "source_urls_json": "[\"https://llm-stats.com/models/diffusiongemma-26b-a4b-it\",\"https://llm-stats.com/benchmarks/mrcr-v2\",\"https://api.llm-stats.com/leaderboard/benchmarks/mrcr-v2\"]", + "raw_score_id": "mrcr-v2::diffusiongemma-26b-a4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "omnidocbench-1.5::omnidocbench-1.5-diffusiongemma-26b-a4b-it", + "evaluation_name": "llm_stats.omnidocbench-1.5", + "source_data": { + "dataset_name": "OmniDocBench 1.5", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/diffusiongemma-26b-a4b-it", + "https://llm-stats.com/benchmarks/omnidocbench-1.5", + "https://api.llm-stats.com/leaderboard/benchmarks/omnidocbench-1.5" + ], + "additional_details": { + "raw_benchmark_id": "omnidocbench-1.5", + "raw_model_id": "diffusiongemma-26b-a4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OmniDocBench 1.5 is a comprehensive benchmark for evaluating multimodal large language models on document understanding tasks, including OCR, document parsing, information extraction, and visual question answering across diverse document types. Lower Overall Edit Distance scores are better.", + "metric_id": "llm_stats.omnidocbench-1.5.score", + "metric_name": "OmniDocBench 1.5 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "omnidocbench-1.5", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OmniDocBench 1.5", + "raw_categories": "[\"multimodal\",\"reasoning\",\"structured_output\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.319, + "details": { + "raw_score": "0.319", + "raw_score_field": "score", + "raw_model_id": "diffusiongemma-26b-a4b-it", + "raw_benchmark_id": "omnidocbench-1.5", + "source_urls_json": "[\"https://llm-stats.com/models/diffusiongemma-26b-a4b-it\",\"https://llm-stats.com/benchmarks/omnidocbench-1.5\",\"https://api.llm-stats.com/leaderboard/benchmarks/omnidocbench-1.5\"]", + "raw_score_id": "omnidocbench-1.5::diffusiongemma-26b-a4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "t2-bench::t2-bench-diffusiongemma-26b-a4b-it", + "evaluation_name": "llm_stats.t2-bench", + "source_data": { + "dataset_name": "t2-bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/diffusiongemma-26b-a4b-it", + "https://llm-stats.com/benchmarks/t2-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/t2-bench" + ], + "additional_details": { + "raw_benchmark_id": "t2-bench", + "raw_model_id": "diffusiongemma-26b-a4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "t2-bench is a benchmark for evaluating agentic tool use capabilities, measuring how well models can select, sequence, and utilize tools to solve complex tasks. It tests autonomous planning and execution in multi-step scenarios.", + "metric_id": "llm_stats.t2-bench.score", + "metric_name": "t2-bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "t2-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "t2-bench", + "raw_categories": "[\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.562, + "details": { + "raw_score": "0.562", + "raw_score_field": "score", + "raw_model_id": "diffusiongemma-26b-a4b-it", + "raw_benchmark_id": "t2-bench", + "source_urls_json": "[\"https://llm-stats.com/models/diffusiongemma-26b-a4b-it\",\"https://llm-stats.com/benchmarks/t2-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/t2-bench\"]", + "raw_score_id": "t2-bench::diffusiongemma-26b-a4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/google/gemini-1.0-pro/626ff0bb-352f-4e77-ae58-2321b4ec106a.json b/data/llm_stats/google/gemini-1.0-pro/626ff0bb-352f-4e77-ae58-2321b4ec106a.json new file mode 100644 index 0000000000000000000000000000000000000000..cc82104715e8f22cdcf3470df28f23affa0b3f1c --- /dev/null +++ b/data/llm_stats/google/gemini-1.0-pro/626ff0bb-352f-4e77-ae58-2321b4ec106a.json @@ -0,0 +1,202 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/other/google_gemini-1.0-pro/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: other scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "other", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Gemini 1.0 Pro", + "id": "google/gemini-1.0-pro", + "developer": "google", + "additional_details": { + "raw_id": "gemini-1.0-pro", + "raw_name": "Gemini 1.0 Pro", + "raw_model_id": "gemini-1.0-pro", + "raw_model_name": "Gemini 1.0 Pro", + "raw_organization_id": "google", + "raw_organization_name": "Google", + "raw_release_date": "2024-02-15", + "raw_announcement_date": "2024-02-15", + "raw_multimodal": "false", + "raw_provider_slug": "google", + "raw_provider_name": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "big-bench::big-bench-gemini-1.0-pro", + "evaluation_name": "llm_stats.big-bench", + "source_data": { + "dataset_name": "BIG-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.0-pro", + "https://llm-stats.com/benchmarks/big-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench" + ], + "additional_details": { + "raw_benchmark_id": "big-bench", + "raw_model_id": "gemini-1.0-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark consisting of 204+ tasks designed to probe large language models and extrapolate their future capabilities. It covers diverse domains including linguistics, mathematics, common-sense reasoning, biology, physics, social bias, software development, and more. The benchmark focuses on tasks believed to be beyond current language model capabilities and includes both English and non-English tasks across multiple languages.", + "metric_id": "llm_stats.big-bench.score", + "metric_name": "BIG-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.75, + "details": { + "raw_score": "0.75", + "raw_score_field": "score", + "raw_model_id": "gemini-1.0-pro", + "raw_benchmark_id": "big-bench", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.0-pro\",\"https://llm-stats.com/benchmarks/big-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench\"]", + "raw_score_id": "big-bench::gemini-1.0-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "fleurs::fleurs-gemini-1.0-pro", + "evaluation_name": "llm_stats.fleurs", + "source_data": { + "dataset_name": "FLEURS", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.0-pro", + "https://llm-stats.com/benchmarks/fleurs", + "https://api.llm-stats.com/leaderboard/benchmarks/fleurs" + ], + "additional_details": { + "raw_benchmark_id": "fleurs", + "raw_model_id": "gemini-1.0-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Few-shot Learning Evaluation of Universal Representations of Speech - a parallel speech dataset in 102 languages built on FLoRes-101 with approximately 12 hours of speech supervision per language for tasks including ASR, speech language identification, translation and retrieval. Scores are shown as speech recognition accuracy (1 - word error rate), so higher is better.", + "metric_id": "llm_stats.fleurs.score", + "metric_name": "FLEURS score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "fleurs", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FLEURS", + "raw_categories": "[\"language\",\"speech_to_text\"]", + "raw_modality": "audio", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.936, + "details": { + "raw_score": "0.936", + "raw_score_field": "score", + "raw_model_id": "gemini-1.0-pro", + "raw_benchmark_id": "fleurs", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.0-pro\",\"https://llm-stats.com/benchmarks/fleurs\",\"https://api.llm-stats.com/leaderboard/benchmarks/fleurs\"]", + "raw_score_id": "fleurs::gemini-1.0-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "wmt23::wmt23-gemini-1.0-pro", + "evaluation_name": "llm_stats.wmt23", + "source_data": { + "dataset_name": "WMT23", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.0-pro", + "https://llm-stats.com/benchmarks/wmt23", + "https://api.llm-stats.com/leaderboard/benchmarks/wmt23" + ], + "additional_details": { + "raw_benchmark_id": "wmt23", + "raw_model_id": "gemini-1.0-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The Eighth Conference on Machine Translation (WMT23) benchmark evaluating machine translation systems across 8 language pairs (14 translation directions) including general, biomedical, literary, and low-resource language translation tasks. Features specialized shared tasks for quality estimation, metrics evaluation, sign language translation, and discourse-level literary translation with professional human assessment.", + "metric_id": "llm_stats.wmt23.score", + "metric_name": "WMT23 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "wmt23", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WMT23", + "raw_categories": "[\"language\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.717, + "details": { + "raw_score": "0.717", + "raw_score_field": "score", + "raw_model_id": "gemini-1.0-pro", + "raw_benchmark_id": "wmt23", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.0-pro\",\"https://llm-stats.com/benchmarks/wmt23\",\"https://api.llm-stats.com/leaderboard/benchmarks/wmt23\"]", + "raw_score_id": "wmt23::gemini-1.0-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/google/gemini-1.5-flash/eeb415ab-7107-4a8a-b9f0-e01a09fd949f.json b/data/llm_stats/google/gemini-1.5-flash/eeb415ab-7107-4a8a-b9f0-e01a09fd949f.json new file mode 100644 index 0000000000000000000000000000000000000000..62c0855a97834161bf76eff49f9740a8a33ea9a9 --- /dev/null +++ b/data/llm_stats/google/gemini-1.5-flash/eeb415ab-7107-4a8a-b9f0-e01a09fd949f.json @@ -0,0 +1,774 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/google_gemini-1.5-flash/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Gemini 1.5 Flash", + "id": "google/gemini-1.5-flash", + "developer": "google", + "additional_details": { + "raw_id": "gemini-1.5-flash", + "raw_name": "Gemini 1.5 Flash", + "raw_model_id": "gemini-1.5-flash", + "raw_model_name": "Gemini 1.5 Flash", + "raw_organization_id": "google", + "raw_organization_name": "Google", + "raw_release_date": "2024-05-01", + "raw_announcement_date": "2024-05-01", + "raw_multimodal": "true", + "raw_provider_slug": "google", + "raw_provider_name": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "amc-2022-23::amc-2022-23-gemini-1.5-flash", + "evaluation_name": "llm_stats.amc-2022-23", + "source_data": { + "dataset_name": "AMC_2022_23", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.5-flash", + "https://llm-stats.com/benchmarks/amc-2022-23", + "https://api.llm-stats.com/leaderboard/benchmarks/amc-2022-23" + ], + "additional_details": { + "raw_benchmark_id": "amc-2022-23", + "raw_model_id": "gemini-1.5-flash", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "American Mathematics Competition problems from the 2022-23 academic year, consisting of multiple-choice mathematics competition problems designed for high school students. These problems require advanced mathematical reasoning, problem-solving strategies, and mathematical knowledge covering topics like algebra, geometry, number theory, and combinatorics. The benchmark is derived from the official AMC competitions sponsored by the Mathematical Association of America.", + "metric_id": "llm_stats.amc-2022-23.score", + "metric_name": "AMC_2022_23 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "amc-2022-23", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AMC_2022_23", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.348, + "details": { + "raw_score": "0.348", + "raw_score_field": "score", + "raw_model_id": "gemini-1.5-flash", + "raw_benchmark_id": "amc-2022-23", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.5-flash\",\"https://llm-stats.com/benchmarks/amc-2022-23\",\"https://api.llm-stats.com/leaderboard/benchmarks/amc-2022-23\"]", + "raw_score_id": "amc-2022-23::gemini-1.5-flash", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "big-bench-hard::big-bench-hard-gemini-1.5-flash", + "evaluation_name": "llm_stats.big-bench-hard", + "source_data": { + "dataset_name": "BIG-Bench Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.5-flash", + "https://llm-stats.com/benchmarks/big-bench-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard" + ], + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_model_id": "gemini-1.5-flash", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIG-Bench Hard (BBH) is a subset of 23 challenging BIG-Bench tasks selected because prior language model evaluations did not outperform average human-rater performance. The benchmark contains 6,511 evaluation examples testing various forms of multi-step reasoning including arithmetic, logical reasoning (Boolean expressions, logical deduction), geometric reasoning, temporal reasoning, and language understanding. Tasks require capabilities such as causal judgment, object counting, navigation, pattern recognition, and complex problem solving.", + "metric_id": "llm_stats.big-bench-hard.score", + "metric_name": "BIG-Bench Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench Hard", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "21" + } + }, + "score_details": { + "score": 0.855, + "details": { + "raw_score": "0.855", + "raw_score_field": "score", + "raw_model_id": "gemini-1.5-flash", + "raw_benchmark_id": "big-bench-hard", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.5-flash\",\"https://llm-stats.com/benchmarks/big-bench-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard\"]", + "raw_score_id": "big-bench-hard::gemini-1.5-flash", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "fleurs::fleurs-gemini-1.5-flash", + "evaluation_name": "llm_stats.fleurs", + "source_data": { + "dataset_name": "FLEURS", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.5-flash", + "https://llm-stats.com/benchmarks/fleurs", + "https://api.llm-stats.com/leaderboard/benchmarks/fleurs" + ], + "additional_details": { + "raw_benchmark_id": "fleurs", + "raw_model_id": "gemini-1.5-flash", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Few-shot Learning Evaluation of Universal Representations of Speech - a parallel speech dataset in 102 languages built on FLoRes-101 with approximately 12 hours of speech supervision per language for tasks including ASR, speech language identification, translation and retrieval. Scores are shown as speech recognition accuracy (1 - word error rate), so higher is better.", + "metric_id": "llm_stats.fleurs.score", + "metric_name": "FLEURS score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "fleurs", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FLEURS", + "raw_categories": "[\"language\",\"speech_to_text\"]", + "raw_modality": "audio", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.904, + "details": { + "raw_score": "0.904", + "raw_score_field": "score", + "raw_model_id": "gemini-1.5-flash", + "raw_benchmark_id": "fleurs", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.5-flash\",\"https://llm-stats.com/benchmarks/fleurs\",\"https://api.llm-stats.com/leaderboard/benchmarks/fleurs\"]", + "raw_score_id": "fleurs::gemini-1.5-flash", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "functionalmath::functionalmath-gemini-1.5-flash", + "evaluation_name": "llm_stats.functionalmath", + "source_data": { + "dataset_name": "FunctionalMATH", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.5-flash", + "https://llm-stats.com/benchmarks/functionalmath", + "https://api.llm-stats.com/leaderboard/benchmarks/functionalmath" + ], + "additional_details": { + "raw_benchmark_id": "functionalmath", + "raw_model_id": "gemini-1.5-flash", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A functional variant of the MATH benchmark that tests language models' ability to generalize reasoning patterns across different problem instances, revealing the reasoning gap between static and functional performance.", + "metric_id": "llm_stats.functionalmath.score", + "metric_name": "FunctionalMATH score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "functionalmath", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FunctionalMATH", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.536, + "details": { + "raw_score": "0.536", + "raw_score_field": "score", + "raw_model_id": "gemini-1.5-flash", + "raw_benchmark_id": "functionalmath", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.5-flash\",\"https://llm-stats.com/benchmarks/functionalmath\",\"https://api.llm-stats.com/leaderboard/benchmarks/functionalmath\"]", + "raw_score_id": "functionalmath::gemini-1.5-flash", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hellaswag::hellaswag-gemini-1.5-flash", + "evaluation_name": "llm_stats.hellaswag", + "source_data": { + "dataset_name": "HellaSwag", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.5-flash", + "https://llm-stats.com/benchmarks/hellaswag", + "https://api.llm-stats.com/leaderboard/benchmarks/hellaswag" + ], + "additional_details": { + "raw_benchmark_id": "hellaswag", + "raw_model_id": "gemini-1.5-flash", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A challenging commonsense natural language inference dataset that uses Adversarial Filtering to create questions trivial for humans (>95% accuracy) but difficult for state-of-the-art models, requiring completion of sentence endings based on physical situations and everyday activities", + "metric_id": "llm_stats.hellaswag.score", + "metric_name": "HellaSwag score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hellaswag", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HellaSwag", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.865, + "details": { + "raw_score": "0.865", + "raw_score_field": "score", + "raw_model_id": "gemini-1.5-flash", + "raw_benchmark_id": "hellaswag", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.5-flash\",\"https://llm-stats.com/benchmarks/hellaswag\",\"https://api.llm-stats.com/leaderboard/benchmarks/hellaswag\"]", + "raw_score_id": "hellaswag::gemini-1.5-flash", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hiddenmath::hiddenmath-gemini-1.5-flash", + "evaluation_name": "llm_stats.hiddenmath", + "source_data": { + "dataset_name": "HiddenMath", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.5-flash", + "https://llm-stats.com/benchmarks/hiddenmath", + "https://api.llm-stats.com/leaderboard/benchmarks/hiddenmath" + ], + "additional_details": { + "raw_benchmark_id": "hiddenmath", + "raw_model_id": "gemini-1.5-flash", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Google DeepMind's internal mathematical reasoning benchmark that introduces novel problems not encountered during model training to evaluate true mathematical reasoning capabilities rather than memorization", + "metric_id": "llm_stats.hiddenmath.score", + "metric_name": "HiddenMath score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hiddenmath", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HiddenMath", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.472, + "details": { + "raw_score": "0.472", + "raw_score_field": "score", + "raw_model_id": "gemini-1.5-flash", + "raw_benchmark_id": "hiddenmath", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.5-flash\",\"https://llm-stats.com/benchmarks/hiddenmath\",\"https://api.llm-stats.com/leaderboard/benchmarks/hiddenmath\"]", + "raw_score_id": "hiddenmath::gemini-1.5-flash", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mgsm::mgsm-gemini-1.5-flash", + "evaluation_name": "llm_stats.mgsm", + "source_data": { + "dataset_name": "MGSM", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.5-flash", + "https://llm-stats.com/benchmarks/mgsm", + "https://api.llm-stats.com/leaderboard/benchmarks/mgsm" + ], + "additional_details": { + "raw_benchmark_id": "mgsm", + "raw_model_id": "gemini-1.5-flash", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MGSM (Multilingual Grade School Math) is a benchmark of grade-school math problems. Contains 250 grade-school math problems manually translated from the GSM8K dataset into ten typologically diverse languages: Spanish, French, German, Russian, Chinese, Japanese, Thai, Swahili, Bengali, and Telugu. Evaluates multilingual mathematical reasoning capabilities.", + "metric_id": "llm_stats.mgsm.score", + "metric_name": "MGSM score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mgsm", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MGSM", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "31" + } + }, + "score_details": { + "score": 0.826, + "details": { + "raw_score": "0.826", + "raw_score_field": "score", + "raw_model_id": "gemini-1.5-flash", + "raw_benchmark_id": "mgsm", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.5-flash\",\"https://llm-stats.com/benchmarks/mgsm\",\"https://api.llm-stats.com/leaderboard/benchmarks/mgsm\"]", + "raw_score_id": "mgsm::gemini-1.5-flash", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mrcr::mrcr-gemini-1.5-flash", + "evaluation_name": "llm_stats.mrcr", + "source_data": { + "dataset_name": "MRCR", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.5-flash", + "https://llm-stats.com/benchmarks/mrcr", + "https://api.llm-stats.com/leaderboard/benchmarks/mrcr" + ], + "additional_details": { + "raw_benchmark_id": "mrcr", + "raw_model_id": "gemini-1.5-flash", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MRCR (Multi-Round Coreference Resolution) is a synthetic long-context reasoning task where models must navigate long conversations to reproduce specific model outputs. It tests the ability to distinguish between similar requests and reason about ordering while maintaining attention across extended contexts.", + "metric_id": "llm_stats.mrcr.score", + "metric_name": "MRCR score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mrcr", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MRCR", + "raw_categories": "[\"long_context\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.719, + "details": { + "raw_score": "0.719", + "raw_score_field": "score", + "raw_model_id": "gemini-1.5-flash", + "raw_benchmark_id": "mrcr", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.5-flash\",\"https://llm-stats.com/benchmarks/mrcr\",\"https://api.llm-stats.com/leaderboard/benchmarks/mrcr\"]", + "raw_score_id": "mrcr::gemini-1.5-flash", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "natural2code::natural2code-gemini-1.5-flash", + "evaluation_name": "llm_stats.natural2code", + "source_data": { + "dataset_name": "Natural2Code", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.5-flash", + "https://llm-stats.com/benchmarks/natural2code", + "https://api.llm-stats.com/leaderboard/benchmarks/natural2code" + ], + "additional_details": { + "raw_benchmark_id": "natural2code", + "raw_model_id": "gemini-1.5-flash", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "NaturalCodeBench (NCB) is a challenging code benchmark designed to mirror the complexity and variety of real-world coding tasks. It comprises 402 high-quality problems in Python and Java, selected from natural user queries from online coding services, covering 6 different domains.", + "metric_id": "llm_stats.natural2code.score", + "metric_name": "Natural2Code score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "natural2code", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Natural2Code", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.798, + "details": { + "raw_score": "0.798", + "raw_score_field": "score", + "raw_model_id": "gemini-1.5-flash", + "raw_benchmark_id": "natural2code", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.5-flash\",\"https://llm-stats.com/benchmarks/natural2code\",\"https://api.llm-stats.com/leaderboard/benchmarks/natural2code\"]", + "raw_score_id": "natural2code::gemini-1.5-flash", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "physicsfinals::physicsfinals-gemini-1.5-flash", + "evaluation_name": "llm_stats.physicsfinals", + "source_data": { + "dataset_name": "PhysicsFinals", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.5-flash", + "https://llm-stats.com/benchmarks/physicsfinals", + "https://api.llm-stats.com/leaderboard/benchmarks/physicsfinals" + ], + "additional_details": { + "raw_benchmark_id": "physicsfinals", + "raw_model_id": "gemini-1.5-flash", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "PHYSICS is a comprehensive benchmark for university-level physics problem solving, containing 1,297 expert-annotated problems covering six core areas: classical mechanics, quantum mechanics, thermodynamics and statistical mechanics, electromagnetism, atomic physics, and optics. Each problem requires advanced physics knowledge and mathematical reasoning. Even advanced models like o3-mini achieve only 59.9% accuracy.", + "metric_id": "llm_stats.physicsfinals.score", + "metric_name": "PhysicsFinals score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "physicsfinals", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PhysicsFinals", + "raw_categories": "[\"math\",\"physics\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.574, + "details": { + "raw_score": "0.574", + "raw_score_field": "score", + "raw_model_id": "gemini-1.5-flash", + "raw_benchmark_id": "physicsfinals", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.5-flash\",\"https://llm-stats.com/benchmarks/physicsfinals\",\"https://api.llm-stats.com/leaderboard/benchmarks/physicsfinals\"]", + "raw_score_id": "physicsfinals::gemini-1.5-flash", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "vibe-eval::vibe-eval-gemini-1.5-flash", + "evaluation_name": "llm_stats.vibe-eval", + "source_data": { + "dataset_name": "Vibe-Eval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.5-flash", + "https://llm-stats.com/benchmarks/vibe-eval", + "https://api.llm-stats.com/leaderboard/benchmarks/vibe-eval" + ], + "additional_details": { + "raw_benchmark_id": "vibe-eval", + "raw_model_id": "gemini-1.5-flash", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VIBE-Eval is a hard evaluation suite for measuring progress of multimodal language models, consisting of 269 visual understanding prompts with gold-standard responses authored by experts. The benchmark has dual objectives: vibe checking multimodal chat models for day-to-day tasks and rigorously testing frontier models, with the hard set containing >50% questions that all frontier models answer incorrectly.", + "metric_id": "llm_stats.vibe-eval.score", + "metric_name": "Vibe-Eval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "vibe-eval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Vibe-Eval", + "raw_categories": "[\"multimodal\",\"general\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.489, + "details": { + "raw_score": "0.489", + "raw_score_field": "score", + "raw_model_id": "gemini-1.5-flash", + "raw_benchmark_id": "vibe-eval", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.5-flash\",\"https://llm-stats.com/benchmarks/vibe-eval\",\"https://api.llm-stats.com/leaderboard/benchmarks/vibe-eval\"]", + "raw_score_id": "vibe-eval::gemini-1.5-flash", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "video-mme::video-mme-gemini-1.5-flash", + "evaluation_name": "llm_stats.video-mme", + "source_data": { + "dataset_name": "Video-MME", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.5-flash", + "https://llm-stats.com/benchmarks/video-mme", + "https://api.llm-stats.com/leaderboard/benchmarks/video-mme" + ], + "additional_details": { + "raw_benchmark_id": "video-mme", + "raw_model_id": "gemini-1.5-flash", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Video-MME is the first-ever comprehensive evaluation benchmark of Multi-modal Large Language Models (MLLMs) in video analysis. It features 900 videos totaling 254 hours with 2,700 human-annotated question-answer pairs across 6 primary visual domains (Knowledge, Film & Television, Sports Competition, Life Record, Multilingual, and others) and 30 subfields. The benchmark evaluates models across diverse temporal dimensions (11 seconds to 1 hour), integrates multi-modal inputs including video frames, subtitles, and audio, and uses rigorous manual labeling by expert annotators for precise assessment.", + "metric_id": "llm_stats.video-mme.score", + "metric_name": "Video-MME score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "video-mme", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Video-MME", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "17" + } + }, + "score_details": { + "score": 0.761, + "details": { + "raw_score": "0.761", + "raw_score_field": "score", + "raw_model_id": "gemini-1.5-flash", + "raw_benchmark_id": "video-mme", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.5-flash\",\"https://llm-stats.com/benchmarks/video-mme\",\"https://api.llm-stats.com/leaderboard/benchmarks/video-mme\"]", + "raw_score_id": "video-mme::gemini-1.5-flash", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "wmt23::wmt23-gemini-1.5-flash", + "evaluation_name": "llm_stats.wmt23", + "source_data": { + "dataset_name": "WMT23", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.5-flash", + "https://llm-stats.com/benchmarks/wmt23", + "https://api.llm-stats.com/leaderboard/benchmarks/wmt23" + ], + "additional_details": { + "raw_benchmark_id": "wmt23", + "raw_model_id": "gemini-1.5-flash", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The Eighth Conference on Machine Translation (WMT23) benchmark evaluating machine translation systems across 8 language pairs (14 translation directions) including general, biomedical, literary, and low-resource language translation tasks. Features specialized shared tasks for quality estimation, metrics evaluation, sign language translation, and discourse-level literary translation with professional human assessment.", + "metric_id": "llm_stats.wmt23.score", + "metric_name": "WMT23 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "wmt23", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WMT23", + "raw_categories": "[\"language\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.741, + "details": { + "raw_score": "0.741", + "raw_score_field": "score", + "raw_model_id": "gemini-1.5-flash", + "raw_benchmark_id": "wmt23", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.5-flash\",\"https://llm-stats.com/benchmarks/wmt23\",\"https://api.llm-stats.com/leaderboard/benchmarks/wmt23\"]", + "raw_score_id": "wmt23::gemini-1.5-flash", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "xstest::xstest-gemini-1.5-flash", + "evaluation_name": "llm_stats.xstest", + "source_data": { + "dataset_name": "XSTest", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.5-flash", + "https://llm-stats.com/benchmarks/xstest", + "https://api.llm-stats.com/leaderboard/benchmarks/xstest" + ], + "additional_details": { + "raw_benchmark_id": "xstest", + "raw_model_id": "gemini-1.5-flash", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "XSTest is a test suite designed to identify exaggerated safety behaviours in large language models. It comprises 450 prompts: 250 safe prompts across ten prompt types that well-calibrated models should not refuse to comply with, and 200 unsafe prompts as contrasts that models should refuse. The benchmark systematically evaluates whether models refuse to respond to clearly safe prompts due to overly cautious safety mechanisms.", + "metric_id": "llm_stats.xstest.score", + "metric_name": "XSTest score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "xstest", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "XSTest", + "raw_categories": "[\"safety\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.97, + "details": { + "raw_score": "0.97", + "raw_score_field": "score", + "raw_model_id": "gemini-1.5-flash", + "raw_benchmark_id": "xstest", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.5-flash\",\"https://llm-stats.com/benchmarks/xstest\",\"https://api.llm-stats.com/leaderboard/benchmarks/xstest\"]", + "raw_score_id": "xstest::gemini-1.5-flash", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/google/gemini-1.5-pro/108f9698-a2cf-4d0c-a186-30214863e73f.json b/data/llm_stats/google/gemini-1.5-pro/108f9698-a2cf-4d0c-a186-30214863e73f.json new file mode 100644 index 0000000000000000000000000000000000000000..421feea5647e7ffc96311fc1636f8e677d67be5e --- /dev/null +++ b/data/llm_stats/google/gemini-1.5-pro/108f9698-a2cf-4d0c-a186-30214863e73f.json @@ -0,0 +1,930 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/google_gemini-1.5-pro/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Gemini 1.5 Pro", + "id": "google/gemini-1.5-pro", + "developer": "google", + "additional_details": { + "raw_id": "gemini-1.5-pro", + "raw_name": "Gemini 1.5 Pro", + "raw_model_id": "gemini-1.5-pro", + "raw_model_name": "Gemini 1.5 Pro", + "raw_organization_id": "google", + "raw_organization_name": "Google", + "raw_release_date": "2024-05-01", + "raw_announcement_date": "2024-05-01", + "raw_multimodal": "true", + "raw_provider_slug": "google", + "raw_provider_name": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "amc-2022-23::amc-2022-23-gemini-1.5-pro", + "evaluation_name": "llm_stats.amc-2022-23", + "source_data": { + "dataset_name": "AMC_2022_23", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.5-pro", + "https://llm-stats.com/benchmarks/amc-2022-23", + "https://api.llm-stats.com/leaderboard/benchmarks/amc-2022-23" + ], + "additional_details": { + "raw_benchmark_id": "amc-2022-23", + "raw_model_id": "gemini-1.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "American Mathematics Competition problems from the 2022-23 academic year, consisting of multiple-choice mathematics competition problems designed for high school students. These problems require advanced mathematical reasoning, problem-solving strategies, and mathematical knowledge covering topics like algebra, geometry, number theory, and combinatorics. The benchmark is derived from the official AMC competitions sponsored by the Mathematical Association of America.", + "metric_id": "llm_stats.amc-2022-23.score", + "metric_name": "AMC_2022_23 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "amc-2022-23", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AMC_2022_23", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.464, + "details": { + "raw_score": "0.464", + "raw_score_field": "score", + "raw_model_id": "gemini-1.5-pro", + "raw_benchmark_id": "amc-2022-23", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.5-pro\",\"https://llm-stats.com/benchmarks/amc-2022-23\",\"https://api.llm-stats.com/leaderboard/benchmarks/amc-2022-23\"]", + "raw_score_id": "amc-2022-23::gemini-1.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "big-bench-hard::big-bench-hard-gemini-1.5-pro", + "evaluation_name": "llm_stats.big-bench-hard", + "source_data": { + "dataset_name": "BIG-Bench Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.5-pro", + "https://llm-stats.com/benchmarks/big-bench-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard" + ], + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_model_id": "gemini-1.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIG-Bench Hard (BBH) is a subset of 23 challenging BIG-Bench tasks selected because prior language model evaluations did not outperform average human-rater performance. The benchmark contains 6,511 evaluation examples testing various forms of multi-step reasoning including arithmetic, logical reasoning (Boolean expressions, logical deduction), geometric reasoning, temporal reasoning, and language understanding. Tasks require capabilities such as causal judgment, object counting, navigation, pattern recognition, and complex problem solving.", + "metric_id": "llm_stats.big-bench-hard.score", + "metric_name": "BIG-Bench Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench Hard", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "21" + } + }, + "score_details": { + "score": 0.892, + "details": { + "raw_score": "0.892", + "raw_score_field": "score", + "raw_model_id": "gemini-1.5-pro", + "raw_benchmark_id": "big-bench-hard", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.5-pro\",\"https://llm-stats.com/benchmarks/big-bench-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard\"]", + "raw_score_id": "big-bench-hard::gemini-1.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "drop::drop-gemini-1.5-pro", + "evaluation_name": "llm_stats.drop", + "source_data": { + "dataset_name": "DROP", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.5-pro", + "https://llm-stats.com/benchmarks/drop", + "https://api.llm-stats.com/leaderboard/benchmarks/drop" + ], + "additional_details": { + "raw_benchmark_id": "drop", + "raw_model_id": "gemini-1.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "DROP (Discrete Reasoning Over Paragraphs) is a reading comprehension benchmark requiring discrete reasoning over paragraph content. It contains crowdsourced, adversarially-created questions that require resolving references and performing discrete operations like addition, counting, or sorting, demanding comprehensive paragraph understanding beyond paraphrase-and-entity-typing shortcuts.", + "metric_id": "llm_stats.drop.score", + "metric_name": "DROP score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "drop", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DROP", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "30" + } + }, + "score_details": { + "score": 0.749, + "details": { + "raw_score": "0.749", + "raw_score_field": "score", + "raw_model_id": "gemini-1.5-pro", + "raw_benchmark_id": "drop", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.5-pro\",\"https://llm-stats.com/benchmarks/drop\",\"https://api.llm-stats.com/leaderboard/benchmarks/drop\"]", + "raw_score_id": "drop::gemini-1.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "fleurs::fleurs-gemini-1.5-pro", + "evaluation_name": "llm_stats.fleurs", + "source_data": { + "dataset_name": "FLEURS", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.5-pro", + "https://llm-stats.com/benchmarks/fleurs", + "https://api.llm-stats.com/leaderboard/benchmarks/fleurs" + ], + "additional_details": { + "raw_benchmark_id": "fleurs", + "raw_model_id": "gemini-1.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Few-shot Learning Evaluation of Universal Representations of Speech - a parallel speech dataset in 102 languages built on FLoRes-101 with approximately 12 hours of speech supervision per language for tasks including ASR, speech language identification, translation and retrieval. Scores are shown as speech recognition accuracy (1 - word error rate), so higher is better.", + "metric_id": "llm_stats.fleurs.score", + "metric_name": "FLEURS score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "fleurs", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FLEURS", + "raw_categories": "[\"language\",\"speech_to_text\"]", + "raw_modality": "audio", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.933, + "details": { + "raw_score": "0.933", + "raw_score_field": "score", + "raw_model_id": "gemini-1.5-pro", + "raw_benchmark_id": "fleurs", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.5-pro\",\"https://llm-stats.com/benchmarks/fleurs\",\"https://api.llm-stats.com/leaderboard/benchmarks/fleurs\"]", + "raw_score_id": "fleurs::gemini-1.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "functionalmath::functionalmath-gemini-1.5-pro", + "evaluation_name": "llm_stats.functionalmath", + "source_data": { + "dataset_name": "FunctionalMATH", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.5-pro", + "https://llm-stats.com/benchmarks/functionalmath", + "https://api.llm-stats.com/leaderboard/benchmarks/functionalmath" + ], + "additional_details": { + "raw_benchmark_id": "functionalmath", + "raw_model_id": "gemini-1.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A functional variant of the MATH benchmark that tests language models' ability to generalize reasoning patterns across different problem instances, revealing the reasoning gap between static and functional performance.", + "metric_id": "llm_stats.functionalmath.score", + "metric_name": "FunctionalMATH score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "functionalmath", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FunctionalMATH", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.646, + "details": { + "raw_score": "0.646", + "raw_score_field": "score", + "raw_model_id": "gemini-1.5-pro", + "raw_benchmark_id": "functionalmath", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.5-pro\",\"https://llm-stats.com/benchmarks/functionalmath\",\"https://api.llm-stats.com/leaderboard/benchmarks/functionalmath\"]", + "raw_score_id": "functionalmath::gemini-1.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hellaswag::hellaswag-gemini-1.5-pro", + "evaluation_name": "llm_stats.hellaswag", + "source_data": { + "dataset_name": "HellaSwag", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.5-pro", + "https://llm-stats.com/benchmarks/hellaswag", + "https://api.llm-stats.com/leaderboard/benchmarks/hellaswag" + ], + "additional_details": { + "raw_benchmark_id": "hellaswag", + "raw_model_id": "gemini-1.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A challenging commonsense natural language inference dataset that uses Adversarial Filtering to create questions trivial for humans (>95% accuracy) but difficult for state-of-the-art models, requiring completion of sentence endings based on physical situations and everyday activities", + "metric_id": "llm_stats.hellaswag.score", + "metric_name": "HellaSwag score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hellaswag", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HellaSwag", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.933, + "details": { + "raw_score": "0.933", + "raw_score_field": "score", + "raw_model_id": "gemini-1.5-pro", + "raw_benchmark_id": "hellaswag", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.5-pro\",\"https://llm-stats.com/benchmarks/hellaswag\",\"https://api.llm-stats.com/leaderboard/benchmarks/hellaswag\"]", + "raw_score_id": "hellaswag::gemini-1.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hiddenmath::hiddenmath-gemini-1.5-pro", + "evaluation_name": "llm_stats.hiddenmath", + "source_data": { + "dataset_name": "HiddenMath", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.5-pro", + "https://llm-stats.com/benchmarks/hiddenmath", + "https://api.llm-stats.com/leaderboard/benchmarks/hiddenmath" + ], + "additional_details": { + "raw_benchmark_id": "hiddenmath", + "raw_model_id": "gemini-1.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Google DeepMind's internal mathematical reasoning benchmark that introduces novel problems not encountered during model training to evaluate true mathematical reasoning capabilities rather than memorization", + "metric_id": "llm_stats.hiddenmath.score", + "metric_name": "HiddenMath score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hiddenmath", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HiddenMath", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.52, + "details": { + "raw_score": "0.52", + "raw_score_field": "score", + "raw_model_id": "gemini-1.5-pro", + "raw_benchmark_id": "hiddenmath", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.5-pro\",\"https://llm-stats.com/benchmarks/hiddenmath\",\"https://api.llm-stats.com/leaderboard/benchmarks/hiddenmath\"]", + "raw_score_id": "hiddenmath::gemini-1.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "math::math-gemini-1.5-pro", + "evaluation_name": "llm_stats.math", + "source_data": { + "dataset_name": "MATH", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.5-pro", + "https://llm-stats.com/benchmarks/math", + "https://api.llm-stats.com/leaderboard/benchmarks/math" + ], + "additional_details": { + "raw_benchmark_id": "math", + "raw_model_id": "gemini-1.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MATH dataset contains 12,500 challenging competition mathematics problems from AMC 10, AMC 12, AIME, and other mathematics competitions. Each problem includes full step-by-step solutions and spans multiple difficulty levels (1-5) across seven mathematical subjects including Prealgebra, Algebra, Number Theory, Counting and Probability, Geometry, Intermediate Algebra, and Precalculus.", + "metric_id": "llm_stats.math.score", + "metric_name": "MATH score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "math", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MATH", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "71" + } + }, + "score_details": { + "score": 0.865, + "details": { + "raw_score": "0.865", + "raw_score_field": "score", + "raw_model_id": "gemini-1.5-pro", + "raw_benchmark_id": "math", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.5-pro\",\"https://llm-stats.com/benchmarks/math\",\"https://api.llm-stats.com/leaderboard/benchmarks/math\"]", + "raw_score_id": "math::gemini-1.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mathvista::mathvista-gemini-1.5-pro", + "evaluation_name": "llm_stats.mathvista", + "source_data": { + "dataset_name": "MathVista", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.5-pro", + "https://llm-stats.com/benchmarks/mathvista", + "https://api.llm-stats.com/leaderboard/benchmarks/mathvista" + ], + "additional_details": { + "raw_benchmark_id": "mathvista", + "raw_model_id": "gemini-1.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MathVista evaluates mathematical reasoning of foundation models in visual contexts. It consists of 6,141 examples derived from 28 existing multimodal datasets and 3 newly created datasets (IQTest, FunctionQA, and PaperQA), combining challenges from diverse mathematical and visual tasks to assess models' ability to understand complex figures and perform rigorous reasoning.", + "metric_id": "llm_stats.mathvista.score", + "metric_name": "MathVista score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mathvista", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MathVista", + "raw_categories": "[\"math\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "39" + } + }, + "score_details": { + "score": 0.681, + "details": { + "raw_score": "0.681", + "raw_score_field": "score", + "raw_model_id": "gemini-1.5-pro", + "raw_benchmark_id": "mathvista", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.5-pro\",\"https://llm-stats.com/benchmarks/mathvista\",\"https://api.llm-stats.com/leaderboard/benchmarks/mathvista\"]", + "raw_score_id": "mathvista::gemini-1.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mgsm::mgsm-gemini-1.5-pro", + "evaluation_name": "llm_stats.mgsm", + "source_data": { + "dataset_name": "MGSM", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.5-pro", + "https://llm-stats.com/benchmarks/mgsm", + "https://api.llm-stats.com/leaderboard/benchmarks/mgsm" + ], + "additional_details": { + "raw_benchmark_id": "mgsm", + "raw_model_id": "gemini-1.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MGSM (Multilingual Grade School Math) is a benchmark of grade-school math problems. Contains 250 grade-school math problems manually translated from the GSM8K dataset into ten typologically diverse languages: Spanish, French, German, Russian, Chinese, Japanese, Thai, Swahili, Bengali, and Telugu. Evaluates multilingual mathematical reasoning capabilities.", + "metric_id": "llm_stats.mgsm.score", + "metric_name": "MGSM score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mgsm", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MGSM", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "31" + } + }, + "score_details": { + "score": 0.875, + "details": { + "raw_score": "0.875", + "raw_score_field": "score", + "raw_model_id": "gemini-1.5-pro", + "raw_benchmark_id": "mgsm", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.5-pro\",\"https://llm-stats.com/benchmarks/mgsm\",\"https://api.llm-stats.com/leaderboard/benchmarks/mgsm\"]", + "raw_score_id": "mgsm::gemini-1.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mrcr::mrcr-gemini-1.5-pro", + "evaluation_name": "llm_stats.mrcr", + "source_data": { + "dataset_name": "MRCR", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.5-pro", + "https://llm-stats.com/benchmarks/mrcr", + "https://api.llm-stats.com/leaderboard/benchmarks/mrcr" + ], + "additional_details": { + "raw_benchmark_id": "mrcr", + "raw_model_id": "gemini-1.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MRCR (Multi-Round Coreference Resolution) is a synthetic long-context reasoning task where models must navigate long conversations to reproduce specific model outputs. It tests the ability to distinguish between similar requests and reason about ordering while maintaining attention across extended contexts.", + "metric_id": "llm_stats.mrcr.score", + "metric_name": "MRCR score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mrcr", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MRCR", + "raw_categories": "[\"long_context\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.826, + "details": { + "raw_score": "0.826", + "raw_score_field": "score", + "raw_model_id": "gemini-1.5-pro", + "raw_benchmark_id": "mrcr", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.5-pro\",\"https://llm-stats.com/benchmarks/mrcr\",\"https://api.llm-stats.com/leaderboard/benchmarks/mrcr\"]", + "raw_score_id": "mrcr::gemini-1.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "natural2code::natural2code-gemini-1.5-pro", + "evaluation_name": "llm_stats.natural2code", + "source_data": { + "dataset_name": "Natural2Code", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.5-pro", + "https://llm-stats.com/benchmarks/natural2code", + "https://api.llm-stats.com/leaderboard/benchmarks/natural2code" + ], + "additional_details": { + "raw_benchmark_id": "natural2code", + "raw_model_id": "gemini-1.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "NaturalCodeBench (NCB) is a challenging code benchmark designed to mirror the complexity and variety of real-world coding tasks. It comprises 402 high-quality problems in Python and Java, selected from natural user queries from online coding services, covering 6 different domains.", + "metric_id": "llm_stats.natural2code.score", + "metric_name": "Natural2Code score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "natural2code", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Natural2Code", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.854, + "details": { + "raw_score": "0.854", + "raw_score_field": "score", + "raw_model_id": "gemini-1.5-pro", + "raw_benchmark_id": "natural2code", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.5-pro\",\"https://llm-stats.com/benchmarks/natural2code\",\"https://api.llm-stats.com/leaderboard/benchmarks/natural2code\"]", + "raw_score_id": "natural2code::gemini-1.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "physicsfinals::physicsfinals-gemini-1.5-pro", + "evaluation_name": "llm_stats.physicsfinals", + "source_data": { + "dataset_name": "PhysicsFinals", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.5-pro", + "https://llm-stats.com/benchmarks/physicsfinals", + "https://api.llm-stats.com/leaderboard/benchmarks/physicsfinals" + ], + "additional_details": { + "raw_benchmark_id": "physicsfinals", + "raw_model_id": "gemini-1.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "PHYSICS is a comprehensive benchmark for university-level physics problem solving, containing 1,297 expert-annotated problems covering six core areas: classical mechanics, quantum mechanics, thermodynamics and statistical mechanics, electromagnetism, atomic physics, and optics. Each problem requires advanced physics knowledge and mathematical reasoning. Even advanced models like o3-mini achieve only 59.9% accuracy.", + "metric_id": "llm_stats.physicsfinals.score", + "metric_name": "PhysicsFinals score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "physicsfinals", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PhysicsFinals", + "raw_categories": "[\"math\",\"physics\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.639, + "details": { + "raw_score": "0.639", + "raw_score_field": "score", + "raw_model_id": "gemini-1.5-pro", + "raw_benchmark_id": "physicsfinals", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.5-pro\",\"https://llm-stats.com/benchmarks/physicsfinals\",\"https://api.llm-stats.com/leaderboard/benchmarks/physicsfinals\"]", + "raw_score_id": "physicsfinals::gemini-1.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "vibe-eval::vibe-eval-gemini-1.5-pro", + "evaluation_name": "llm_stats.vibe-eval", + "source_data": { + "dataset_name": "Vibe-Eval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.5-pro", + "https://llm-stats.com/benchmarks/vibe-eval", + "https://api.llm-stats.com/leaderboard/benchmarks/vibe-eval" + ], + "additional_details": { + "raw_benchmark_id": "vibe-eval", + "raw_model_id": "gemini-1.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VIBE-Eval is a hard evaluation suite for measuring progress of multimodal language models, consisting of 269 visual understanding prompts with gold-standard responses authored by experts. The benchmark has dual objectives: vibe checking multimodal chat models for day-to-day tasks and rigorously testing frontier models, with the hard set containing >50% questions that all frontier models answer incorrectly.", + "metric_id": "llm_stats.vibe-eval.score", + "metric_name": "Vibe-Eval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "vibe-eval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Vibe-Eval", + "raw_categories": "[\"multimodal\",\"general\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.539, + "details": { + "raw_score": "0.539", + "raw_score_field": "score", + "raw_model_id": "gemini-1.5-pro", + "raw_benchmark_id": "vibe-eval", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.5-pro\",\"https://llm-stats.com/benchmarks/vibe-eval\",\"https://api.llm-stats.com/leaderboard/benchmarks/vibe-eval\"]", + "raw_score_id": "vibe-eval::gemini-1.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "video-mme::video-mme-gemini-1.5-pro", + "evaluation_name": "llm_stats.video-mme", + "source_data": { + "dataset_name": "Video-MME", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.5-pro", + "https://llm-stats.com/benchmarks/video-mme", + "https://api.llm-stats.com/leaderboard/benchmarks/video-mme" + ], + "additional_details": { + "raw_benchmark_id": "video-mme", + "raw_model_id": "gemini-1.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Video-MME is the first-ever comprehensive evaluation benchmark of Multi-modal Large Language Models (MLLMs) in video analysis. It features 900 videos totaling 254 hours with 2,700 human-annotated question-answer pairs across 6 primary visual domains (Knowledge, Film & Television, Sports Competition, Life Record, Multilingual, and others) and 30 subfields. The benchmark evaluates models across diverse temporal dimensions (11 seconds to 1 hour), integrates multi-modal inputs including video frames, subtitles, and audio, and uses rigorous manual labeling by expert annotators for precise assessment.", + "metric_id": "llm_stats.video-mme.score", + "metric_name": "Video-MME score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "video-mme", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Video-MME", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "17" + } + }, + "score_details": { + "score": 0.786, + "details": { + "raw_score": "0.786", + "raw_score_field": "score", + "raw_model_id": "gemini-1.5-pro", + "raw_benchmark_id": "video-mme", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.5-pro\",\"https://llm-stats.com/benchmarks/video-mme\",\"https://api.llm-stats.com/leaderboard/benchmarks/video-mme\"]", + "raw_score_id": "video-mme::gemini-1.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "wmt23::wmt23-gemini-1.5-pro", + "evaluation_name": "llm_stats.wmt23", + "source_data": { + "dataset_name": "WMT23", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.5-pro", + "https://llm-stats.com/benchmarks/wmt23", + "https://api.llm-stats.com/leaderboard/benchmarks/wmt23" + ], + "additional_details": { + "raw_benchmark_id": "wmt23", + "raw_model_id": "gemini-1.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The Eighth Conference on Machine Translation (WMT23) benchmark evaluating machine translation systems across 8 language pairs (14 translation directions) including general, biomedical, literary, and low-resource language translation tasks. Features specialized shared tasks for quality estimation, metrics evaluation, sign language translation, and discourse-level literary translation with professional human assessment.", + "metric_id": "llm_stats.wmt23.score", + "metric_name": "WMT23 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "wmt23", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WMT23", + "raw_categories": "[\"language\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.751, + "details": { + "raw_score": "0.751", + "raw_score_field": "score", + "raw_model_id": "gemini-1.5-pro", + "raw_benchmark_id": "wmt23", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.5-pro\",\"https://llm-stats.com/benchmarks/wmt23\",\"https://api.llm-stats.com/leaderboard/benchmarks/wmt23\"]", + "raw_score_id": "wmt23::gemini-1.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "xstest::xstest-gemini-1.5-pro", + "evaluation_name": "llm_stats.xstest", + "source_data": { + "dataset_name": "XSTest", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-1.5-pro", + "https://llm-stats.com/benchmarks/xstest", + "https://api.llm-stats.com/leaderboard/benchmarks/xstest" + ], + "additional_details": { + "raw_benchmark_id": "xstest", + "raw_model_id": "gemini-1.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "XSTest is a test suite designed to identify exaggerated safety behaviours in large language models. It comprises 450 prompts: 250 safe prompts across ten prompt types that well-calibrated models should not refuse to comply with, and 200 unsafe prompts as contrasts that models should refuse. The benchmark systematically evaluates whether models refuse to respond to clearly safe prompts due to overly cautious safety mechanisms.", + "metric_id": "llm_stats.xstest.score", + "metric_name": "XSTest score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "xstest", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "XSTest", + "raw_categories": "[\"safety\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.988, + "details": { + "raw_score": "0.988", + "raw_score_field": "score", + "raw_model_id": "gemini-1.5-pro", + "raw_benchmark_id": "xstest", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-1.5-pro\",\"https://llm-stats.com/benchmarks/xstest\",\"https://api.llm-stats.com/leaderboard/benchmarks/xstest\"]", + "raw_score_id": "xstest::gemini-1.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/google/gemini-3.1-pro-preview/6b9beadd-a44e-402d-9964-5c7e198f297e.json b/data/llm_stats/google/gemini-3.1-pro-preview/6b9beadd-a44e-402d-9964-5c7e198f297e.json new file mode 100644 index 0000000000000000000000000000000000000000..4a7ae5c47ec3aa192b778db7acbf07dd970b049c --- /dev/null +++ b/data/llm_stats/google/gemini-3.1-pro-preview/6b9beadd-a44e-402d-9964-5c7e198f297e.json @@ -0,0 +1,776 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/google_gemini-3.1-pro-preview/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-pro-preview", + "id": "google/gemini-3.1-pro-preview", + "developer": "google", + "additional_details": { + "raw_id": "gemini-3.1-pro-preview", + "raw_model_name": "gemini-3.1-pro-preview", + "raw_display_name": "Gemini 3.1 Pro", + "raw_organization_id": "google", + "raw_organization_name": "Google", + "raw_context_length": "1048576", + "raw_input_modalities": "[\"image\",\"text\",\"video\"]", + "raw_output_modalities": "[\"text\"]", + "raw_input_price": "2.5", + "raw_output_price": "15.0", + "raw_model_id": "gemini-3.1-pro-preview", + "raw_provider_slug": "google", + "raw_provider_name": "ZeroEval" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "apex-agents::apex-agents-gemini-3.1-pro-preview", + "evaluation_name": "llm_stats.apex-agents", + "source_data": { + "dataset_name": "APEX-Agents", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-3.1-pro-preview", + "https://llm-stats.com/benchmarks/apex-agents", + "https://api.llm-stats.com/leaderboard/benchmarks/apex-agents" + ], + "additional_details": { + "raw_benchmark_id": "apex-agents", + "raw_model_id": "gemini-3.1-pro-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "APEX-Agents is a benchmark evaluating AI agents on long horizon professional tasks that require sustained reasoning, planning, and execution across complex multi-step workflows.", + "metric_id": "llm_stats.apex-agents.score", + "metric_name": "APEX-Agents score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "apex-agents", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "APEX-Agents", + "raw_categories": "[\"reasoning\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.335, + "details": { + "raw_score": "0.335", + "raw_score_field": "score", + "raw_model_id": "gemini-3.1-pro-preview", + "raw_benchmark_id": "apex-agents", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-3.1-pro-preview\",\"https://llm-stats.com/benchmarks/apex-agents\",\"https://api.llm-stats.com/leaderboard/benchmarks/apex-agents\"]", + "raw_score_id": "apex-agents::gemini-3.1-pro-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "arc-agi-v2::arc-agi-v2-gemini-3.1-pro-preview", + "evaluation_name": "llm_stats.arc-agi-v2", + "source_data": { + "dataset_name": "ARC-AGI v2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-3.1-pro-preview", + "https://llm-stats.com/benchmarks/arc-agi-v2", + "https://api.llm-stats.com/leaderboard/benchmarks/arc-agi-v2" + ], + "additional_details": { + "raw_benchmark_id": "arc-agi-v2", + "raw_model_id": "gemini-3.1-pro-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ARC-AGI-2 is an upgraded benchmark for measuring abstract reasoning and problem-solving abilities in AI systems through visual grid transformation tasks. It evaluates fluid intelligence via input-output grid pairs (1x1 to 30x30) using colored cells (0-9), requiring models to identify underlying transformation rules from demonstration examples and apply them to test cases. Designed to be easy for humans but challenging for AI, focusing on core cognitive abilities like spatial reasoning, pattern recognition, and compositional generalization.", + "metric_id": "llm_stats.arc-agi-v2.score", + "metric_name": "ARC-AGI v2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arc-agi-v2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ARC-AGI v2", + "raw_categories": "[\"reasoning\",\"spatial_reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.771, + "details": { + "raw_score": "0.771", + "raw_score_field": "score", + "raw_model_id": "gemini-3.1-pro-preview", + "raw_benchmark_id": "arc-agi-v2", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-3.1-pro-preview\",\"https://llm-stats.com/benchmarks/arc-agi-v2\",\"https://api.llm-stats.com/leaderboard/benchmarks/arc-agi-v2\"]", + "raw_score_id": "arc-agi-v2::gemini-3.1-pro-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "browsecomp::browsecomp-gemini-3.1-pro-preview", + "evaluation_name": "llm_stats.browsecomp", + "source_data": { + "dataset_name": "BrowseComp", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-3.1-pro-preview", + "https://llm-stats.com/benchmarks/browsecomp", + "https://api.llm-stats.com/leaderboard/benchmarks/browsecomp" + ], + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_model_id": "gemini-3.1-pro-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BrowseComp is a benchmark comprising 1,266 questions that challenge AI agents to persistently navigate the internet in search of hard-to-find, entangled information. The benchmark measures agents' ability to exercise persistence in information gathering, demonstrate creativity in web navigation, and find concise, verifiable answers. Despite the difficulty of the questions, BrowseComp is simple and easy-to-use, as predicted answers are short and easily verifiable against reference answers.", + "metric_id": "llm_stats.browsecomp.score", + "metric_name": "BrowseComp score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BrowseComp", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "52" + } + }, + "score_details": { + "score": 0.859, + "details": { + "raw_score": "0.859", + "raw_score_field": "score", + "raw_model_id": "gemini-3.1-pro-preview", + "raw_benchmark_id": "browsecomp", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-3.1-pro-preview\",\"https://llm-stats.com/benchmarks/browsecomp\",\"https://api.llm-stats.com/leaderboard/benchmarks/browsecomp\"]", + "raw_score_id": "browsecomp::gemini-3.1-pro-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gpqa::gpqa-gemini-3.1-pro-preview", + "evaluation_name": "llm_stats.gpqa", + "source_data": { + "dataset_name": "GPQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-3.1-pro-preview", + "https://llm-stats.com/benchmarks/gpqa", + "https://api.llm-stats.com/leaderboard/benchmarks/gpqa" + ], + "additional_details": { + "raw_benchmark_id": "gpqa", + "raw_model_id": "gemini-3.1-pro-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A challenging dataset of 448 multiple-choice questions written by domain experts in biology, physics, and chemistry. Questions are Google-proof and extremely difficult, with PhD experts reaching 65% accuracy.", + "metric_id": "llm_stats.gpqa.score", + "metric_name": "GPQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gpqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GPQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\",\"biology\",\"chemistry\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "226" + } + }, + "score_details": { + "score": 0.943, + "details": { + "raw_score": "0.943", + "raw_score_field": "score", + "raw_model_id": "gemini-3.1-pro-preview", + "raw_benchmark_id": "gpqa", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-3.1-pro-preview\",\"https://llm-stats.com/benchmarks/gpqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/gpqa\"]", + "raw_score_id": "gpqa::gemini-3.1-pro-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "humanity-s-last-exam::humanity-s-last-exam-gemini-3.1-pro-preview", + "evaluation_name": "llm_stats.humanity-s-last-exam", + "source_data": { + "dataset_name": "Humanity's Last Exam", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-3.1-pro-preview", + "https://llm-stats.com/benchmarks/humanity-s-last-exam", + "https://api.llm-stats.com/leaderboard/benchmarks/humanity's-last-exam" + ], + "additional_details": { + "raw_benchmark_id": "humanity's-last-exam", + "raw_model_id": "gemini-3.1-pro-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Humanity's Last Exam (HLE) is a multi-modal academic benchmark with 2,500 questions across mathematics, humanities, and natural sciences, designed to test LLM capabilities at the frontier of human knowledge with unambiguous, verifiable solutions", + "metric_id": "llm_stats.humanity-s-last-exam.score", + "metric_name": "Humanity's Last Exam score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "humanity's-last-exam", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Humanity's Last Exam", + "raw_categories": "[\"math\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "88" + } + }, + "score_details": { + "score": 0.514, + "details": { + "raw_score": "0.514", + "raw_score_field": "score", + "raw_model_id": "gemini-3.1-pro-preview", + "raw_benchmark_id": "humanity's-last-exam", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-3.1-pro-preview\",\"https://llm-stats.com/benchmarks/humanity-s-last-exam\",\"https://api.llm-stats.com/leaderboard/benchmarks/humanity's-last-exam\"]", + "raw_score_id": "humanity's-last-exam::gemini-3.1-pro-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "livecodebench-pro::livecodebench-pro-gemini-3.1-pro-preview", + "evaluation_name": "llm_stats.livecodebench-pro", + "source_data": { + "dataset_name": "LiveCodeBench Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-3.1-pro-preview", + "https://llm-stats.com/benchmarks/livecodebench-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/livecodebench-pro" + ], + "additional_details": { + "raw_benchmark_id": "livecodebench-pro", + "raw_model_id": "gemini-3.1-pro-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LiveCodeBench Pro is an advanced evaluation benchmark for large language models for code that uses Elo ratings to rank models based on their performance on coding tasks. It evaluates models on real-world coding problems from programming contests (LeetCode, AtCoder, CodeForces) and provides a relative ranking system where higher Elo scores indicate superior performance.", + "metric_id": "llm_stats.livecodebench-pro.score", + "metric_name": "LiveCodeBench Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "points", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 2887.0, + "additional_details": { + "raw_benchmark_id": "livecodebench-pro", + "raw_score_field": "score", + "bound_strategy": "observed_value", + "raw_name": "LiveCodeBench Pro", + "raw_categories": "[\"reasoning\",\"general\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 2887.0, + "details": { + "raw_score": "2887.0", + "raw_score_field": "score", + "raw_model_id": "gemini-3.1-pro-preview", + "raw_benchmark_id": "livecodebench-pro", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-3.1-pro-preview\",\"https://llm-stats.com/benchmarks/livecodebench-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/livecodebench-pro\"]", + "raw_score_id": "livecodebench-pro::gemini-3.1-pro-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mcp-atlas::mcp-atlas-gemini-3.1-pro-preview", + "evaluation_name": "llm_stats.mcp-atlas", + "source_data": { + "dataset_name": "MCP Atlas", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-3.1-pro-preview", + "https://llm-stats.com/benchmarks/mcp-atlas", + "https://api.llm-stats.com/leaderboard/benchmarks/mcp-atlas" + ], + "additional_details": { + "raw_benchmark_id": "mcp-atlas", + "raw_model_id": "gemini-3.1-pro-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MCP Atlas is a benchmark for evaluating AI models on scaled tool use capabilities, measuring how well models can coordinate and utilize multiple tools across complex multi-step tasks.", + "metric_id": "llm_stats.mcp-atlas.score", + "metric_name": "MCP Atlas score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mcp-atlas", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MCP Atlas", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.692, + "details": { + "raw_score": "0.692", + "raw_score_field": "score", + "raw_model_id": "gemini-3.1-pro-preview", + "raw_benchmark_id": "mcp-atlas", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-3.1-pro-preview\",\"https://llm-stats.com/benchmarks/mcp-atlas\",\"https://api.llm-stats.com/leaderboard/benchmarks/mcp-atlas\"]", + "raw_score_id": "mcp-atlas::gemini-3.1-pro-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmmlu::mmmlu-gemini-3.1-pro-preview", + "evaluation_name": "llm_stats.mmmlu", + "source_data": { + "dataset_name": "MMMLU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-3.1-pro-preview", + "https://llm-stats.com/benchmarks/mmmlu", + "https://api.llm-stats.com/leaderboard/benchmarks/mmmlu" + ], + "additional_details": { + "raw_benchmark_id": "mmmlu", + "raw_model_id": "gemini-3.1-pro-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Multilingual Massive Multitask Language Understanding dataset released by OpenAI, featuring professionally translated MMLU test questions across 14 languages including Arabic, Bengali, German, Spanish, French, Hindi, Indonesian, Italian, Japanese, Korean, Portuguese, Swahili, Yoruba, and Chinese. Contains approximately 15,908 multiple-choice questions per language covering 57 subjects.", + "metric_id": "llm_stats.mmmlu.score", + "metric_name": "MMMLU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmmlu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMMLU", + "raw_categories": "[\"language\",\"math\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "49" + } + }, + "score_details": { + "score": 0.926, + "details": { + "raw_score": "0.926", + "raw_score_field": "score", + "raw_model_id": "gemini-3.1-pro-preview", + "raw_benchmark_id": "mmmlu", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-3.1-pro-preview\",\"https://llm-stats.com/benchmarks/mmmlu\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmmlu\"]", + "raw_score_id": "mmmlu::gemini-3.1-pro-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmmu-pro::mmmu-pro-gemini-3.1-pro-preview", + "evaluation_name": "llm_stats.mmmu-pro", + "source_data": { + "dataset_name": "MMMU-Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-3.1-pro-preview", + "https://llm-stats.com/benchmarks/mmmu-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/mmmu-pro" + ], + "additional_details": { + "raw_benchmark_id": "mmmu-pro", + "raw_model_id": "gemini-3.1-pro-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A more robust multi-discipline multimodal understanding benchmark that enhances MMMU through a three-step process: filtering text-only answerable questions, augmenting candidate options, and introducing vision-only input settings. Achieves significantly lower model performance (16.8-26.9%) compared to original MMMU, providing more rigorous evaluation that closely mimics real-world scenarios.", + "metric_id": "llm_stats.mmmu-pro.score", + "metric_name": "MMMU-Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmmu-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMMU-Pro", + "raw_categories": "[\"multimodal\",\"reasoning\",\"general\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "60" + } + }, + "score_details": { + "score": 0.805, + "details": { + "raw_score": "0.805", + "raw_score_field": "score", + "raw_model_id": "gemini-3.1-pro-preview", + "raw_benchmark_id": "mmmu-pro", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-3.1-pro-preview\",\"https://llm-stats.com/benchmarks/mmmu-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmmu-pro\"]", + "raw_score_id": "mmmu-pro::gemini-3.1-pro-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mrcr-v2-8-needle::mrcr-v2-8-needle-gemini-3.1-pro-preview", + "evaluation_name": "llm_stats.mrcr-v2-8-needle", + "source_data": { + "dataset_name": "MRCR v2 (8-needle)", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-3.1-pro-preview", + "https://llm-stats.com/benchmarks/mrcr-v2-8-needle", + "https://api.llm-stats.com/leaderboard/benchmarks/mrcr-v2-(8-needle)" + ], + "additional_details": { + "raw_benchmark_id": "mrcr-v2-(8-needle)", + "raw_model_id": "gemini-3.1-pro-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MRCR v2 (8-needle) is a variant of the Multi-Round Coreference Resolution benchmark that includes 8 needle items to retrieve from long contexts. This tests models' ability to simultaneously track and reason about multiple pieces of information across extended conversations.", + "metric_id": "llm_stats.mrcr-v2-8-needle.score", + "metric_name": "MRCR v2 (8-needle) score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mrcr-v2-(8-needle)", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MRCR v2 (8-needle)", + "raw_categories": "[\"long_context\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.263, + "details": { + "raw_score": "0.263", + "raw_score_field": "score", + "raw_model_id": "gemini-3.1-pro-preview", + "raw_benchmark_id": "mrcr-v2-(8-needle)", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-3.1-pro-preview\",\"https://llm-stats.com/benchmarks/mrcr-v2-8-needle\",\"https://api.llm-stats.com/leaderboard/benchmarks/mrcr-v2-(8-needle)\"]", + "raw_score_id": "mrcr-v2-(8-needle)::gemini-3.1-pro-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "scicode::scicode-gemini-3.1-pro-preview", + "evaluation_name": "llm_stats.scicode", + "source_data": { + "dataset_name": "SciCode", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-3.1-pro-preview", + "https://llm-stats.com/benchmarks/scicode", + "https://api.llm-stats.com/leaderboard/benchmarks/scicode" + ], + "additional_details": { + "raw_benchmark_id": "scicode", + "raw_model_id": "gemini-3.1-pro-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SciCode is a research coding benchmark curated by scientists that challenges language models to code solutions for scientific problems. It contains 338 subproblems decomposed from 80 challenging main problems across 16 natural science sub-fields including mathematics, physics, chemistry, biology, and materials science. Problems require knowledge recall, reasoning, and code synthesis skills.", + "metric_id": "llm_stats.scicode.score", + "metric_name": "SciCode score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "scicode", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SciCode", + "raw_categories": "[\"math\",\"physics\",\"reasoning\",\"biology\",\"chemistry\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.59, + "details": { + "raw_score": "0.59", + "raw_score_field": "score", + "raw_model_id": "gemini-3.1-pro-preview", + "raw_benchmark_id": "scicode", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-3.1-pro-preview\",\"https://llm-stats.com/benchmarks/scicode\",\"https://api.llm-stats.com/leaderboard/benchmarks/scicode\"]", + "raw_score_id": "scicode::gemini-3.1-pro-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-verified::swe-bench-verified-gemini-3.1-pro-preview", + "evaluation_name": "llm_stats.swe-bench-verified", + "source_data": { + "dataset_name": "SWE-Bench Verified", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-3.1-pro-preview", + "https://llm-stats.com/benchmarks/swe-bench-verified", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-verified" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-verified", + "raw_model_id": "gemini-3.1-pro-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A verified subset of 500 software engineering problems from real GitHub issues, validated by human annotators for evaluating language models' ability to resolve real-world coding issues by generating patches for Python codebases.", + "metric_id": "llm_stats.swe-bench-verified.score", + "metric_name": "SWE-Bench Verified score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-verified", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Bench Verified", + "raw_categories": "[\"reasoning\",\"frontend_development\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "103" + } + }, + "score_details": { + "score": 0.806, + "details": { + "raw_score": "0.806", + "raw_score_field": "score", + "raw_model_id": "gemini-3.1-pro-preview", + "raw_benchmark_id": "swe-bench-verified", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-3.1-pro-preview\",\"https://llm-stats.com/benchmarks/swe-bench-verified\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-verified\"]", + "raw_score_id": "swe-bench-verified::gemini-3.1-pro-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "t2-bench::t2-bench-gemini-3.1-pro-preview", + "evaluation_name": "llm_stats.t2-bench", + "source_data": { + "dataset_name": "t2-bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-3.1-pro-preview", + "https://llm-stats.com/benchmarks/t2-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/t2-bench" + ], + "additional_details": { + "raw_benchmark_id": "t2-bench", + "raw_model_id": "gemini-3.1-pro-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "t2-bench is a benchmark for evaluating agentic tool use capabilities, measuring how well models can select, sequence, and utilize tools to solve complex tasks. It tests autonomous planning and execution in multi-step scenarios.", + "metric_id": "llm_stats.t2-bench.score", + "metric_name": "t2-bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "t2-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "t2-bench", + "raw_categories": "[\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.993, + "details": { + "raw_score": "0.993", + "raw_score_field": "score", + "raw_model_id": "gemini-3.1-pro-preview", + "raw_benchmark_id": "t2-bench", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-3.1-pro-preview\",\"https://llm-stats.com/benchmarks/t2-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/t2-bench\"]", + "raw_score_id": "t2-bench::gemini-3.1-pro-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2::terminal-bench-2-gemini-3.1-pro-preview", + "evaluation_name": "llm_stats.terminal-bench-2", + "source_data": { + "dataset_name": "Terminal-Bench 2.0", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-3.1-pro-preview", + "https://llm-stats.com/benchmarks/terminal-bench-2", + "https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2" + ], + "additional_details": { + "raw_benchmark_id": "terminal-bench-2", + "raw_model_id": "gemini-3.1-pro-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Terminal-Bench 2.0 is an updated benchmark for testing AI agents' tool use ability to operate a computer via terminal. It evaluates how well models can handle real-world, end-to-end tasks autonomously, including compiling code, training models, setting up servers, system administration, security tasks, data science workflows, and cybersecurity vulnerabilities.", + "metric_id": "llm_stats.terminal-bench-2.score", + "metric_name": "Terminal-Bench 2.0 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "terminal-bench-2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Terminal-Bench 2.0", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "49" + } + }, + "score_details": { + "score": 0.685, + "details": { + "raw_score": "0.685", + "raw_score_field": "score", + "raw_model_id": "gemini-3.1-pro-preview", + "raw_benchmark_id": "terminal-bench-2", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-3.1-pro-preview\",\"https://llm-stats.com/benchmarks/terminal-bench-2\",\"https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2\"]", + "raw_score_id": "terminal-bench-2::gemini-3.1-pro-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/google/gemini-diffusion/1f5011a8-54cf-4a65-bad7-93af38b1c1ae.json b/data/llm_stats/google/gemini-diffusion/1f5011a8-54cf-4a65-bad7-93af38b1c1ae.json new file mode 100644 index 0000000000000000000000000000000000000000..3a02c3d39049f04c0189e27990ab7551b6629201 --- /dev/null +++ b/data/llm_stats/google/gemini-diffusion/1f5011a8-54cf-4a65-bad7-93af38b1c1ae.json @@ -0,0 +1,358 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/google_gemini-diffusion/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Gemini Diffusion", + "id": "google/gemini-diffusion", + "developer": "google", + "additional_details": { + "raw_id": "gemini-diffusion", + "raw_name": "Gemini Diffusion", + "raw_model_id": "gemini-diffusion", + "raw_model_name": "Gemini Diffusion", + "raw_organization_id": "google", + "raw_organization_name": "Google", + "raw_release_date": "2025-05-20", + "raw_announcement_date": "2025-05-20", + "raw_multimodal": "false", + "raw_provider_slug": "google", + "raw_provider_name": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "big-bench-extra-hard::big-bench-extra-hard-gemini-diffusion", + "evaluation_name": "llm_stats.big-bench-extra-hard", + "source_data": { + "dataset_name": "BIG-Bench Extra Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-diffusion", + "https://llm-stats.com/benchmarks/big-bench-extra-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench-extra-hard" + ], + "additional_details": { + "raw_benchmark_id": "big-bench-extra-hard", + "raw_model_id": "gemini-diffusion", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIG-Bench Extra Hard (BBEH) is a challenging benchmark that replaces each task in BIG-Bench Hard with a novel task that probes similar reasoning capabilities but exhibits significantly increased difficulty. The benchmark contains 23 tasks testing diverse reasoning skills including many-hop reasoning, causal understanding, spatial reasoning, temporal arithmetic, geometric reasoning, linguistic reasoning, logic puzzles, and humor understanding. Designed to address saturation on existing benchmarks where state-of-the-art models achieve near-perfect scores, BBEH shows substantial room for improvement with best models achieving only 9.8-44.8% average accuracy.", + "metric_id": "llm_stats.big-bench-extra-hard.score", + "metric_name": "BIG-Bench Extra Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench-extra-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench Extra Hard", + "raw_categories": "[\"language\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.15, + "details": { + "raw_score": "0.15", + "raw_score_field": "score", + "raw_model_id": "gemini-diffusion", + "raw_benchmark_id": "big-bench-extra-hard", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-diffusion\",\"https://llm-stats.com/benchmarks/big-bench-extra-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench-extra-hard\"]", + "raw_score_id": "big-bench-extra-hard::gemini-diffusion", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bigcodebench::bigcodebench-gemini-diffusion", + "evaluation_name": "llm_stats.bigcodebench", + "source_data": { + "dataset_name": "BigCodeBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-diffusion", + "https://llm-stats.com/benchmarks/bigcodebench", + "https://api.llm-stats.com/leaderboard/benchmarks/bigcodebench" + ], + "additional_details": { + "raw_benchmark_id": "bigcodebench", + "raw_model_id": "gemini-diffusion", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A benchmark that challenges LLMs to invoke multiple function calls as tools from 139 libraries and 7 domains for 1,140 fine-grained programming tasks. Evaluates code generation with diverse function calls and complex instructions, featuring two variants: Complete (code completion based on comprehensive docstrings) and Instruct (generating code from natural language instructions).", + "metric_id": "llm_stats.bigcodebench.score", + "metric_name": "BigCodeBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bigcodebench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BigCodeBench", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.454, + "details": { + "raw_score": "0.454", + "raw_score_field": "score", + "raw_model_id": "gemini-diffusion", + "raw_benchmark_id": "bigcodebench", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-diffusion\",\"https://llm-stats.com/benchmarks/bigcodebench\",\"https://api.llm-stats.com/leaderboard/benchmarks/bigcodebench\"]", + "raw_score_id": "bigcodebench::gemini-diffusion", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "global-mmlu-lite::global-mmlu-lite-gemini-diffusion", + "evaluation_name": "llm_stats.global-mmlu-lite", + "source_data": { + "dataset_name": "Global-MMLU-Lite", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-diffusion", + "https://llm-stats.com/benchmarks/global-mmlu-lite", + "https://api.llm-stats.com/leaderboard/benchmarks/global-mmlu-lite" + ], + "additional_details": { + "raw_benchmark_id": "global-mmlu-lite", + "raw_model_id": "gemini-diffusion", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A lightweight version of Global MMLU benchmark that evaluates language models across multiple languages while addressing cultural and linguistic biases in multilingual evaluation.", + "metric_id": "llm_stats.global-mmlu-lite.score", + "metric_name": "Global-MMLU-Lite score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "global-mmlu-lite", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Global-MMLU-Lite", + "raw_categories": "[\"language\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "14" + } + }, + "score_details": { + "score": 0.691, + "details": { + "raw_score": "0.691", + "raw_score_field": "score", + "raw_model_id": "gemini-diffusion", + "raw_benchmark_id": "global-mmlu-lite", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-diffusion\",\"https://llm-stats.com/benchmarks/global-mmlu-lite\",\"https://api.llm-stats.com/leaderboard/benchmarks/global-mmlu-lite\"]", + "raw_score_id": "global-mmlu-lite::gemini-diffusion", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "humaneval::humaneval-gemini-diffusion", + "evaluation_name": "llm_stats.humaneval", + "source_data": { + "dataset_name": "HumanEval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-diffusion", + "https://llm-stats.com/benchmarks/humaneval", + "https://api.llm-stats.com/leaderboard/benchmarks/humaneval" + ], + "additional_details": { + "raw_benchmark_id": "humaneval", + "raw_model_id": "gemini-diffusion", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A benchmark that measures functional correctness for synthesizing programs from docstrings, consisting of 164 original programming problems assessing language comprehension, algorithms, and simple mathematics", + "metric_id": "llm_stats.humaneval.score", + "metric_name": "HumanEval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "humaneval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HumanEval", + "raw_categories": "[\"reasoning\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "66" + } + }, + "score_details": { + "score": 0.896, + "details": { + "raw_score": "0.896", + "raw_score_field": "score", + "raw_model_id": "gemini-diffusion", + "raw_benchmark_id": "humaneval", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-diffusion\",\"https://llm-stats.com/benchmarks/humaneval\",\"https://api.llm-stats.com/leaderboard/benchmarks/humaneval\"]", + "raw_score_id": "humaneval::gemini-diffusion", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "lbpp-v2::lbpp-v2-gemini-diffusion", + "evaluation_name": "llm_stats.lbpp-v2", + "source_data": { + "dataset_name": "LBPP (v2)", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-diffusion", + "https://llm-stats.com/benchmarks/lbpp-v2", + "https://api.llm-stats.com/leaderboard/benchmarks/lbpp-(v2)" + ], + "additional_details": { + "raw_benchmark_id": "lbpp-(v2)", + "raw_model_id": "gemini-diffusion", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LBPP (v2) benchmark - specific documentation not found in official sources, possibly related to language-based planning problems", + "metric_id": "llm_stats.lbpp-v2.score", + "metric_name": "LBPP (v2) score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "lbpp-(v2)", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LBPP (v2)", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.568, + "details": { + "raw_score": "0.568", + "raw_score_field": "score", + "raw_model_id": "gemini-diffusion", + "raw_benchmark_id": "lbpp-(v2)", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-diffusion\",\"https://llm-stats.com/benchmarks/lbpp-v2\",\"https://api.llm-stats.com/leaderboard/benchmarks/lbpp-(v2)\"]", + "raw_score_id": "lbpp-(v2)::gemini-diffusion", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mbpp::mbpp-gemini-diffusion", + "evaluation_name": "llm_stats.mbpp", + "source_data": { + "dataset_name": "MBPP", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemini-diffusion", + "https://llm-stats.com/benchmarks/mbpp", + "https://api.llm-stats.com/leaderboard/benchmarks/mbpp" + ], + "additional_details": { + "raw_benchmark_id": "mbpp", + "raw_model_id": "gemini-diffusion", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MBPP (Mostly Basic Python Problems) is a benchmark of 974 crowd-sourced Python programming problems designed to be solvable by entry-level programmers. Each problem consists of a task description, code solution, and 3 automated test cases covering programming fundamentals and standard library functionality.", + "metric_id": "llm_stats.mbpp.score", + "metric_name": "MBPP score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mbpp", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MBPP", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "33" + } + }, + "score_details": { + "score": 0.76, + "details": { + "raw_score": "0.76", + "raw_score_field": "score", + "raw_model_id": "gemini-diffusion", + "raw_benchmark_id": "mbpp", + "source_urls_json": "[\"https://llm-stats.com/models/gemini-diffusion\",\"https://llm-stats.com/benchmarks/mbpp\",\"https://api.llm-stats.com/leaderboard/benchmarks/mbpp\"]", + "raw_score_id": "mbpp::gemini-diffusion", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/google/gemma-2-27b-it/36ab0038-fb95-4a29-98d7-5e70b3b63e2e.json b/data/llm_stats/google/gemma-2-27b-it/36ab0038-fb95-4a29-98d7-5e70b3b63e2e.json new file mode 100644 index 0000000000000000000000000000000000000000..bc5d0bc354965211c1f1a64151a86c20bbad4262 --- /dev/null +++ b/data/llm_stats/google/gemma-2-27b-it/36ab0038-fb95-4a29-98d7-5e70b3b63e2e.json @@ -0,0 +1,619 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/google_gemma-2-27b-it/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Gemma 2 27B", + "id": "google/gemma-2-27b-it", + "developer": "google", + "additional_details": { + "raw_id": "gemma-2-27b-it", + "raw_name": "Gemma 2 27B", + "raw_model_id": "gemma-2-27b-it", + "raw_model_name": "Gemma 2 27B", + "raw_organization_id": "google", + "raw_organization_name": "Google", + "raw_release_date": "2024-06-27", + "raw_announcement_date": "2024-06-27", + "raw_multimodal": "false", + "raw_param_count": "27200000000", + "raw_provider_slug": "google", + "raw_provider_name": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "agieval::agieval-gemma-2-27b-it", + "evaluation_name": "llm_stats.agieval", + "source_data": { + "dataset_name": "AGIEval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-2-27b-it", + "https://llm-stats.com/benchmarks/agieval", + "https://api.llm-stats.com/leaderboard/benchmarks/agieval" + ], + "additional_details": { + "raw_benchmark_id": "agieval", + "raw_model_id": "gemma-2-27b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A human-centric benchmark for evaluating foundation models on standardized exams including college entrance exams (Gaokao, SAT), law school admission tests (LSAT), math competitions, lawyer qualification tests, and civil service exams. Contains 20 tasks (18 multiple-choice, 2 cloze) designed to assess understanding, knowledge, reasoning, and calculation abilities in real-world academic and professional contexts.", + "metric_id": "llm_stats.agieval.score", + "metric_name": "AGIEval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "agieval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AGIEval", + "raw_categories": "[\"legal\",\"math\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.551, + "details": { + "raw_score": "0.551", + "raw_score_field": "score", + "raw_model_id": "gemma-2-27b-it", + "raw_benchmark_id": "agieval", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-2-27b-it\",\"https://llm-stats.com/benchmarks/agieval\",\"https://api.llm-stats.com/leaderboard/benchmarks/agieval\"]", + "raw_score_id": "agieval::gemma-2-27b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "arc-c::arc-c-gemma-2-27b-it", + "evaluation_name": "llm_stats.arc-c", + "source_data": { + "dataset_name": "ARC-C", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-2-27b-it", + "https://llm-stats.com/benchmarks/arc-c", + "https://api.llm-stats.com/leaderboard/benchmarks/arc-c" + ], + "additional_details": { + "raw_benchmark_id": "arc-c", + "raw_model_id": "gemma-2-27b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The AI2 Reasoning Challenge (ARC) Challenge Set is a multiple-choice question-answering benchmark containing grade-school level science questions that require advanced reasoning capabilities. ARC-C specifically contains questions that were answered incorrectly by both retrieval-based and word co-occurrence algorithms, making it a particularly challenging subset designed to test commonsense reasoning abilities in AI systems.", + "metric_id": "llm_stats.arc-c.score", + "metric_name": "ARC-C score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arc-c", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ARC-C", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "34" + } + }, + "score_details": { + "score": 0.714, + "details": { + "raw_score": "0.714", + "raw_score_field": "score", + "raw_model_id": "gemma-2-27b-it", + "raw_benchmark_id": "arc-c", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-2-27b-it\",\"https://llm-stats.com/benchmarks/arc-c\",\"https://api.llm-stats.com/leaderboard/benchmarks/arc-c\"]", + "raw_score_id": "arc-c::gemma-2-27b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "arc-e::arc-e-gemma-2-27b-it", + "evaluation_name": "llm_stats.arc-e", + "source_data": { + "dataset_name": "ARC-E", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-2-27b-it", + "https://llm-stats.com/benchmarks/arc-e", + "https://api.llm-stats.com/leaderboard/benchmarks/arc-e" + ], + "additional_details": { + "raw_benchmark_id": "arc-e", + "raw_model_id": "gemma-2-27b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ARC-E (AI2 Reasoning Challenge - Easy Set) is a subset of grade-school level, multiple-choice science questions that requires knowledge and reasoning capabilities. Part of the AI2 Reasoning Challenge dataset containing 5,197 questions that test scientific reasoning and factual knowledge. The Easy Set contains questions that are answerable by retrieval-based and word co-occurrence algorithms, making them more accessible than the Challenge Set.", + "metric_id": "llm_stats.arc-e.score", + "metric_name": "ARC-E score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arc-e", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ARC-E", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.886, + "details": { + "raw_score": "0.886", + "raw_score_field": "score", + "raw_model_id": "gemma-2-27b-it", + "raw_benchmark_id": "arc-e", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-2-27b-it\",\"https://llm-stats.com/benchmarks/arc-e\",\"https://api.llm-stats.com/leaderboard/benchmarks/arc-e\"]", + "raw_score_id": "arc-e::gemma-2-27b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "big-bench::big-bench-gemma-2-27b-it", + "evaluation_name": "llm_stats.big-bench", + "source_data": { + "dataset_name": "BIG-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-2-27b-it", + "https://llm-stats.com/benchmarks/big-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench" + ], + "additional_details": { + "raw_benchmark_id": "big-bench", + "raw_model_id": "gemma-2-27b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark consisting of 204+ tasks designed to probe large language models and extrapolate their future capabilities. It covers diverse domains including linguistics, mathematics, common-sense reasoning, biology, physics, social bias, software development, and more. The benchmark focuses on tasks believed to be beyond current language model capabilities and includes both English and non-English tasks across multiple languages.", + "metric_id": "llm_stats.big-bench.score", + "metric_name": "BIG-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.749, + "details": { + "raw_score": "0.749", + "raw_score_field": "score", + "raw_model_id": "gemma-2-27b-it", + "raw_benchmark_id": "big-bench", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-2-27b-it\",\"https://llm-stats.com/benchmarks/big-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench\"]", + "raw_score_id": "big-bench::gemma-2-27b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "boolq::boolq-gemma-2-27b-it", + "evaluation_name": "llm_stats.boolq", + "source_data": { + "dataset_name": "BoolQ", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-2-27b-it", + "https://llm-stats.com/benchmarks/boolq", + "https://api.llm-stats.com/leaderboard/benchmarks/boolq" + ], + "additional_details": { + "raw_benchmark_id": "boolq", + "raw_model_id": "gemma-2-27b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BoolQ is a reading comprehension dataset for yes/no questions containing 15,942 naturally occurring examples. Each example consists of a question, passage, and boolean answer, where questions are generated in unprompted and unconstrained settings. The dataset challenges models with complex, non-factoid information requiring entailment-like inference to solve.", + "metric_id": "llm_stats.boolq.score", + "metric_name": "BoolQ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "boolq", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BoolQ", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.848, + "details": { + "raw_score": "0.848", + "raw_score_field": "score", + "raw_model_id": "gemma-2-27b-it", + "raw_benchmark_id": "boolq", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-2-27b-it\",\"https://llm-stats.com/benchmarks/boolq\",\"https://api.llm-stats.com/leaderboard/benchmarks/boolq\"]", + "raw_score_id": "boolq::gemma-2-27b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hellaswag::hellaswag-gemma-2-27b-it", + "evaluation_name": "llm_stats.hellaswag", + "source_data": { + "dataset_name": "HellaSwag", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-2-27b-it", + "https://llm-stats.com/benchmarks/hellaswag", + "https://api.llm-stats.com/leaderboard/benchmarks/hellaswag" + ], + "additional_details": { + "raw_benchmark_id": "hellaswag", + "raw_model_id": "gemma-2-27b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A challenging commonsense natural language inference dataset that uses Adversarial Filtering to create questions trivial for humans (>95% accuracy) but difficult for state-of-the-art models, requiring completion of sentence endings based on physical situations and everyday activities", + "metric_id": "llm_stats.hellaswag.score", + "metric_name": "HellaSwag score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hellaswag", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HellaSwag", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.864, + "details": { + "raw_score": "0.864", + "raw_score_field": "score", + "raw_model_id": "gemma-2-27b-it", + "raw_benchmark_id": "hellaswag", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-2-27b-it\",\"https://llm-stats.com/benchmarks/hellaswag\",\"https://api.llm-stats.com/leaderboard/benchmarks/hellaswag\"]", + "raw_score_id": "hellaswag::gemma-2-27b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "natural-questions::natural-questions-gemma-2-27b-it", + "evaluation_name": "llm_stats.natural-questions", + "source_data": { + "dataset_name": "Natural Questions", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-2-27b-it", + "https://llm-stats.com/benchmarks/natural-questions", + "https://api.llm-stats.com/leaderboard/benchmarks/natural-questions" + ], + "additional_details": { + "raw_benchmark_id": "natural-questions", + "raw_model_id": "gemma-2-27b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Natural Questions is a question answering dataset featuring real anonymized queries issued to Google search engine. It contains 307,373 training examples where annotators provide long answers (passages) and short answers (entities) from Wikipedia pages, or mark them as unanswerable.", + "metric_id": "llm_stats.natural-questions.score", + "metric_name": "Natural Questions score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "natural-questions", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Natural Questions", + "raw_categories": "[\"reasoning\",\"search\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.345, + "details": { + "raw_score": "0.345", + "raw_score_field": "score", + "raw_model_id": "gemma-2-27b-it", + "raw_benchmark_id": "natural-questions", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-2-27b-it\",\"https://llm-stats.com/benchmarks/natural-questions\",\"https://api.llm-stats.com/leaderboard/benchmarks/natural-questions\"]", + "raw_score_id": "natural-questions::gemma-2-27b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "piqa::piqa-gemma-2-27b-it", + "evaluation_name": "llm_stats.piqa", + "source_data": { + "dataset_name": "PIQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-2-27b-it", + "https://llm-stats.com/benchmarks/piqa", + "https://api.llm-stats.com/leaderboard/benchmarks/piqa" + ], + "additional_details": { + "raw_benchmark_id": "piqa", + "raw_model_id": "gemma-2-27b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "PIQA (Physical Interaction: Question Answering) is a benchmark dataset for physical commonsense reasoning in natural language. It tests AI systems' ability to answer questions requiring physical world knowledge through multiple choice questions with everyday situations, focusing on atypical solutions inspired by instructables.com. The dataset contains 21,000 multiple choice questions where models must choose the most appropriate solution for physical interactions.", + "metric_id": "llm_stats.piqa.score", + "metric_name": "PIQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "piqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PIQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.832, + "details": { + "raw_score": "0.832", + "raw_score_field": "score", + "raw_model_id": "gemma-2-27b-it", + "raw_benchmark_id": "piqa", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-2-27b-it\",\"https://llm-stats.com/benchmarks/piqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/piqa\"]", + "raw_score_id": "piqa::gemma-2-27b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "social-iqa::social-iqa-gemma-2-27b-it", + "evaluation_name": "llm_stats.social-iqa", + "source_data": { + "dataset_name": "Social IQa", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-2-27b-it", + "https://llm-stats.com/benchmarks/social-iqa", + "https://api.llm-stats.com/leaderboard/benchmarks/social-iqa" + ], + "additional_details": { + "raw_benchmark_id": "social-iqa", + "raw_model_id": "gemma-2-27b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The first large-scale benchmark for commonsense reasoning about social situations. Contains 38,000 multiple choice questions probing emotional and social intelligence in everyday situations, testing commonsense understanding of social interactions and theory of mind reasoning about the implied emotions and behavior of others.", + "metric_id": "llm_stats.social-iqa.score", + "metric_name": "Social IQa score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "social-iqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Social IQa", + "raw_categories": "[\"psychology\",\"reasoning\",\"creativity\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.537, + "details": { + "raw_score": "0.537", + "raw_score_field": "score", + "raw_model_id": "gemma-2-27b-it", + "raw_benchmark_id": "social-iqa", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-2-27b-it\",\"https://llm-stats.com/benchmarks/social-iqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/social-iqa\"]", + "raw_score_id": "social-iqa::gemma-2-27b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "triviaqa::triviaqa-gemma-2-27b-it", + "evaluation_name": "llm_stats.triviaqa", + "source_data": { + "dataset_name": "TriviaQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-2-27b-it", + "https://llm-stats.com/benchmarks/triviaqa", + "https://api.llm-stats.com/leaderboard/benchmarks/triviaqa" + ], + "additional_details": { + "raw_benchmark_id": "triviaqa", + "raw_model_id": "gemma-2-27b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A large-scale reading comprehension dataset containing over 650K question-answer-evidence triples. TriviaQA includes 95K question-answer pairs authored by trivia enthusiasts and independently gathered evidence documents (six per question on average) that provide high quality distant supervision for answering the questions. The dataset features relatively complex, compositional questions with considerable syntactic and lexical variability, requiring cross-sentence reasoning to find answers.", + "metric_id": "llm_stats.triviaqa.score", + "metric_name": "TriviaQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "triviaqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TriviaQA", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.837, + "details": { + "raw_score": "0.837", + "raw_score_field": "score", + "raw_model_id": "gemma-2-27b-it", + "raw_benchmark_id": "triviaqa", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-2-27b-it\",\"https://llm-stats.com/benchmarks/triviaqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/triviaqa\"]", + "raw_score_id": "triviaqa::gemma-2-27b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "winogrande::winogrande-gemma-2-27b-it", + "evaluation_name": "llm_stats.winogrande", + "source_data": { + "dataset_name": "Winogrande", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-2-27b-it", + "https://llm-stats.com/benchmarks/winogrande", + "https://api.llm-stats.com/leaderboard/benchmarks/winogrande" + ], + "additional_details": { + "raw_benchmark_id": "winogrande", + "raw_model_id": "gemma-2-27b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WinoGrande: An Adversarial Winograd Schema Challenge at Scale. A large-scale dataset of 44,000 pronoun resolution problems designed to test machine commonsense reasoning. Uses adversarial filtering to reduce spurious biases and provides a more robust evaluation of whether AI systems truly understand commonsense or exploit statistical shortcuts. Current best AI methods achieve 59.4-79.1% accuracy, significantly below human performance of 94.0%.", + "metric_id": "llm_stats.winogrande.score", + "metric_name": "Winogrande score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "winogrande", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Winogrande", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "22" + } + }, + "score_details": { + "score": 0.837, + "details": { + "raw_score": "0.837", + "raw_score_field": "score", + "raw_model_id": "gemma-2-27b-it", + "raw_benchmark_id": "winogrande", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-2-27b-it\",\"https://llm-stats.com/benchmarks/winogrande\",\"https://api.llm-stats.com/leaderboard/benchmarks/winogrande\"]", + "raw_score_id": "winogrande::gemma-2-27b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/google/gemma-2-9b-it/17a1e708-8194-4990-92d1-8dd7c7bb900b.json b/data/llm_stats/google/gemma-2-9b-it/17a1e708-8194-4990-92d1-8dd7c7bb900b.json new file mode 100644 index 0000000000000000000000000000000000000000..053c8da24948673e748e4b2ea8b8ea34d8162641 --- /dev/null +++ b/data/llm_stats/google/gemma-2-9b-it/17a1e708-8194-4990-92d1-8dd7c7bb900b.json @@ -0,0 +1,567 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/google_gemma-2-9b-it/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Gemma 2 9B", + "id": "google/gemma-2-9b-it", + "developer": "google", + "additional_details": { + "raw_id": "gemma-2-9b-it", + "raw_name": "Gemma 2 9B", + "raw_model_id": "gemma-2-9b-it", + "raw_model_name": "Gemma 2 9B", + "raw_organization_id": "google", + "raw_organization_name": "Google", + "raw_release_date": "2024-06-27", + "raw_announcement_date": "2024-06-27", + "raw_multimodal": "false", + "raw_param_count": "9240000000", + "raw_provider_slug": "google", + "raw_provider_name": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "agieval::agieval-gemma-2-9b-it", + "evaluation_name": "llm_stats.agieval", + "source_data": { + "dataset_name": "AGIEval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-2-9b-it", + "https://llm-stats.com/benchmarks/agieval", + "https://api.llm-stats.com/leaderboard/benchmarks/agieval" + ], + "additional_details": { + "raw_benchmark_id": "agieval", + "raw_model_id": "gemma-2-9b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A human-centric benchmark for evaluating foundation models on standardized exams including college entrance exams (Gaokao, SAT), law school admission tests (LSAT), math competitions, lawyer qualification tests, and civil service exams. Contains 20 tasks (18 multiple-choice, 2 cloze) designed to assess understanding, knowledge, reasoning, and calculation abilities in real-world academic and professional contexts.", + "metric_id": "llm_stats.agieval.score", + "metric_name": "AGIEval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "agieval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AGIEval", + "raw_categories": "[\"legal\",\"math\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.528, + "details": { + "raw_score": "0.528", + "raw_score_field": "score", + "raw_model_id": "gemma-2-9b-it", + "raw_benchmark_id": "agieval", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-2-9b-it\",\"https://llm-stats.com/benchmarks/agieval\",\"https://api.llm-stats.com/leaderboard/benchmarks/agieval\"]", + "raw_score_id": "agieval::gemma-2-9b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "arc-e::arc-e-gemma-2-9b-it", + "evaluation_name": "llm_stats.arc-e", + "source_data": { + "dataset_name": "ARC-E", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-2-9b-it", + "https://llm-stats.com/benchmarks/arc-e", + "https://api.llm-stats.com/leaderboard/benchmarks/arc-e" + ], + "additional_details": { + "raw_benchmark_id": "arc-e", + "raw_model_id": "gemma-2-9b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ARC-E (AI2 Reasoning Challenge - Easy Set) is a subset of grade-school level, multiple-choice science questions that requires knowledge and reasoning capabilities. Part of the AI2 Reasoning Challenge dataset containing 5,197 questions that test scientific reasoning and factual knowledge. The Easy Set contains questions that are answerable by retrieval-based and word co-occurrence algorithms, making them more accessible than the Challenge Set.", + "metric_id": "llm_stats.arc-e.score", + "metric_name": "ARC-E score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arc-e", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ARC-E", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.88, + "details": { + "raw_score": "0.88", + "raw_score_field": "score", + "raw_model_id": "gemma-2-9b-it", + "raw_benchmark_id": "arc-e", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-2-9b-it\",\"https://llm-stats.com/benchmarks/arc-e\",\"https://api.llm-stats.com/leaderboard/benchmarks/arc-e\"]", + "raw_score_id": "arc-e::gemma-2-9b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "big-bench::big-bench-gemma-2-9b-it", + "evaluation_name": "llm_stats.big-bench", + "source_data": { + "dataset_name": "BIG-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-2-9b-it", + "https://llm-stats.com/benchmarks/big-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench" + ], + "additional_details": { + "raw_benchmark_id": "big-bench", + "raw_model_id": "gemma-2-9b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark consisting of 204+ tasks designed to probe large language models and extrapolate their future capabilities. It covers diverse domains including linguistics, mathematics, common-sense reasoning, biology, physics, social bias, software development, and more. The benchmark focuses on tasks believed to be beyond current language model capabilities and includes both English and non-English tasks across multiple languages.", + "metric_id": "llm_stats.big-bench.score", + "metric_name": "BIG-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.682, + "details": { + "raw_score": "0.682", + "raw_score_field": "score", + "raw_model_id": "gemma-2-9b-it", + "raw_benchmark_id": "big-bench", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-2-9b-it\",\"https://llm-stats.com/benchmarks/big-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench\"]", + "raw_score_id": "big-bench::gemma-2-9b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "boolq::boolq-gemma-2-9b-it", + "evaluation_name": "llm_stats.boolq", + "source_data": { + "dataset_name": "BoolQ", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-2-9b-it", + "https://llm-stats.com/benchmarks/boolq", + "https://api.llm-stats.com/leaderboard/benchmarks/boolq" + ], + "additional_details": { + "raw_benchmark_id": "boolq", + "raw_model_id": "gemma-2-9b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BoolQ is a reading comprehension dataset for yes/no questions containing 15,942 naturally occurring examples. Each example consists of a question, passage, and boolean answer, where questions are generated in unprompted and unconstrained settings. The dataset challenges models with complex, non-factoid information requiring entailment-like inference to solve.", + "metric_id": "llm_stats.boolq.score", + "metric_name": "BoolQ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "boolq", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BoolQ", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.842, + "details": { + "raw_score": "0.842", + "raw_score_field": "score", + "raw_model_id": "gemma-2-9b-it", + "raw_benchmark_id": "boolq", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-2-9b-it\",\"https://llm-stats.com/benchmarks/boolq\",\"https://api.llm-stats.com/leaderboard/benchmarks/boolq\"]", + "raw_score_id": "boolq::gemma-2-9b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hellaswag::hellaswag-gemma-2-9b-it", + "evaluation_name": "llm_stats.hellaswag", + "source_data": { + "dataset_name": "HellaSwag", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-2-9b-it", + "https://llm-stats.com/benchmarks/hellaswag", + "https://api.llm-stats.com/leaderboard/benchmarks/hellaswag" + ], + "additional_details": { + "raw_benchmark_id": "hellaswag", + "raw_model_id": "gemma-2-9b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A challenging commonsense natural language inference dataset that uses Adversarial Filtering to create questions trivial for humans (>95% accuracy) but difficult for state-of-the-art models, requiring completion of sentence endings based on physical situations and everyday activities", + "metric_id": "llm_stats.hellaswag.score", + "metric_name": "HellaSwag score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hellaswag", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HellaSwag", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.819, + "details": { + "raw_score": "0.819", + "raw_score_field": "score", + "raw_model_id": "gemma-2-9b-it", + "raw_benchmark_id": "hellaswag", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-2-9b-it\",\"https://llm-stats.com/benchmarks/hellaswag\",\"https://api.llm-stats.com/leaderboard/benchmarks/hellaswag\"]", + "raw_score_id": "hellaswag::gemma-2-9b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "natural-questions::natural-questions-gemma-2-9b-it", + "evaluation_name": "llm_stats.natural-questions", + "source_data": { + "dataset_name": "Natural Questions", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-2-9b-it", + "https://llm-stats.com/benchmarks/natural-questions", + "https://api.llm-stats.com/leaderboard/benchmarks/natural-questions" + ], + "additional_details": { + "raw_benchmark_id": "natural-questions", + "raw_model_id": "gemma-2-9b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Natural Questions is a question answering dataset featuring real anonymized queries issued to Google search engine. It contains 307,373 training examples where annotators provide long answers (passages) and short answers (entities) from Wikipedia pages, or mark them as unanswerable.", + "metric_id": "llm_stats.natural-questions.score", + "metric_name": "Natural Questions score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "natural-questions", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Natural Questions", + "raw_categories": "[\"reasoning\",\"search\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.292, + "details": { + "raw_score": "0.292", + "raw_score_field": "score", + "raw_model_id": "gemma-2-9b-it", + "raw_benchmark_id": "natural-questions", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-2-9b-it\",\"https://llm-stats.com/benchmarks/natural-questions\",\"https://api.llm-stats.com/leaderboard/benchmarks/natural-questions\"]", + "raw_score_id": "natural-questions::gemma-2-9b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "piqa::piqa-gemma-2-9b-it", + "evaluation_name": "llm_stats.piqa", + "source_data": { + "dataset_name": "PIQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-2-9b-it", + "https://llm-stats.com/benchmarks/piqa", + "https://api.llm-stats.com/leaderboard/benchmarks/piqa" + ], + "additional_details": { + "raw_benchmark_id": "piqa", + "raw_model_id": "gemma-2-9b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "PIQA (Physical Interaction: Question Answering) is a benchmark dataset for physical commonsense reasoning in natural language. It tests AI systems' ability to answer questions requiring physical world knowledge through multiple choice questions with everyday situations, focusing on atypical solutions inspired by instructables.com. The dataset contains 21,000 multiple choice questions where models must choose the most appropriate solution for physical interactions.", + "metric_id": "llm_stats.piqa.score", + "metric_name": "PIQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "piqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PIQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.817, + "details": { + "raw_score": "0.817", + "raw_score_field": "score", + "raw_model_id": "gemma-2-9b-it", + "raw_benchmark_id": "piqa", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-2-9b-it\",\"https://llm-stats.com/benchmarks/piqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/piqa\"]", + "raw_score_id": "piqa::gemma-2-9b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "social-iqa::social-iqa-gemma-2-9b-it", + "evaluation_name": "llm_stats.social-iqa", + "source_data": { + "dataset_name": "Social IQa", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-2-9b-it", + "https://llm-stats.com/benchmarks/social-iqa", + "https://api.llm-stats.com/leaderboard/benchmarks/social-iqa" + ], + "additional_details": { + "raw_benchmark_id": "social-iqa", + "raw_model_id": "gemma-2-9b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The first large-scale benchmark for commonsense reasoning about social situations. Contains 38,000 multiple choice questions probing emotional and social intelligence in everyday situations, testing commonsense understanding of social interactions and theory of mind reasoning about the implied emotions and behavior of others.", + "metric_id": "llm_stats.social-iqa.score", + "metric_name": "Social IQa score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "social-iqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Social IQa", + "raw_categories": "[\"psychology\",\"reasoning\",\"creativity\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.534, + "details": { + "raw_score": "0.534", + "raw_score_field": "score", + "raw_model_id": "gemma-2-9b-it", + "raw_benchmark_id": "social-iqa", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-2-9b-it\",\"https://llm-stats.com/benchmarks/social-iqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/social-iqa\"]", + "raw_score_id": "social-iqa::gemma-2-9b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "triviaqa::triviaqa-gemma-2-9b-it", + "evaluation_name": "llm_stats.triviaqa", + "source_data": { + "dataset_name": "TriviaQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-2-9b-it", + "https://llm-stats.com/benchmarks/triviaqa", + "https://api.llm-stats.com/leaderboard/benchmarks/triviaqa" + ], + "additional_details": { + "raw_benchmark_id": "triviaqa", + "raw_model_id": "gemma-2-9b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A large-scale reading comprehension dataset containing over 650K question-answer-evidence triples. TriviaQA includes 95K question-answer pairs authored by trivia enthusiasts and independently gathered evidence documents (six per question on average) that provide high quality distant supervision for answering the questions. The dataset features relatively complex, compositional questions with considerable syntactic and lexical variability, requiring cross-sentence reasoning to find answers.", + "metric_id": "llm_stats.triviaqa.score", + "metric_name": "TriviaQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "triviaqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TriviaQA", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.766, + "details": { + "raw_score": "0.766", + "raw_score_field": "score", + "raw_model_id": "gemma-2-9b-it", + "raw_benchmark_id": "triviaqa", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-2-9b-it\",\"https://llm-stats.com/benchmarks/triviaqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/triviaqa\"]", + "raw_score_id": "triviaqa::gemma-2-9b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "winogrande::winogrande-gemma-2-9b-it", + "evaluation_name": "llm_stats.winogrande", + "source_data": { + "dataset_name": "Winogrande", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-2-9b-it", + "https://llm-stats.com/benchmarks/winogrande", + "https://api.llm-stats.com/leaderboard/benchmarks/winogrande" + ], + "additional_details": { + "raw_benchmark_id": "winogrande", + "raw_model_id": "gemma-2-9b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WinoGrande: An Adversarial Winograd Schema Challenge at Scale. A large-scale dataset of 44,000 pronoun resolution problems designed to test machine commonsense reasoning. Uses adversarial filtering to reduce spurious biases and provides a more robust evaluation of whether AI systems truly understand commonsense or exploit statistical shortcuts. Current best AI methods achieve 59.4-79.1% accuracy, significantly below human performance of 94.0%.", + "metric_id": "llm_stats.winogrande.score", + "metric_name": "Winogrande score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "winogrande", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Winogrande", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "22" + } + }, + "score_details": { + "score": 0.806, + "details": { + "raw_score": "0.806", + "raw_score_field": "score", + "raw_model_id": "gemma-2-9b-it", + "raw_benchmark_id": "winogrande", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-2-9b-it\",\"https://llm-stats.com/benchmarks/winogrande\",\"https://api.llm-stats.com/leaderboard/benchmarks/winogrande\"]", + "raw_score_id": "winogrande::gemma-2-9b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/google/gemma-3-12b-it/85f235db-d906-472d-bc8d-44ff735a3de1.json b/data/llm_stats/google/gemma-3-12b-it/85f235db-d906-472d-bc8d-44ff735a3de1.json new file mode 100644 index 0000000000000000000000000000000000000000..a916d0ba7f50b9426807f9e14a647f676be00655 --- /dev/null +++ b/data/llm_stats/google/gemma-3-12b-it/85f235db-d906-472d-bc8d-44ff735a3de1.json @@ -0,0 +1,827 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/google_gemma-3-12b-it/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Gemma 3 12B", + "id": "google/gemma-3-12b-it", + "developer": "google", + "additional_details": { + "raw_id": "gemma-3-12b-it", + "raw_name": "Gemma 3 12B", + "raw_model_id": "gemma-3-12b-it", + "raw_model_name": "Gemma 3 12B", + "raw_organization_id": "google", + "raw_organization_name": "Google", + "raw_release_date": "2025-03-12", + "raw_announcement_date": "2025-03-12", + "raw_multimodal": "true", + "raw_param_count": "12000000000", + "raw_provider_slug": "google", + "raw_provider_name": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "big-bench-extra-hard::big-bench-extra-hard-gemma-3-12b-it", + "evaluation_name": "llm_stats.big-bench-extra-hard", + "source_data": { + "dataset_name": "BIG-Bench Extra Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-12b-it", + "https://llm-stats.com/benchmarks/big-bench-extra-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench-extra-hard" + ], + "additional_details": { + "raw_benchmark_id": "big-bench-extra-hard", + "raw_model_id": "gemma-3-12b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIG-Bench Extra Hard (BBEH) is a challenging benchmark that replaces each task in BIG-Bench Hard with a novel task that probes similar reasoning capabilities but exhibits significantly increased difficulty. The benchmark contains 23 tasks testing diverse reasoning skills including many-hop reasoning, causal understanding, spatial reasoning, temporal arithmetic, geometric reasoning, linguistic reasoning, logic puzzles, and humor understanding. Designed to address saturation on existing benchmarks where state-of-the-art models achieve near-perfect scores, BBEH shows substantial room for improvement with best models achieving only 9.8-44.8% average accuracy.", + "metric_id": "llm_stats.big-bench-extra-hard.score", + "metric_name": "BIG-Bench Extra Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench-extra-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench Extra Hard", + "raw_categories": "[\"language\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.163, + "details": { + "raw_score": "0.163", + "raw_score_field": "score", + "raw_model_id": "gemma-3-12b-it", + "raw_benchmark_id": "big-bench-extra-hard", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-12b-it\",\"https://llm-stats.com/benchmarks/big-bench-extra-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench-extra-hard\"]", + "raw_score_id": "big-bench-extra-hard::gemma-3-12b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "big-bench-hard::big-bench-hard-gemma-3-12b-it", + "evaluation_name": "llm_stats.big-bench-hard", + "source_data": { + "dataset_name": "BIG-Bench Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-12b-it", + "https://llm-stats.com/benchmarks/big-bench-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard" + ], + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_model_id": "gemma-3-12b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIG-Bench Hard (BBH) is a subset of 23 challenging BIG-Bench tasks selected because prior language model evaluations did not outperform average human-rater performance. The benchmark contains 6,511 evaluation examples testing various forms of multi-step reasoning including arithmetic, logical reasoning (Boolean expressions, logical deduction), geometric reasoning, temporal reasoning, and language understanding. Tasks require capabilities such as causal judgment, object counting, navigation, pattern recognition, and complex problem solving.", + "metric_id": "llm_stats.big-bench-hard.score", + "metric_name": "BIG-Bench Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench Hard", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "21" + } + }, + "score_details": { + "score": 0.857, + "details": { + "raw_score": "0.857", + "raw_score_field": "score", + "raw_model_id": "gemma-3-12b-it", + "raw_benchmark_id": "big-bench-hard", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-12b-it\",\"https://llm-stats.com/benchmarks/big-bench-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard\"]", + "raw_score_id": "big-bench-hard::gemma-3-12b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bird-sql-dev::bird-sql-dev-gemma-3-12b-it", + "evaluation_name": "llm_stats.bird-sql-dev", + "source_data": { + "dataset_name": "Bird-SQL (dev)", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-12b-it", + "https://llm-stats.com/benchmarks/bird-sql-dev", + "https://api.llm-stats.com/leaderboard/benchmarks/bird-sql-(dev)" + ], + "additional_details": { + "raw_benchmark_id": "bird-sql-(dev)", + "raw_model_id": "gemma-3-12b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIRD (BIg Bench for LaRge-scale Database Grounded Text-to-SQLs) is a comprehensive text-to-SQL benchmark containing 12,751 question-SQL pairs across 95 databases (33.4 GB total) spanning 37+ professional domains. It evaluates large language models' ability to convert natural language to executable SQL queries in real-world scenarios with complex database schemas and dirty data.", + "metric_id": "llm_stats.bird-sql-dev.score", + "metric_name": "Bird-SQL (dev) score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bird-sql-(dev)", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Bird-SQL (dev)", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.479, + "details": { + "raw_score": "0.479", + "raw_score_field": "score", + "raw_model_id": "gemma-3-12b-it", + "raw_benchmark_id": "bird-sql-(dev)", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-12b-it\",\"https://llm-stats.com/benchmarks/bird-sql-dev\",\"https://api.llm-stats.com/leaderboard/benchmarks/bird-sql-(dev)\"]", + "raw_score_id": "bird-sql-(dev)::gemma-3-12b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "eclektic::eclektic-gemma-3-12b-it", + "evaluation_name": "llm_stats.eclektic", + "source_data": { + "dataset_name": "ECLeKTic", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-12b-it", + "https://llm-stats.com/benchmarks/eclektic", + "https://api.llm-stats.com/leaderboard/benchmarks/eclektic" + ], + "additional_details": { + "raw_benchmark_id": "eclektic", + "raw_model_id": "gemma-3-12b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multilingual closed-book question answering dataset that evaluates cross-lingual knowledge transfer in large language models across 12 languages, using knowledge-seeking questions based on Wikipedia articles that exist only in one language", + "metric_id": "llm_stats.eclektic.score", + "metric_name": "ECLeKTic score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "eclektic", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ECLeKTic", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.103, + "details": { + "raw_score": "0.103", + "raw_score_field": "score", + "raw_model_id": "gemma-3-12b-it", + "raw_benchmark_id": "eclektic", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-12b-it\",\"https://llm-stats.com/benchmarks/eclektic\",\"https://api.llm-stats.com/leaderboard/benchmarks/eclektic\"]", + "raw_score_id": "eclektic::gemma-3-12b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "facts-grounding::facts-grounding-gemma-3-12b-it", + "evaluation_name": "llm_stats.facts-grounding", + "source_data": { + "dataset_name": "FACTS Grounding", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-12b-it", + "https://llm-stats.com/benchmarks/facts-grounding", + "https://api.llm-stats.com/leaderboard/benchmarks/facts-grounding" + ], + "additional_details": { + "raw_benchmark_id": "facts-grounding", + "raw_model_id": "gemma-3-12b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A benchmark evaluating language models' ability to generate factually accurate and well-grounded responses based on long-form input context, comprising 1,719 examples with documents up to 32k tokens requiring detailed responses that are fully grounded in provided documents", + "metric_id": "llm_stats.facts-grounding.score", + "metric_name": "FACTS Grounding score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "facts-grounding", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FACTS Grounding", + "raw_categories": "[\"reasoning\",\"factuality\",\"grounding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.758, + "details": { + "raw_score": "0.758", + "raw_score_field": "score", + "raw_model_id": "gemma-3-12b-it", + "raw_benchmark_id": "facts-grounding", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-12b-it\",\"https://llm-stats.com/benchmarks/facts-grounding\",\"https://api.llm-stats.com/leaderboard/benchmarks/facts-grounding\"]", + "raw_score_id": "facts-grounding::gemma-3-12b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "global-mmlu-lite::global-mmlu-lite-gemma-3-12b-it", + "evaluation_name": "llm_stats.global-mmlu-lite", + "source_data": { + "dataset_name": "Global-MMLU-Lite", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-12b-it", + "https://llm-stats.com/benchmarks/global-mmlu-lite", + "https://api.llm-stats.com/leaderboard/benchmarks/global-mmlu-lite" + ], + "additional_details": { + "raw_benchmark_id": "global-mmlu-lite", + "raw_model_id": "gemma-3-12b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A lightweight version of Global MMLU benchmark that evaluates language models across multiple languages while addressing cultural and linguistic biases in multilingual evaluation.", + "metric_id": "llm_stats.global-mmlu-lite.score", + "metric_name": "Global-MMLU-Lite score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "global-mmlu-lite", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Global-MMLU-Lite", + "raw_categories": "[\"language\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "14" + } + }, + "score_details": { + "score": 0.695, + "details": { + "raw_score": "0.695", + "raw_score_field": "score", + "raw_model_id": "gemma-3-12b-it", + "raw_benchmark_id": "global-mmlu-lite", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-12b-it\",\"https://llm-stats.com/benchmarks/global-mmlu-lite\",\"https://api.llm-stats.com/leaderboard/benchmarks/global-mmlu-lite\"]", + "raw_score_id": "global-mmlu-lite::gemma-3-12b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gsm8k::gsm8k-gemma-3-12b-it", + "evaluation_name": "llm_stats.gsm8k", + "source_data": { + "dataset_name": "GSM8k", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-12b-it", + "https://llm-stats.com/benchmarks/gsm8k", + "https://api.llm-stats.com/leaderboard/benchmarks/gsm8k" + ], + "additional_details": { + "raw_benchmark_id": "gsm8k", + "raw_model_id": "gemma-3-12b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Grade School Math 8K, a dataset of 8.5K high-quality linguistically diverse grade school math word problems requiring multi-step reasoning and elementary arithmetic operations.", + "metric_id": "llm_stats.gsm8k.score", + "metric_name": "GSM8k score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gsm8k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GSM8k", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "48" + } + }, + "score_details": { + "score": 0.944, + "details": { + "raw_score": "0.944", + "raw_score_field": "score", + "raw_model_id": "gemma-3-12b-it", + "raw_benchmark_id": "gsm8k", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-12b-it\",\"https://llm-stats.com/benchmarks/gsm8k\",\"https://api.llm-stats.com/leaderboard/benchmarks/gsm8k\"]", + "raw_score_id": "gsm8k::gemma-3-12b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hiddenmath::hiddenmath-gemma-3-12b-it", + "evaluation_name": "llm_stats.hiddenmath", + "source_data": { + "dataset_name": "HiddenMath", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-12b-it", + "https://llm-stats.com/benchmarks/hiddenmath", + "https://api.llm-stats.com/leaderboard/benchmarks/hiddenmath" + ], + "additional_details": { + "raw_benchmark_id": "hiddenmath", + "raw_model_id": "gemma-3-12b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Google DeepMind's internal mathematical reasoning benchmark that introduces novel problems not encountered during model training to evaluate true mathematical reasoning capabilities rather than memorization", + "metric_id": "llm_stats.hiddenmath.score", + "metric_name": "HiddenMath score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hiddenmath", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HiddenMath", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.545, + "details": { + "raw_score": "0.545", + "raw_score_field": "score", + "raw_model_id": "gemma-3-12b-it", + "raw_benchmark_id": "hiddenmath", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-12b-it\",\"https://llm-stats.com/benchmarks/hiddenmath\",\"https://api.llm-stats.com/leaderboard/benchmarks/hiddenmath\"]", + "raw_score_id": "hiddenmath::gemma-3-12b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "infovqa::infovqa-gemma-3-12b-it", + "evaluation_name": "llm_stats.infovqa", + "source_data": { + "dataset_name": "InfoVQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-12b-it", + "https://llm-stats.com/benchmarks/infovqa", + "https://api.llm-stats.com/leaderboard/benchmarks/infovqa" + ], + "additional_details": { + "raw_benchmark_id": "infovqa", + "raw_model_id": "gemma-3-12b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "InfoVQA dataset with 30,000 questions and 5,000 infographic images requiring joint reasoning over document layout, textual content, graphical elements, and data visualizations with elementary reasoning and arithmetic skills", + "metric_id": "llm_stats.infovqa.score", + "metric_name": "InfoVQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "infovqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "InfoVQA", + "raw_categories": "[\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.649, + "details": { + "raw_score": "0.649", + "raw_score_field": "score", + "raw_model_id": "gemma-3-12b-it", + "raw_benchmark_id": "infovqa", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-12b-it\",\"https://llm-stats.com/benchmarks/infovqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/infovqa\"]", + "raw_score_id": "infovqa::gemma-3-12b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "math::math-gemma-3-12b-it", + "evaluation_name": "llm_stats.math", + "source_data": { + "dataset_name": "MATH", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-12b-it", + "https://llm-stats.com/benchmarks/math", + "https://api.llm-stats.com/leaderboard/benchmarks/math" + ], + "additional_details": { + "raw_benchmark_id": "math", + "raw_model_id": "gemma-3-12b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MATH dataset contains 12,500 challenging competition mathematics problems from AMC 10, AMC 12, AIME, and other mathematics competitions. Each problem includes full step-by-step solutions and spans multiple difficulty levels (1-5) across seven mathematical subjects including Prealgebra, Algebra, Number Theory, Counting and Probability, Geometry, Intermediate Algebra, and Precalculus.", + "metric_id": "llm_stats.math.score", + "metric_name": "MATH score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "math", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MATH", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "71" + } + }, + "score_details": { + "score": 0.838, + "details": { + "raw_score": "0.838", + "raw_score_field": "score", + "raw_model_id": "gemma-3-12b-it", + "raw_benchmark_id": "math", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-12b-it\",\"https://llm-stats.com/benchmarks/math\",\"https://api.llm-stats.com/leaderboard/benchmarks/math\"]", + "raw_score_id": "math::gemma-3-12b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmmu-val::mmmu-val-gemma-3-12b-it", + "evaluation_name": "llm_stats.mmmu-val", + "source_data": { + "dataset_name": "MMMU (val)", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-12b-it", + "https://llm-stats.com/benchmarks/mmmu-val", + "https://api.llm-stats.com/leaderboard/benchmarks/mmmu-(val)" + ], + "additional_details": { + "raw_benchmark_id": "mmmu-(val)", + "raw_model_id": "gemma-3-12b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Validation set of the Massive Multi-discipline Multimodal Understanding and Reasoning benchmark. Features college-level multimodal questions across 6 core disciplines (Art & Design, Business, Science, Health & Medicine, Humanities & Social Science, Tech & Engineering) spanning 30 subjects and 183 subfields with diverse image types including charts, diagrams, maps, and tables.", + "metric_id": "llm_stats.mmmu-val.score", + "metric_name": "MMMU (val) score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmmu-(val)", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMMU (val)", + "raw_categories": "[\"multimodal\",\"reasoning\",\"general\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.596, + "details": { + "raw_score": "0.596", + "raw_score_field": "score", + "raw_model_id": "gemma-3-12b-it", + "raw_benchmark_id": "mmmu-(val)", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-12b-it\",\"https://llm-stats.com/benchmarks/mmmu-val\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmmu-(val)\"]", + "raw_score_id": "mmmu-(val)::gemma-3-12b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "natural2code::natural2code-gemma-3-12b-it", + "evaluation_name": "llm_stats.natural2code", + "source_data": { + "dataset_name": "Natural2Code", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-12b-it", + "https://llm-stats.com/benchmarks/natural2code", + "https://api.llm-stats.com/leaderboard/benchmarks/natural2code" + ], + "additional_details": { + "raw_benchmark_id": "natural2code", + "raw_model_id": "gemma-3-12b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "NaturalCodeBench (NCB) is a challenging code benchmark designed to mirror the complexity and variety of real-world coding tasks. It comprises 402 high-quality problems in Python and Java, selected from natural user queries from online coding services, covering 6 different domains.", + "metric_id": "llm_stats.natural2code.score", + "metric_name": "Natural2Code score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "natural2code", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Natural2Code", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.807, + "details": { + "raw_score": "0.807", + "raw_score_field": "score", + "raw_model_id": "gemma-3-12b-it", + "raw_benchmark_id": "natural2code", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-12b-it\",\"https://llm-stats.com/benchmarks/natural2code\",\"https://api.llm-stats.com/leaderboard/benchmarks/natural2code\"]", + "raw_score_id": "natural2code::gemma-3-12b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "textvqa::textvqa-gemma-3-12b-it", + "evaluation_name": "llm_stats.textvqa", + "source_data": { + "dataset_name": "TextVQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-12b-it", + "https://llm-stats.com/benchmarks/textvqa", + "https://api.llm-stats.com/leaderboard/benchmarks/textvqa" + ], + "additional_details": { + "raw_benchmark_id": "textvqa", + "raw_model_id": "gemma-3-12b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "TextVQA contains 45,336 questions on 28,408 images that require reasoning about text to answer. Introduced to benchmark VQA models' ability to read and reason about text within images, particularly for assistive technologies for visually impaired users. The dataset addresses the gap where existing VQA datasets had few text-based questions or were too small.", + "metric_id": "llm_stats.textvqa.score", + "metric_name": "TextVQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "textvqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TextVQA", + "raw_categories": "[\"image_to_text\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "15" + } + }, + "score_details": { + "score": 0.677, + "details": { + "raw_score": "0.677", + "raw_score_field": "score", + "raw_model_id": "gemma-3-12b-it", + "raw_benchmark_id": "textvqa", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-12b-it\",\"https://llm-stats.com/benchmarks/textvqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/textvqa\"]", + "raw_score_id": "textvqa::gemma-3-12b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "vqav2-val::vqav2-val-gemma-3-12b-it", + "evaluation_name": "llm_stats.vqav2-val", + "source_data": { + "dataset_name": "VQAv2 (val)", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-12b-it", + "https://llm-stats.com/benchmarks/vqav2-val", + "https://api.llm-stats.com/leaderboard/benchmarks/vqav2-(val)" + ], + "additional_details": { + "raw_benchmark_id": "vqav2-(val)", + "raw_model_id": "gemma-3-12b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VQAv2 is a balanced Visual Question Answering dataset containing open-ended questions about images that require understanding of vision, language, and commonsense knowledge to answer. VQAv2 addresses bias issues from the original VQA dataset by collecting complementary images such that every question is associated with similar images that result in different answers, forcing models to actually understand visual content rather than relying on language priors.", + "metric_id": "llm_stats.vqav2-val.score", + "metric_name": "VQAv2 (val) score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "vqav2-(val)", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VQAv2 (val)", + "raw_categories": "[\"image_to_text\",\"language\",\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.716, + "details": { + "raw_score": "0.716", + "raw_score_field": "score", + "raw_model_id": "gemma-3-12b-it", + "raw_benchmark_id": "vqav2-(val)", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-12b-it\",\"https://llm-stats.com/benchmarks/vqav2-val\",\"https://api.llm-stats.com/leaderboard/benchmarks/vqav2-(val)\"]", + "raw_score_id": "vqav2-(val)::gemma-3-12b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "wmt24::wmt24-gemma-3-12b-it", + "evaluation_name": "llm_stats.wmt24", + "source_data": { + "dataset_name": "WMT24++", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-12b-it", + "https://llm-stats.com/benchmarks/wmt24", + "https://api.llm-stats.com/leaderboard/benchmarks/wmt24++" + ], + "additional_details": { + "raw_benchmark_id": "wmt24++", + "raw_model_id": "gemma-3-12b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WMT24++ is a comprehensive multilingual machine translation benchmark that expands the WMT24 dataset to cover 55 languages and dialects. It includes human-written references and post-edits across four domains (literary, news, social, and speech) to evaluate machine translation systems and large language models across diverse linguistic contexts.", + "metric_id": "llm_stats.wmt24.score", + "metric_name": "WMT24++ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "wmt24++", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WMT24++", + "raw_categories": "[\"language\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.516, + "details": { + "raw_score": "0.516", + "raw_score_field": "score", + "raw_model_id": "gemma-3-12b-it", + "raw_benchmark_id": "wmt24++", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-12b-it\",\"https://llm-stats.com/benchmarks/wmt24\",\"https://api.llm-stats.com/leaderboard/benchmarks/wmt24++\"]", + "raw_score_id": "wmt24++::gemma-3-12b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/google/gemma-3-1b-it/b589669a-eb2c-4de9-bd1e-88f0aba0b95d.json b/data/llm_stats/google/gemma-3-1b-it/b589669a-eb2c-4de9-bd1e-88f0aba0b95d.json new file mode 100644 index 0000000000000000000000000000000000000000..7fb392c56cd6dc998f41cb8ecef24c4777cbeecc --- /dev/null +++ b/data/llm_stats/google/gemma-3-1b-it/b589669a-eb2c-4de9-bd1e-88f0aba0b95d.json @@ -0,0 +1,411 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/google_gemma-3-1b-it/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Gemma 3 1B", + "id": "google/gemma-3-1b-it", + "developer": "google", + "additional_details": { + "raw_id": "gemma-3-1b-it", + "raw_name": "Gemma 3 1B", + "raw_model_id": "gemma-3-1b-it", + "raw_model_name": "Gemma 3 1B", + "raw_organization_id": "google", + "raw_organization_name": "Google", + "raw_release_date": "2025-03-12", + "raw_announcement_date": "2025-03-12", + "raw_multimodal": "false", + "raw_param_count": "1000000000", + "raw_provider_slug": "google", + "raw_provider_name": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "big-bench-extra-hard::big-bench-extra-hard-gemma-3-1b-it", + "evaluation_name": "llm_stats.big-bench-extra-hard", + "source_data": { + "dataset_name": "BIG-Bench Extra Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-1b-it", + "https://llm-stats.com/benchmarks/big-bench-extra-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench-extra-hard" + ], + "additional_details": { + "raw_benchmark_id": "big-bench-extra-hard", + "raw_model_id": "gemma-3-1b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIG-Bench Extra Hard (BBEH) is a challenging benchmark that replaces each task in BIG-Bench Hard with a novel task that probes similar reasoning capabilities but exhibits significantly increased difficulty. The benchmark contains 23 tasks testing diverse reasoning skills including many-hop reasoning, causal understanding, spatial reasoning, temporal arithmetic, geometric reasoning, linguistic reasoning, logic puzzles, and humor understanding. Designed to address saturation on existing benchmarks where state-of-the-art models achieve near-perfect scores, BBEH shows substantial room for improvement with best models achieving only 9.8-44.8% average accuracy.", + "metric_id": "llm_stats.big-bench-extra-hard.score", + "metric_name": "BIG-Bench Extra Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench-extra-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench Extra Hard", + "raw_categories": "[\"language\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.072, + "details": { + "raw_score": "0.072", + "raw_score_field": "score", + "raw_model_id": "gemma-3-1b-it", + "raw_benchmark_id": "big-bench-extra-hard", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-1b-it\",\"https://llm-stats.com/benchmarks/big-bench-extra-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench-extra-hard\"]", + "raw_score_id": "big-bench-extra-hard::gemma-3-1b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bird-sql-dev::bird-sql-dev-gemma-3-1b-it", + "evaluation_name": "llm_stats.bird-sql-dev", + "source_data": { + "dataset_name": "Bird-SQL (dev)", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-1b-it", + "https://llm-stats.com/benchmarks/bird-sql-dev", + "https://api.llm-stats.com/leaderboard/benchmarks/bird-sql-(dev)" + ], + "additional_details": { + "raw_benchmark_id": "bird-sql-(dev)", + "raw_model_id": "gemma-3-1b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIRD (BIg Bench for LaRge-scale Database Grounded Text-to-SQLs) is a comprehensive text-to-SQL benchmark containing 12,751 question-SQL pairs across 95 databases (33.4 GB total) spanning 37+ professional domains. It evaluates large language models' ability to convert natural language to executable SQL queries in real-world scenarios with complex database schemas and dirty data.", + "metric_id": "llm_stats.bird-sql-dev.score", + "metric_name": "Bird-SQL (dev) score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bird-sql-(dev)", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Bird-SQL (dev)", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.064, + "details": { + "raw_score": "0.064", + "raw_score_field": "score", + "raw_model_id": "gemma-3-1b-it", + "raw_benchmark_id": "bird-sql-(dev)", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-1b-it\",\"https://llm-stats.com/benchmarks/bird-sql-dev\",\"https://api.llm-stats.com/leaderboard/benchmarks/bird-sql-(dev)\"]", + "raw_score_id": "bird-sql-(dev)::gemma-3-1b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "eclektic::eclektic-gemma-3-1b-it", + "evaluation_name": "llm_stats.eclektic", + "source_data": { + "dataset_name": "ECLeKTic", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-1b-it", + "https://llm-stats.com/benchmarks/eclektic", + "https://api.llm-stats.com/leaderboard/benchmarks/eclektic" + ], + "additional_details": { + "raw_benchmark_id": "eclektic", + "raw_model_id": "gemma-3-1b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multilingual closed-book question answering dataset that evaluates cross-lingual knowledge transfer in large language models across 12 languages, using knowledge-seeking questions based on Wikipedia articles that exist only in one language", + "metric_id": "llm_stats.eclektic.score", + "metric_name": "ECLeKTic score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "eclektic", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ECLeKTic", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.014, + "details": { + "raw_score": "0.014", + "raw_score_field": "score", + "raw_model_id": "gemma-3-1b-it", + "raw_benchmark_id": "eclektic", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-1b-it\",\"https://llm-stats.com/benchmarks/eclektic\",\"https://api.llm-stats.com/leaderboard/benchmarks/eclektic\"]", + "raw_score_id": "eclektic::gemma-3-1b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "facts-grounding::facts-grounding-gemma-3-1b-it", + "evaluation_name": "llm_stats.facts-grounding", + "source_data": { + "dataset_name": "FACTS Grounding", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-1b-it", + "https://llm-stats.com/benchmarks/facts-grounding", + "https://api.llm-stats.com/leaderboard/benchmarks/facts-grounding" + ], + "additional_details": { + "raw_benchmark_id": "facts-grounding", + "raw_model_id": "gemma-3-1b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A benchmark evaluating language models' ability to generate factually accurate and well-grounded responses based on long-form input context, comprising 1,719 examples with documents up to 32k tokens requiring detailed responses that are fully grounded in provided documents", + "metric_id": "llm_stats.facts-grounding.score", + "metric_name": "FACTS Grounding score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "facts-grounding", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FACTS Grounding", + "raw_categories": "[\"reasoning\",\"factuality\",\"grounding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.364, + "details": { + "raw_score": "0.364", + "raw_score_field": "score", + "raw_model_id": "gemma-3-1b-it", + "raw_benchmark_id": "facts-grounding", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-1b-it\",\"https://llm-stats.com/benchmarks/facts-grounding\",\"https://api.llm-stats.com/leaderboard/benchmarks/facts-grounding\"]", + "raw_score_id": "facts-grounding::gemma-3-1b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "global-mmlu-lite::global-mmlu-lite-gemma-3-1b-it", + "evaluation_name": "llm_stats.global-mmlu-lite", + "source_data": { + "dataset_name": "Global-MMLU-Lite", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-1b-it", + "https://llm-stats.com/benchmarks/global-mmlu-lite", + "https://api.llm-stats.com/leaderboard/benchmarks/global-mmlu-lite" + ], + "additional_details": { + "raw_benchmark_id": "global-mmlu-lite", + "raw_model_id": "gemma-3-1b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A lightweight version of Global MMLU benchmark that evaluates language models across multiple languages while addressing cultural and linguistic biases in multilingual evaluation.", + "metric_id": "llm_stats.global-mmlu-lite.score", + "metric_name": "Global-MMLU-Lite score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "global-mmlu-lite", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Global-MMLU-Lite", + "raw_categories": "[\"language\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "14" + } + }, + "score_details": { + "score": 0.342, + "details": { + "raw_score": "0.342", + "raw_score_field": "score", + "raw_model_id": "gemma-3-1b-it", + "raw_benchmark_id": "global-mmlu-lite", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-1b-it\",\"https://llm-stats.com/benchmarks/global-mmlu-lite\",\"https://api.llm-stats.com/leaderboard/benchmarks/global-mmlu-lite\"]", + "raw_score_id": "global-mmlu-lite::gemma-3-1b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hiddenmath::hiddenmath-gemma-3-1b-it", + "evaluation_name": "llm_stats.hiddenmath", + "source_data": { + "dataset_name": "HiddenMath", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-1b-it", + "https://llm-stats.com/benchmarks/hiddenmath", + "https://api.llm-stats.com/leaderboard/benchmarks/hiddenmath" + ], + "additional_details": { + "raw_benchmark_id": "hiddenmath", + "raw_model_id": "gemma-3-1b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Google DeepMind's internal mathematical reasoning benchmark that introduces novel problems not encountered during model training to evaluate true mathematical reasoning capabilities rather than memorization", + "metric_id": "llm_stats.hiddenmath.score", + "metric_name": "HiddenMath score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hiddenmath", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HiddenMath", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.158, + "details": { + "raw_score": "0.158", + "raw_score_field": "score", + "raw_model_id": "gemma-3-1b-it", + "raw_benchmark_id": "hiddenmath", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-1b-it\",\"https://llm-stats.com/benchmarks/hiddenmath\",\"https://api.llm-stats.com/leaderboard/benchmarks/hiddenmath\"]", + "raw_score_id": "hiddenmath::gemma-3-1b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "natural2code::natural2code-gemma-3-1b-it", + "evaluation_name": "llm_stats.natural2code", + "source_data": { + "dataset_name": "Natural2Code", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-1b-it", + "https://llm-stats.com/benchmarks/natural2code", + "https://api.llm-stats.com/leaderboard/benchmarks/natural2code" + ], + "additional_details": { + "raw_benchmark_id": "natural2code", + "raw_model_id": "gemma-3-1b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "NaturalCodeBench (NCB) is a challenging code benchmark designed to mirror the complexity and variety of real-world coding tasks. It comprises 402 high-quality problems in Python and Java, selected from natural user queries from online coding services, covering 6 different domains.", + "metric_id": "llm_stats.natural2code.score", + "metric_name": "Natural2Code score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "natural2code", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Natural2Code", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.56, + "details": { + "raw_score": "0.56", + "raw_score_field": "score", + "raw_model_id": "gemma-3-1b-it", + "raw_benchmark_id": "natural2code", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-1b-it\",\"https://llm-stats.com/benchmarks/natural2code\",\"https://api.llm-stats.com/leaderboard/benchmarks/natural2code\"]", + "raw_score_id": "natural2code::gemma-3-1b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/google/gemma-3-27b-it/848817d5-bb60-4f2c-a6d3-17b0e9ac1293.json b/data/llm_stats/google/gemma-3-27b-it/848817d5-bb60-4f2c-a6d3-17b0e9ac1293.json new file mode 100644 index 0000000000000000000000000000000000000000..e84e7fc135c760282104a8dfcc69d9bd381d0af9 --- /dev/null +++ b/data/llm_stats/google/gemma-3-27b-it/848817d5-bb60-4f2c-a6d3-17b0e9ac1293.json @@ -0,0 +1,931 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/google_gemma-3-27b-it/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Gemma 3 27B", + "id": "google/gemma-3-27b-it", + "developer": "google", + "additional_details": { + "raw_id": "gemma-3-27b-it", + "raw_name": "Gemma 3 27B", + "raw_model_id": "gemma-3-27b-it", + "raw_model_name": "Gemma 3 27B", + "raw_organization_id": "google", + "raw_organization_name": "Google", + "raw_release_date": "2025-03-12", + "raw_announcement_date": "2025-03-12", + "raw_multimodal": "true", + "raw_param_count": "27000000000", + "raw_provider_slug": "google", + "raw_provider_name": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "big-bench-extra-hard::big-bench-extra-hard-gemma-3-27b-it", + "evaluation_name": "llm_stats.big-bench-extra-hard", + "source_data": { + "dataset_name": "BIG-Bench Extra Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-27b-it", + "https://llm-stats.com/benchmarks/big-bench-extra-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench-extra-hard" + ], + "additional_details": { + "raw_benchmark_id": "big-bench-extra-hard", + "raw_model_id": "gemma-3-27b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIG-Bench Extra Hard (BBEH) is a challenging benchmark that replaces each task in BIG-Bench Hard with a novel task that probes similar reasoning capabilities but exhibits significantly increased difficulty. The benchmark contains 23 tasks testing diverse reasoning skills including many-hop reasoning, causal understanding, spatial reasoning, temporal arithmetic, geometric reasoning, linguistic reasoning, logic puzzles, and humor understanding. Designed to address saturation on existing benchmarks where state-of-the-art models achieve near-perfect scores, BBEH shows substantial room for improvement with best models achieving only 9.8-44.8% average accuracy.", + "metric_id": "llm_stats.big-bench-extra-hard.score", + "metric_name": "BIG-Bench Extra Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench-extra-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench Extra Hard", + "raw_categories": "[\"language\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.193, + "details": { + "raw_score": "0.193", + "raw_score_field": "score", + "raw_model_id": "gemma-3-27b-it", + "raw_benchmark_id": "big-bench-extra-hard", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-27b-it\",\"https://llm-stats.com/benchmarks/big-bench-extra-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench-extra-hard\"]", + "raw_score_id": "big-bench-extra-hard::gemma-3-27b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "big-bench-hard::big-bench-hard-gemma-3-27b-it", + "evaluation_name": "llm_stats.big-bench-hard", + "source_data": { + "dataset_name": "BIG-Bench Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-27b-it", + "https://llm-stats.com/benchmarks/big-bench-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard" + ], + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_model_id": "gemma-3-27b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIG-Bench Hard (BBH) is a subset of 23 challenging BIG-Bench tasks selected because prior language model evaluations did not outperform average human-rater performance. The benchmark contains 6,511 evaluation examples testing various forms of multi-step reasoning including arithmetic, logical reasoning (Boolean expressions, logical deduction), geometric reasoning, temporal reasoning, and language understanding. Tasks require capabilities such as causal judgment, object counting, navigation, pattern recognition, and complex problem solving.", + "metric_id": "llm_stats.big-bench-hard.score", + "metric_name": "BIG-Bench Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench Hard", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "21" + } + }, + "score_details": { + "score": 0.876, + "details": { + "raw_score": "0.876", + "raw_score_field": "score", + "raw_model_id": "gemma-3-27b-it", + "raw_benchmark_id": "big-bench-hard", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-27b-it\",\"https://llm-stats.com/benchmarks/big-bench-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard\"]", + "raw_score_id": "big-bench-hard::gemma-3-27b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bird-sql-dev::bird-sql-dev-gemma-3-27b-it", + "evaluation_name": "llm_stats.bird-sql-dev", + "source_data": { + "dataset_name": "Bird-SQL (dev)", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-27b-it", + "https://llm-stats.com/benchmarks/bird-sql-dev", + "https://api.llm-stats.com/leaderboard/benchmarks/bird-sql-(dev)" + ], + "additional_details": { + "raw_benchmark_id": "bird-sql-(dev)", + "raw_model_id": "gemma-3-27b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIRD (BIg Bench for LaRge-scale Database Grounded Text-to-SQLs) is a comprehensive text-to-SQL benchmark containing 12,751 question-SQL pairs across 95 databases (33.4 GB total) spanning 37+ professional domains. It evaluates large language models' ability to convert natural language to executable SQL queries in real-world scenarios with complex database schemas and dirty data.", + "metric_id": "llm_stats.bird-sql-dev.score", + "metric_name": "Bird-SQL (dev) score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bird-sql-(dev)", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Bird-SQL (dev)", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.544, + "details": { + "raw_score": "0.544", + "raw_score_field": "score", + "raw_model_id": "gemma-3-27b-it", + "raw_benchmark_id": "bird-sql-(dev)", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-27b-it\",\"https://llm-stats.com/benchmarks/bird-sql-dev\",\"https://api.llm-stats.com/leaderboard/benchmarks/bird-sql-(dev)\"]", + "raw_score_id": "bird-sql-(dev)::gemma-3-27b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "eclektic::eclektic-gemma-3-27b-it", + "evaluation_name": "llm_stats.eclektic", + "source_data": { + "dataset_name": "ECLeKTic", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-27b-it", + "https://llm-stats.com/benchmarks/eclektic", + "https://api.llm-stats.com/leaderboard/benchmarks/eclektic" + ], + "additional_details": { + "raw_benchmark_id": "eclektic", + "raw_model_id": "gemma-3-27b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multilingual closed-book question answering dataset that evaluates cross-lingual knowledge transfer in large language models across 12 languages, using knowledge-seeking questions based on Wikipedia articles that exist only in one language", + "metric_id": "llm_stats.eclektic.score", + "metric_name": "ECLeKTic score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "eclektic", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ECLeKTic", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.167, + "details": { + "raw_score": "0.167", + "raw_score_field": "score", + "raw_model_id": "gemma-3-27b-it", + "raw_benchmark_id": "eclektic", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-27b-it\",\"https://llm-stats.com/benchmarks/eclektic\",\"https://api.llm-stats.com/leaderboard/benchmarks/eclektic\"]", + "raw_score_id": "eclektic::gemma-3-27b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "facts-grounding::facts-grounding-gemma-3-27b-it", + "evaluation_name": "llm_stats.facts-grounding", + "source_data": { + "dataset_name": "FACTS Grounding", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-27b-it", + "https://llm-stats.com/benchmarks/facts-grounding", + "https://api.llm-stats.com/leaderboard/benchmarks/facts-grounding" + ], + "additional_details": { + "raw_benchmark_id": "facts-grounding", + "raw_model_id": "gemma-3-27b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A benchmark evaluating language models' ability to generate factually accurate and well-grounded responses based on long-form input context, comprising 1,719 examples with documents up to 32k tokens requiring detailed responses that are fully grounded in provided documents", + "metric_id": "llm_stats.facts-grounding.score", + "metric_name": "FACTS Grounding score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "facts-grounding", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FACTS Grounding", + "raw_categories": "[\"reasoning\",\"factuality\",\"grounding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.749, + "details": { + "raw_score": "0.749", + "raw_score_field": "score", + "raw_model_id": "gemma-3-27b-it", + "raw_benchmark_id": "facts-grounding", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-27b-it\",\"https://llm-stats.com/benchmarks/facts-grounding\",\"https://api.llm-stats.com/leaderboard/benchmarks/facts-grounding\"]", + "raw_score_id": "facts-grounding::gemma-3-27b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "global-mmlu-lite::global-mmlu-lite-gemma-3-27b-it", + "evaluation_name": "llm_stats.global-mmlu-lite", + "source_data": { + "dataset_name": "Global-MMLU-Lite", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-27b-it", + "https://llm-stats.com/benchmarks/global-mmlu-lite", + "https://api.llm-stats.com/leaderboard/benchmarks/global-mmlu-lite" + ], + "additional_details": { + "raw_benchmark_id": "global-mmlu-lite", + "raw_model_id": "gemma-3-27b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A lightweight version of Global MMLU benchmark that evaluates language models across multiple languages while addressing cultural and linguistic biases in multilingual evaluation.", + "metric_id": "llm_stats.global-mmlu-lite.score", + "metric_name": "Global-MMLU-Lite score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "global-mmlu-lite", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Global-MMLU-Lite", + "raw_categories": "[\"language\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "14" + } + }, + "score_details": { + "score": 0.751, + "details": { + "raw_score": "0.751", + "raw_score_field": "score", + "raw_model_id": "gemma-3-27b-it", + "raw_benchmark_id": "global-mmlu-lite", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-27b-it\",\"https://llm-stats.com/benchmarks/global-mmlu-lite\",\"https://api.llm-stats.com/leaderboard/benchmarks/global-mmlu-lite\"]", + "raw_score_id": "global-mmlu-lite::gemma-3-27b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gsm8k::gsm8k-gemma-3-27b-it", + "evaluation_name": "llm_stats.gsm8k", + "source_data": { + "dataset_name": "GSM8k", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-27b-it", + "https://llm-stats.com/benchmarks/gsm8k", + "https://api.llm-stats.com/leaderboard/benchmarks/gsm8k" + ], + "additional_details": { + "raw_benchmark_id": "gsm8k", + "raw_model_id": "gemma-3-27b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Grade School Math 8K, a dataset of 8.5K high-quality linguistically diverse grade school math word problems requiring multi-step reasoning and elementary arithmetic operations.", + "metric_id": "llm_stats.gsm8k.score", + "metric_name": "GSM8k score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gsm8k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GSM8k", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "48" + } + }, + "score_details": { + "score": 0.959, + "details": { + "raw_score": "0.959", + "raw_score_field": "score", + "raw_model_id": "gemma-3-27b-it", + "raw_benchmark_id": "gsm8k", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-27b-it\",\"https://llm-stats.com/benchmarks/gsm8k\",\"https://api.llm-stats.com/leaderboard/benchmarks/gsm8k\"]", + "raw_score_id": "gsm8k::gemma-3-27b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hiddenmath::hiddenmath-gemma-3-27b-it", + "evaluation_name": "llm_stats.hiddenmath", + "source_data": { + "dataset_name": "HiddenMath", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-27b-it", + "https://llm-stats.com/benchmarks/hiddenmath", + "https://api.llm-stats.com/leaderboard/benchmarks/hiddenmath" + ], + "additional_details": { + "raw_benchmark_id": "hiddenmath", + "raw_model_id": "gemma-3-27b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Google DeepMind's internal mathematical reasoning benchmark that introduces novel problems not encountered during model training to evaluate true mathematical reasoning capabilities rather than memorization", + "metric_id": "llm_stats.hiddenmath.score", + "metric_name": "HiddenMath score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hiddenmath", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HiddenMath", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.603, + "details": { + "raw_score": "0.603", + "raw_score_field": "score", + "raw_model_id": "gemma-3-27b-it", + "raw_benchmark_id": "hiddenmath", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-27b-it\",\"https://llm-stats.com/benchmarks/hiddenmath\",\"https://api.llm-stats.com/leaderboard/benchmarks/hiddenmath\"]", + "raw_score_id": "hiddenmath::gemma-3-27b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ifeval::ifeval-gemma-3-27b-it", + "evaluation_name": "llm_stats.ifeval", + "source_data": { + "dataset_name": "IFEval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-27b-it", + "https://llm-stats.com/benchmarks/ifeval", + "https://api.llm-stats.com/leaderboard/benchmarks/ifeval" + ], + "additional_details": { + "raw_benchmark_id": "ifeval", + "raw_model_id": "gemma-3-27b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Instruction-Following Evaluation (IFEval) benchmark for large language models, focusing on verifiable instructions with 25 types of instructions and around 500 prompts containing one or more verifiable constraints", + "metric_id": "llm_stats.ifeval.score", + "metric_name": "IFEval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ifeval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IFEval", + "raw_categories": "[\"instruction_following\",\"structured_output\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "65" + } + }, + "score_details": { + "score": 0.904, + "details": { + "raw_score": "0.904", + "raw_score_field": "score", + "raw_model_id": "gemma-3-27b-it", + "raw_benchmark_id": "ifeval", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-27b-it\",\"https://llm-stats.com/benchmarks/ifeval\",\"https://api.llm-stats.com/leaderboard/benchmarks/ifeval\"]", + "raw_score_id": "ifeval::gemma-3-27b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "infovqa::infovqa-gemma-3-27b-it", + "evaluation_name": "llm_stats.infovqa", + "source_data": { + "dataset_name": "InfoVQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-27b-it", + "https://llm-stats.com/benchmarks/infovqa", + "https://api.llm-stats.com/leaderboard/benchmarks/infovqa" + ], + "additional_details": { + "raw_benchmark_id": "infovqa", + "raw_model_id": "gemma-3-27b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "InfoVQA dataset with 30,000 questions and 5,000 infographic images requiring joint reasoning over document layout, textual content, graphical elements, and data visualizations with elementary reasoning and arithmetic skills", + "metric_id": "llm_stats.infovqa.score", + "metric_name": "InfoVQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "infovqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "InfoVQA", + "raw_categories": "[\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.706, + "details": { + "raw_score": "0.706", + "raw_score_field": "score", + "raw_model_id": "gemma-3-27b-it", + "raw_benchmark_id": "infovqa", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-27b-it\",\"https://llm-stats.com/benchmarks/infovqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/infovqa\"]", + "raw_score_id": "infovqa::gemma-3-27b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "math::math-gemma-3-27b-it", + "evaluation_name": "llm_stats.math", + "source_data": { + "dataset_name": "MATH", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-27b-it", + "https://llm-stats.com/benchmarks/math", + "https://api.llm-stats.com/leaderboard/benchmarks/math" + ], + "additional_details": { + "raw_benchmark_id": "math", + "raw_model_id": "gemma-3-27b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MATH dataset contains 12,500 challenging competition mathematics problems from AMC 10, AMC 12, AIME, and other mathematics competitions. Each problem includes full step-by-step solutions and spans multiple difficulty levels (1-5) across seven mathematical subjects including Prealgebra, Algebra, Number Theory, Counting and Probability, Geometry, Intermediate Algebra, and Precalculus.", + "metric_id": "llm_stats.math.score", + "metric_name": "MATH score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "math", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MATH", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "71" + } + }, + "score_details": { + "score": 0.89, + "details": { + "raw_score": "0.89", + "raw_score_field": "score", + "raw_model_id": "gemma-3-27b-it", + "raw_benchmark_id": "math", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-27b-it\",\"https://llm-stats.com/benchmarks/math\",\"https://api.llm-stats.com/leaderboard/benchmarks/math\"]", + "raw_score_id": "math::gemma-3-27b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mbpp::mbpp-gemma-3-27b-it", + "evaluation_name": "llm_stats.mbpp", + "source_data": { + "dataset_name": "MBPP", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-27b-it", + "https://llm-stats.com/benchmarks/mbpp", + "https://api.llm-stats.com/leaderboard/benchmarks/mbpp" + ], + "additional_details": { + "raw_benchmark_id": "mbpp", + "raw_model_id": "gemma-3-27b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MBPP (Mostly Basic Python Problems) is a benchmark of 974 crowd-sourced Python programming problems designed to be solvable by entry-level programmers. Each problem consists of a task description, code solution, and 3 automated test cases covering programming fundamentals and standard library functionality.", + "metric_id": "llm_stats.mbpp.score", + "metric_name": "MBPP score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mbpp", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MBPP", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "33" + } + }, + "score_details": { + "score": 0.744, + "details": { + "raw_score": "0.744", + "raw_score_field": "score", + "raw_model_id": "gemma-3-27b-it", + "raw_benchmark_id": "mbpp", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-27b-it\",\"https://llm-stats.com/benchmarks/mbpp\",\"https://api.llm-stats.com/leaderboard/benchmarks/mbpp\"]", + "raw_score_id": "mbpp::gemma-3-27b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmmu-val::mmmu-val-gemma-3-27b-it", + "evaluation_name": "llm_stats.mmmu-val", + "source_data": { + "dataset_name": "MMMU (val)", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-27b-it", + "https://llm-stats.com/benchmarks/mmmu-val", + "https://api.llm-stats.com/leaderboard/benchmarks/mmmu-(val)" + ], + "additional_details": { + "raw_benchmark_id": "mmmu-(val)", + "raw_model_id": "gemma-3-27b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Validation set of the Massive Multi-discipline Multimodal Understanding and Reasoning benchmark. Features college-level multimodal questions across 6 core disciplines (Art & Design, Business, Science, Health & Medicine, Humanities & Social Science, Tech & Engineering) spanning 30 subjects and 183 subfields with diverse image types including charts, diagrams, maps, and tables.", + "metric_id": "llm_stats.mmmu-val.score", + "metric_name": "MMMU (val) score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmmu-(val)", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMMU (val)", + "raw_categories": "[\"multimodal\",\"reasoning\",\"general\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.649, + "details": { + "raw_score": "0.649", + "raw_score_field": "score", + "raw_model_id": "gemma-3-27b-it", + "raw_benchmark_id": "mmmu-(val)", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-27b-it\",\"https://llm-stats.com/benchmarks/mmmu-val\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmmu-(val)\"]", + "raw_score_id": "mmmu-(val)::gemma-3-27b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "natural2code::natural2code-gemma-3-27b-it", + "evaluation_name": "llm_stats.natural2code", + "source_data": { + "dataset_name": "Natural2Code", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-27b-it", + "https://llm-stats.com/benchmarks/natural2code", + "https://api.llm-stats.com/leaderboard/benchmarks/natural2code" + ], + "additional_details": { + "raw_benchmark_id": "natural2code", + "raw_model_id": "gemma-3-27b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "NaturalCodeBench (NCB) is a challenging code benchmark designed to mirror the complexity and variety of real-world coding tasks. It comprises 402 high-quality problems in Python and Java, selected from natural user queries from online coding services, covering 6 different domains.", + "metric_id": "llm_stats.natural2code.score", + "metric_name": "Natural2Code score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "natural2code", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Natural2Code", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.845, + "details": { + "raw_score": "0.845", + "raw_score_field": "score", + "raw_model_id": "gemma-3-27b-it", + "raw_benchmark_id": "natural2code", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-27b-it\",\"https://llm-stats.com/benchmarks/natural2code\",\"https://api.llm-stats.com/leaderboard/benchmarks/natural2code\"]", + "raw_score_id": "natural2code::gemma-3-27b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "textvqa::textvqa-gemma-3-27b-it", + "evaluation_name": "llm_stats.textvqa", + "source_data": { + "dataset_name": "TextVQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-27b-it", + "https://llm-stats.com/benchmarks/textvqa", + "https://api.llm-stats.com/leaderboard/benchmarks/textvqa" + ], + "additional_details": { + "raw_benchmark_id": "textvqa", + "raw_model_id": "gemma-3-27b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "TextVQA contains 45,336 questions on 28,408 images that require reasoning about text to answer. Introduced to benchmark VQA models' ability to read and reason about text within images, particularly for assistive technologies for visually impaired users. The dataset addresses the gap where existing VQA datasets had few text-based questions or were too small.", + "metric_id": "llm_stats.textvqa.score", + "metric_name": "TextVQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "textvqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TextVQA", + "raw_categories": "[\"image_to_text\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "15" + } + }, + "score_details": { + "score": 0.651, + "details": { + "raw_score": "0.651", + "raw_score_field": "score", + "raw_model_id": "gemma-3-27b-it", + "raw_benchmark_id": "textvqa", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-27b-it\",\"https://llm-stats.com/benchmarks/textvqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/textvqa\"]", + "raw_score_id": "textvqa::gemma-3-27b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "vqav2-val::vqav2-val-gemma-3-27b-it", + "evaluation_name": "llm_stats.vqav2-val", + "source_data": { + "dataset_name": "VQAv2 (val)", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-27b-it", + "https://llm-stats.com/benchmarks/vqav2-val", + "https://api.llm-stats.com/leaderboard/benchmarks/vqav2-(val)" + ], + "additional_details": { + "raw_benchmark_id": "vqav2-(val)", + "raw_model_id": "gemma-3-27b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VQAv2 is a balanced Visual Question Answering dataset containing open-ended questions about images that require understanding of vision, language, and commonsense knowledge to answer. VQAv2 addresses bias issues from the original VQA dataset by collecting complementary images such that every question is associated with similar images that result in different answers, forcing models to actually understand visual content rather than relying on language priors.", + "metric_id": "llm_stats.vqav2-val.score", + "metric_name": "VQAv2 (val) score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "vqav2-(val)", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VQAv2 (val)", + "raw_categories": "[\"image_to_text\",\"language\",\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.71, + "details": { + "raw_score": "0.71", + "raw_score_field": "score", + "raw_model_id": "gemma-3-27b-it", + "raw_benchmark_id": "vqav2-(val)", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-27b-it\",\"https://llm-stats.com/benchmarks/vqav2-val\",\"https://api.llm-stats.com/leaderboard/benchmarks/vqav2-(val)\"]", + "raw_score_id": "vqav2-(val)::gemma-3-27b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "wmt24::wmt24-gemma-3-27b-it", + "evaluation_name": "llm_stats.wmt24", + "source_data": { + "dataset_name": "WMT24++", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-27b-it", + "https://llm-stats.com/benchmarks/wmt24", + "https://api.llm-stats.com/leaderboard/benchmarks/wmt24++" + ], + "additional_details": { + "raw_benchmark_id": "wmt24++", + "raw_model_id": "gemma-3-27b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WMT24++ is a comprehensive multilingual machine translation benchmark that expands the WMT24 dataset to cover 55 languages and dialects. It includes human-written references and post-edits across four domains (literary, news, social, and speech) to evaluate machine translation systems and large language models across diverse linguistic contexts.", + "metric_id": "llm_stats.wmt24.score", + "metric_name": "WMT24++ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "wmt24++", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WMT24++", + "raw_categories": "[\"language\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.534, + "details": { + "raw_score": "0.534", + "raw_score_field": "score", + "raw_model_id": "gemma-3-27b-it", + "raw_benchmark_id": "wmt24++", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-27b-it\",\"https://llm-stats.com/benchmarks/wmt24\",\"https://api.llm-stats.com/leaderboard/benchmarks/wmt24++\"]", + "raw_score_id": "wmt24++::gemma-3-27b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/google/gemma-3-4b-it/04877a22-3fe2-4e0a-9ae5-afc8ec5a7955.json b/data/llm_stats/google/gemma-3-4b-it/04877a22-3fe2-4e0a-9ae5-afc8ec5a7955.json new file mode 100644 index 0000000000000000000000000000000000000000..9f9d3a9a7cb922165f397277f82a14f5ab016a9a --- /dev/null +++ b/data/llm_stats/google/gemma-3-4b-it/04877a22-3fe2-4e0a-9ae5-afc8ec5a7955.json @@ -0,0 +1,775 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/google_gemma-3-4b-it/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Gemma 3 4B", + "id": "google/gemma-3-4b-it", + "developer": "google", + "additional_details": { + "raw_id": "gemma-3-4b-it", + "raw_name": "Gemma 3 4B", + "raw_model_id": "gemma-3-4b-it", + "raw_model_name": "Gemma 3 4B", + "raw_organization_id": "google", + "raw_organization_name": "Google", + "raw_release_date": "2025-03-12", + "raw_announcement_date": "2025-03-12", + "raw_multimodal": "true", + "raw_param_count": "4000000000", + "raw_provider_slug": "google", + "raw_provider_name": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "big-bench-extra-hard::big-bench-extra-hard-gemma-3-4b-it", + "evaluation_name": "llm_stats.big-bench-extra-hard", + "source_data": { + "dataset_name": "BIG-Bench Extra Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-4b-it", + "https://llm-stats.com/benchmarks/big-bench-extra-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench-extra-hard" + ], + "additional_details": { + "raw_benchmark_id": "big-bench-extra-hard", + "raw_model_id": "gemma-3-4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIG-Bench Extra Hard (BBEH) is a challenging benchmark that replaces each task in BIG-Bench Hard with a novel task that probes similar reasoning capabilities but exhibits significantly increased difficulty. The benchmark contains 23 tasks testing diverse reasoning skills including many-hop reasoning, causal understanding, spatial reasoning, temporal arithmetic, geometric reasoning, linguistic reasoning, logic puzzles, and humor understanding. Designed to address saturation on existing benchmarks where state-of-the-art models achieve near-perfect scores, BBEH shows substantial room for improvement with best models achieving only 9.8-44.8% average accuracy.", + "metric_id": "llm_stats.big-bench-extra-hard.score", + "metric_name": "BIG-Bench Extra Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench-extra-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench Extra Hard", + "raw_categories": "[\"language\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.11, + "details": { + "raw_score": "0.11", + "raw_score_field": "score", + "raw_model_id": "gemma-3-4b-it", + "raw_benchmark_id": "big-bench-extra-hard", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-4b-it\",\"https://llm-stats.com/benchmarks/big-bench-extra-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench-extra-hard\"]", + "raw_score_id": "big-bench-extra-hard::gemma-3-4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "big-bench-hard::big-bench-hard-gemma-3-4b-it", + "evaluation_name": "llm_stats.big-bench-hard", + "source_data": { + "dataset_name": "BIG-Bench Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-4b-it", + "https://llm-stats.com/benchmarks/big-bench-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard" + ], + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_model_id": "gemma-3-4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIG-Bench Hard (BBH) is a subset of 23 challenging BIG-Bench tasks selected because prior language model evaluations did not outperform average human-rater performance. The benchmark contains 6,511 evaluation examples testing various forms of multi-step reasoning including arithmetic, logical reasoning (Boolean expressions, logical deduction), geometric reasoning, temporal reasoning, and language understanding. Tasks require capabilities such as causal judgment, object counting, navigation, pattern recognition, and complex problem solving.", + "metric_id": "llm_stats.big-bench-hard.score", + "metric_name": "BIG-Bench Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench Hard", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "21" + } + }, + "score_details": { + "score": 0.722, + "details": { + "raw_score": "0.722", + "raw_score_field": "score", + "raw_model_id": "gemma-3-4b-it", + "raw_benchmark_id": "big-bench-hard", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-4b-it\",\"https://llm-stats.com/benchmarks/big-bench-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard\"]", + "raw_score_id": "big-bench-hard::gemma-3-4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bird-sql-dev::bird-sql-dev-gemma-3-4b-it", + "evaluation_name": "llm_stats.bird-sql-dev", + "source_data": { + "dataset_name": "Bird-SQL (dev)", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-4b-it", + "https://llm-stats.com/benchmarks/bird-sql-dev", + "https://api.llm-stats.com/leaderboard/benchmarks/bird-sql-(dev)" + ], + "additional_details": { + "raw_benchmark_id": "bird-sql-(dev)", + "raw_model_id": "gemma-3-4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIRD (BIg Bench for LaRge-scale Database Grounded Text-to-SQLs) is a comprehensive text-to-SQL benchmark containing 12,751 question-SQL pairs across 95 databases (33.4 GB total) spanning 37+ professional domains. It evaluates large language models' ability to convert natural language to executable SQL queries in real-world scenarios with complex database schemas and dirty data.", + "metric_id": "llm_stats.bird-sql-dev.score", + "metric_name": "Bird-SQL (dev) score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bird-sql-(dev)", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Bird-SQL (dev)", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.363, + "details": { + "raw_score": "0.363", + "raw_score_field": "score", + "raw_model_id": "gemma-3-4b-it", + "raw_benchmark_id": "bird-sql-(dev)", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-4b-it\",\"https://llm-stats.com/benchmarks/bird-sql-dev\",\"https://api.llm-stats.com/leaderboard/benchmarks/bird-sql-(dev)\"]", + "raw_score_id": "bird-sql-(dev)::gemma-3-4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "eclektic::eclektic-gemma-3-4b-it", + "evaluation_name": "llm_stats.eclektic", + "source_data": { + "dataset_name": "ECLeKTic", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-4b-it", + "https://llm-stats.com/benchmarks/eclektic", + "https://api.llm-stats.com/leaderboard/benchmarks/eclektic" + ], + "additional_details": { + "raw_benchmark_id": "eclektic", + "raw_model_id": "gemma-3-4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multilingual closed-book question answering dataset that evaluates cross-lingual knowledge transfer in large language models across 12 languages, using knowledge-seeking questions based on Wikipedia articles that exist only in one language", + "metric_id": "llm_stats.eclektic.score", + "metric_name": "ECLeKTic score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "eclektic", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ECLeKTic", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.046, + "details": { + "raw_score": "0.046", + "raw_score_field": "score", + "raw_model_id": "gemma-3-4b-it", + "raw_benchmark_id": "eclektic", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-4b-it\",\"https://llm-stats.com/benchmarks/eclektic\",\"https://api.llm-stats.com/leaderboard/benchmarks/eclektic\"]", + "raw_score_id": "eclektic::gemma-3-4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "facts-grounding::facts-grounding-gemma-3-4b-it", + "evaluation_name": "llm_stats.facts-grounding", + "source_data": { + "dataset_name": "FACTS Grounding", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-4b-it", + "https://llm-stats.com/benchmarks/facts-grounding", + "https://api.llm-stats.com/leaderboard/benchmarks/facts-grounding" + ], + "additional_details": { + "raw_benchmark_id": "facts-grounding", + "raw_model_id": "gemma-3-4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A benchmark evaluating language models' ability to generate factually accurate and well-grounded responses based on long-form input context, comprising 1,719 examples with documents up to 32k tokens requiring detailed responses that are fully grounded in provided documents", + "metric_id": "llm_stats.facts-grounding.score", + "metric_name": "FACTS Grounding score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "facts-grounding", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FACTS Grounding", + "raw_categories": "[\"reasoning\",\"factuality\",\"grounding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.701, + "details": { + "raw_score": "0.701", + "raw_score_field": "score", + "raw_model_id": "gemma-3-4b-it", + "raw_benchmark_id": "facts-grounding", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-4b-it\",\"https://llm-stats.com/benchmarks/facts-grounding\",\"https://api.llm-stats.com/leaderboard/benchmarks/facts-grounding\"]", + "raw_score_id": "facts-grounding::gemma-3-4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "global-mmlu-lite::global-mmlu-lite-gemma-3-4b-it", + "evaluation_name": "llm_stats.global-mmlu-lite", + "source_data": { + "dataset_name": "Global-MMLU-Lite", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-4b-it", + "https://llm-stats.com/benchmarks/global-mmlu-lite", + "https://api.llm-stats.com/leaderboard/benchmarks/global-mmlu-lite" + ], + "additional_details": { + "raw_benchmark_id": "global-mmlu-lite", + "raw_model_id": "gemma-3-4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A lightweight version of Global MMLU benchmark that evaluates language models across multiple languages while addressing cultural and linguistic biases in multilingual evaluation.", + "metric_id": "llm_stats.global-mmlu-lite.score", + "metric_name": "Global-MMLU-Lite score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "global-mmlu-lite", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Global-MMLU-Lite", + "raw_categories": "[\"language\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "14" + } + }, + "score_details": { + "score": 0.545, + "details": { + "raw_score": "0.545", + "raw_score_field": "score", + "raw_model_id": "gemma-3-4b-it", + "raw_benchmark_id": "global-mmlu-lite", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-4b-it\",\"https://llm-stats.com/benchmarks/global-mmlu-lite\",\"https://api.llm-stats.com/leaderboard/benchmarks/global-mmlu-lite\"]", + "raw_score_id": "global-mmlu-lite::gemma-3-4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hiddenmath::hiddenmath-gemma-3-4b-it", + "evaluation_name": "llm_stats.hiddenmath", + "source_data": { + "dataset_name": "HiddenMath", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-4b-it", + "https://llm-stats.com/benchmarks/hiddenmath", + "https://api.llm-stats.com/leaderboard/benchmarks/hiddenmath" + ], + "additional_details": { + "raw_benchmark_id": "hiddenmath", + "raw_model_id": "gemma-3-4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Google DeepMind's internal mathematical reasoning benchmark that introduces novel problems not encountered during model training to evaluate true mathematical reasoning capabilities rather than memorization", + "metric_id": "llm_stats.hiddenmath.score", + "metric_name": "HiddenMath score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hiddenmath", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HiddenMath", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.43, + "details": { + "raw_score": "0.43", + "raw_score_field": "score", + "raw_model_id": "gemma-3-4b-it", + "raw_benchmark_id": "hiddenmath", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-4b-it\",\"https://llm-stats.com/benchmarks/hiddenmath\",\"https://api.llm-stats.com/leaderboard/benchmarks/hiddenmath\"]", + "raw_score_id": "hiddenmath::gemma-3-4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ifeval::ifeval-gemma-3-4b-it", + "evaluation_name": "llm_stats.ifeval", + "source_data": { + "dataset_name": "IFEval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-4b-it", + "https://llm-stats.com/benchmarks/ifeval", + "https://api.llm-stats.com/leaderboard/benchmarks/ifeval" + ], + "additional_details": { + "raw_benchmark_id": "ifeval", + "raw_model_id": "gemma-3-4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Instruction-Following Evaluation (IFEval) benchmark for large language models, focusing on verifiable instructions with 25 types of instructions and around 500 prompts containing one or more verifiable constraints", + "metric_id": "llm_stats.ifeval.score", + "metric_name": "IFEval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ifeval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IFEval", + "raw_categories": "[\"instruction_following\",\"structured_output\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "65" + } + }, + "score_details": { + "score": 0.902, + "details": { + "raw_score": "0.902", + "raw_score_field": "score", + "raw_model_id": "gemma-3-4b-it", + "raw_benchmark_id": "ifeval", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-4b-it\",\"https://llm-stats.com/benchmarks/ifeval\",\"https://api.llm-stats.com/leaderboard/benchmarks/ifeval\"]", + "raw_score_id": "ifeval::gemma-3-4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "infovqa::infovqa-gemma-3-4b-it", + "evaluation_name": "llm_stats.infovqa", + "source_data": { + "dataset_name": "InfoVQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-4b-it", + "https://llm-stats.com/benchmarks/infovqa", + "https://api.llm-stats.com/leaderboard/benchmarks/infovqa" + ], + "additional_details": { + "raw_benchmark_id": "infovqa", + "raw_model_id": "gemma-3-4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "InfoVQA dataset with 30,000 questions and 5,000 infographic images requiring joint reasoning over document layout, textual content, graphical elements, and data visualizations with elementary reasoning and arithmetic skills", + "metric_id": "llm_stats.infovqa.score", + "metric_name": "InfoVQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "infovqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "InfoVQA", + "raw_categories": "[\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.5, + "details": { + "raw_score": "0.5", + "raw_score_field": "score", + "raw_model_id": "gemma-3-4b-it", + "raw_benchmark_id": "infovqa", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-4b-it\",\"https://llm-stats.com/benchmarks/infovqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/infovqa\"]", + "raw_score_id": "infovqa::gemma-3-4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmmu-val::mmmu-val-gemma-3-4b-it", + "evaluation_name": "llm_stats.mmmu-val", + "source_data": { + "dataset_name": "MMMU (val)", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-4b-it", + "https://llm-stats.com/benchmarks/mmmu-val", + "https://api.llm-stats.com/leaderboard/benchmarks/mmmu-(val)" + ], + "additional_details": { + "raw_benchmark_id": "mmmu-(val)", + "raw_model_id": "gemma-3-4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Validation set of the Massive Multi-discipline Multimodal Understanding and Reasoning benchmark. Features college-level multimodal questions across 6 core disciplines (Art & Design, Business, Science, Health & Medicine, Humanities & Social Science, Tech & Engineering) spanning 30 subjects and 183 subfields with diverse image types including charts, diagrams, maps, and tables.", + "metric_id": "llm_stats.mmmu-val.score", + "metric_name": "MMMU (val) score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmmu-(val)", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMMU (val)", + "raw_categories": "[\"multimodal\",\"reasoning\",\"general\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.488, + "details": { + "raw_score": "0.488", + "raw_score_field": "score", + "raw_model_id": "gemma-3-4b-it", + "raw_benchmark_id": "mmmu-(val)", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-4b-it\",\"https://llm-stats.com/benchmarks/mmmu-val\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmmu-(val)\"]", + "raw_score_id": "mmmu-(val)::gemma-3-4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "natural2code::natural2code-gemma-3-4b-it", + "evaluation_name": "llm_stats.natural2code", + "source_data": { + "dataset_name": "Natural2Code", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-4b-it", + "https://llm-stats.com/benchmarks/natural2code", + "https://api.llm-stats.com/leaderboard/benchmarks/natural2code" + ], + "additional_details": { + "raw_benchmark_id": "natural2code", + "raw_model_id": "gemma-3-4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "NaturalCodeBench (NCB) is a challenging code benchmark designed to mirror the complexity and variety of real-world coding tasks. It comprises 402 high-quality problems in Python and Java, selected from natural user queries from online coding services, covering 6 different domains.", + "metric_id": "llm_stats.natural2code.score", + "metric_name": "Natural2Code score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "natural2code", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Natural2Code", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.703, + "details": { + "raw_score": "0.703", + "raw_score_field": "score", + "raw_model_id": "gemma-3-4b-it", + "raw_benchmark_id": "natural2code", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-4b-it\",\"https://llm-stats.com/benchmarks/natural2code\",\"https://api.llm-stats.com/leaderboard/benchmarks/natural2code\"]", + "raw_score_id": "natural2code::gemma-3-4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "textvqa::textvqa-gemma-3-4b-it", + "evaluation_name": "llm_stats.textvqa", + "source_data": { + "dataset_name": "TextVQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-4b-it", + "https://llm-stats.com/benchmarks/textvqa", + "https://api.llm-stats.com/leaderboard/benchmarks/textvqa" + ], + "additional_details": { + "raw_benchmark_id": "textvqa", + "raw_model_id": "gemma-3-4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "TextVQA contains 45,336 questions on 28,408 images that require reasoning about text to answer. Introduced to benchmark VQA models' ability to read and reason about text within images, particularly for assistive technologies for visually impaired users. The dataset addresses the gap where existing VQA datasets had few text-based questions or were too small.", + "metric_id": "llm_stats.textvqa.score", + "metric_name": "TextVQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "textvqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TextVQA", + "raw_categories": "[\"image_to_text\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "15" + } + }, + "score_details": { + "score": 0.578, + "details": { + "raw_score": "0.578", + "raw_score_field": "score", + "raw_model_id": "gemma-3-4b-it", + "raw_benchmark_id": "textvqa", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-4b-it\",\"https://llm-stats.com/benchmarks/textvqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/textvqa\"]", + "raw_score_id": "textvqa::gemma-3-4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "vqav2-val::vqav2-val-gemma-3-4b-it", + "evaluation_name": "llm_stats.vqav2-val", + "source_data": { + "dataset_name": "VQAv2 (val)", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-4b-it", + "https://llm-stats.com/benchmarks/vqav2-val", + "https://api.llm-stats.com/leaderboard/benchmarks/vqav2-(val)" + ], + "additional_details": { + "raw_benchmark_id": "vqav2-(val)", + "raw_model_id": "gemma-3-4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VQAv2 is a balanced Visual Question Answering dataset containing open-ended questions about images that require understanding of vision, language, and commonsense knowledge to answer. VQAv2 addresses bias issues from the original VQA dataset by collecting complementary images such that every question is associated with similar images that result in different answers, forcing models to actually understand visual content rather than relying on language priors.", + "metric_id": "llm_stats.vqav2-val.score", + "metric_name": "VQAv2 (val) score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "vqav2-(val)", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VQAv2 (val)", + "raw_categories": "[\"image_to_text\",\"language\",\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.624, + "details": { + "raw_score": "0.624", + "raw_score_field": "score", + "raw_model_id": "gemma-3-4b-it", + "raw_benchmark_id": "vqav2-(val)", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-4b-it\",\"https://llm-stats.com/benchmarks/vqav2-val\",\"https://api.llm-stats.com/leaderboard/benchmarks/vqav2-(val)\"]", + "raw_score_id": "vqav2-(val)::gemma-3-4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "wmt24::wmt24-gemma-3-4b-it", + "evaluation_name": "llm_stats.wmt24", + "source_data": { + "dataset_name": "WMT24++", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3-4b-it", + "https://llm-stats.com/benchmarks/wmt24", + "https://api.llm-stats.com/leaderboard/benchmarks/wmt24++" + ], + "additional_details": { + "raw_benchmark_id": "wmt24++", + "raw_model_id": "gemma-3-4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WMT24++ is a comprehensive multilingual machine translation benchmark that expands the WMT24 dataset to cover 55 languages and dialects. It includes human-written references and post-edits across four domains (literary, news, social, and speech) to evaluate machine translation systems and large language models across diverse linguistic contexts.", + "metric_id": "llm_stats.wmt24.score", + "metric_name": "WMT24++ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "wmt24++", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WMT24++", + "raw_categories": "[\"language\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.468, + "details": { + "raw_score": "0.468", + "raw_score_field": "score", + "raw_model_id": "gemma-3-4b-it", + "raw_benchmark_id": "wmt24++", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3-4b-it\",\"https://llm-stats.com/benchmarks/wmt24\",\"https://api.llm-stats.com/leaderboard/benchmarks/wmt24++\"]", + "raw_score_id": "wmt24++::gemma-3-4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/google/gemma-3n-e2b-it-litert-preview/9a341c6b-234a-47bc-8775-bc69c6f7e113.json b/data/llm_stats/google/gemma-3n-e2b-it-litert-preview/9a341c6b-234a-47bc-8775-bc69c6f7e113.json new file mode 100644 index 0000000000000000000000000000000000000000..52b2df39ce0d7e660a80f50c8eefffa763db7e43 --- /dev/null +++ b/data/llm_stats/google/gemma-3n-e2b-it-litert-preview/9a341c6b-234a-47bc-8775-bc69c6f7e113.json @@ -0,0 +1,723 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/google_gemma-3n-e2b-it-litert-preview/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Gemma 3n E2B Instructed LiteRT (Preview)", + "id": "google/gemma-3n-e2b-it-litert-preview", + "developer": "google", + "additional_details": { + "raw_id": "gemma-3n-e2b-it-litert-preview", + "raw_name": "Gemma 3n E2B Instructed LiteRT (Preview)", + "raw_model_id": "gemma-3n-e2b-it-litert-preview", + "raw_model_name": "Gemma 3n E2B Instructed LiteRT (Preview)", + "raw_organization_id": "google", + "raw_organization_name": "Google", + "raw_release_date": "2025-05-20", + "raw_announcement_date": "2025-05-20", + "raw_multimodal": "true", + "raw_param_count": "1910000000", + "raw_provider_slug": "google", + "raw_provider_name": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "arc-e::arc-e-gemma-3n-e2b-it-litert-preview", + "evaluation_name": "llm_stats.arc-e", + "source_data": { + "dataset_name": "ARC-E", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e2b-it-litert-preview", + "https://llm-stats.com/benchmarks/arc-e", + "https://api.llm-stats.com/leaderboard/benchmarks/arc-e" + ], + "additional_details": { + "raw_benchmark_id": "arc-e", + "raw_model_id": "gemma-3n-e2b-it-litert-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ARC-E (AI2 Reasoning Challenge - Easy Set) is a subset of grade-school level, multiple-choice science questions that requires knowledge and reasoning capabilities. Part of the AI2 Reasoning Challenge dataset containing 5,197 questions that test scientific reasoning and factual knowledge. The Easy Set contains questions that are answerable by retrieval-based and word co-occurrence algorithms, making them more accessible than the Challenge Set.", + "metric_id": "llm_stats.arc-e.score", + "metric_name": "ARC-E score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arc-e", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ARC-E", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.758, + "details": { + "raw_score": "0.758", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e2b-it-litert-preview", + "raw_benchmark_id": "arc-e", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e2b-it-litert-preview\",\"https://llm-stats.com/benchmarks/arc-e\",\"https://api.llm-stats.com/leaderboard/benchmarks/arc-e\"]", + "raw_score_id": "arc-e::gemma-3n-e2b-it-litert-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "big-bench-hard::big-bench-hard-gemma-3n-e2b-it-litert-preview", + "evaluation_name": "llm_stats.big-bench-hard", + "source_data": { + "dataset_name": "BIG-Bench Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e2b-it-litert-preview", + "https://llm-stats.com/benchmarks/big-bench-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard" + ], + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_model_id": "gemma-3n-e2b-it-litert-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIG-Bench Hard (BBH) is a subset of 23 challenging BIG-Bench tasks selected because prior language model evaluations did not outperform average human-rater performance. The benchmark contains 6,511 evaluation examples testing various forms of multi-step reasoning including arithmetic, logical reasoning (Boolean expressions, logical deduction), geometric reasoning, temporal reasoning, and language understanding. Tasks require capabilities such as causal judgment, object counting, navigation, pattern recognition, and complex problem solving.", + "metric_id": "llm_stats.big-bench-hard.score", + "metric_name": "BIG-Bench Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench Hard", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "21" + } + }, + "score_details": { + "score": 0.443, + "details": { + "raw_score": "0.443", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e2b-it-litert-preview", + "raw_benchmark_id": "big-bench-hard", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e2b-it-litert-preview\",\"https://llm-stats.com/benchmarks/big-bench-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard\"]", + "raw_score_id": "big-bench-hard::gemma-3n-e2b-it-litert-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "boolq::boolq-gemma-3n-e2b-it-litert-preview", + "evaluation_name": "llm_stats.boolq", + "source_data": { + "dataset_name": "BoolQ", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e2b-it-litert-preview", + "https://llm-stats.com/benchmarks/boolq", + "https://api.llm-stats.com/leaderboard/benchmarks/boolq" + ], + "additional_details": { + "raw_benchmark_id": "boolq", + "raw_model_id": "gemma-3n-e2b-it-litert-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BoolQ is a reading comprehension dataset for yes/no questions containing 15,942 naturally occurring examples. Each example consists of a question, passage, and boolean answer, where questions are generated in unprompted and unconstrained settings. The dataset challenges models with complex, non-factoid information requiring entailment-like inference to solve.", + "metric_id": "llm_stats.boolq.score", + "metric_name": "BoolQ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "boolq", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BoolQ", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.764, + "details": { + "raw_score": "0.764", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e2b-it-litert-preview", + "raw_benchmark_id": "boolq", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e2b-it-litert-preview\",\"https://llm-stats.com/benchmarks/boolq\",\"https://api.llm-stats.com/leaderboard/benchmarks/boolq\"]", + "raw_score_id": "boolq::gemma-3n-e2b-it-litert-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "codegolf-v2.2::codegolf-v2.2-gemma-3n-e2b-it-litert-preview", + "evaluation_name": "llm_stats.codegolf-v2.2", + "source_data": { + "dataset_name": "Codegolf v2.2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e2b-it-litert-preview", + "https://llm-stats.com/benchmarks/codegolf-v2.2", + "https://api.llm-stats.com/leaderboard/benchmarks/codegolf-v2.2" + ], + "additional_details": { + "raw_benchmark_id": "codegolf-v2.2", + "raw_model_id": "gemma-3n-e2b-it-litert-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Codegolf v2.2 benchmark", + "metric_id": "llm_stats.codegolf-v2.2.score", + "metric_name": "Codegolf v2.2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "codegolf-v2.2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Codegolf v2.2", + "raw_categories": "[\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.11, + "details": { + "raw_score": "0.11", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e2b-it-litert-preview", + "raw_benchmark_id": "codegolf-v2.2", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e2b-it-litert-preview\",\"https://llm-stats.com/benchmarks/codegolf-v2.2\",\"https://api.llm-stats.com/leaderboard/benchmarks/codegolf-v2.2\"]", + "raw_score_id": "codegolf-v2.2::gemma-3n-e2b-it-litert-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "eclektic::eclektic-gemma-3n-e2b-it-litert-preview", + "evaluation_name": "llm_stats.eclektic", + "source_data": { + "dataset_name": "ECLeKTic", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e2b-it-litert-preview", + "https://llm-stats.com/benchmarks/eclektic", + "https://api.llm-stats.com/leaderboard/benchmarks/eclektic" + ], + "additional_details": { + "raw_benchmark_id": "eclektic", + "raw_model_id": "gemma-3n-e2b-it-litert-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multilingual closed-book question answering dataset that evaluates cross-lingual knowledge transfer in large language models across 12 languages, using knowledge-seeking questions based on Wikipedia articles that exist only in one language", + "metric_id": "llm_stats.eclektic.score", + "metric_name": "ECLeKTic score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "eclektic", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ECLeKTic", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.025, + "details": { + "raw_score": "0.025", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e2b-it-litert-preview", + "raw_benchmark_id": "eclektic", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e2b-it-litert-preview\",\"https://llm-stats.com/benchmarks/eclektic\",\"https://api.llm-stats.com/leaderboard/benchmarks/eclektic\"]", + "raw_score_id": "eclektic::gemma-3n-e2b-it-litert-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "global-mmlu-lite::global-mmlu-lite-gemma-3n-e2b-it-litert-preview", + "evaluation_name": "llm_stats.global-mmlu-lite", + "source_data": { + "dataset_name": "Global-MMLU-Lite", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e2b-it-litert-preview", + "https://llm-stats.com/benchmarks/global-mmlu-lite", + "https://api.llm-stats.com/leaderboard/benchmarks/global-mmlu-lite" + ], + "additional_details": { + "raw_benchmark_id": "global-mmlu-lite", + "raw_model_id": "gemma-3n-e2b-it-litert-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A lightweight version of Global MMLU benchmark that evaluates language models across multiple languages while addressing cultural and linguistic biases in multilingual evaluation.", + "metric_id": "llm_stats.global-mmlu-lite.score", + "metric_name": "Global-MMLU-Lite score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "global-mmlu-lite", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Global-MMLU-Lite", + "raw_categories": "[\"language\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "14" + } + }, + "score_details": { + "score": 0.59, + "details": { + "raw_score": "0.59", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e2b-it-litert-preview", + "raw_benchmark_id": "global-mmlu-lite", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e2b-it-litert-preview\",\"https://llm-stats.com/benchmarks/global-mmlu-lite\",\"https://api.llm-stats.com/leaderboard/benchmarks/global-mmlu-lite\"]", + "raw_score_id": "global-mmlu-lite::gemma-3n-e2b-it-litert-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "global-mmlu::global-mmlu-gemma-3n-e2b-it-litert-preview", + "evaluation_name": "llm_stats.global-mmlu", + "source_data": { + "dataset_name": "Global-MMLU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e2b-it-litert-preview", + "https://llm-stats.com/benchmarks/global-mmlu", + "https://api.llm-stats.com/leaderboard/benchmarks/global-mmlu" + ], + "additional_details": { + "raw_benchmark_id": "global-mmlu", + "raw_model_id": "gemma-3n-e2b-it-litert-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A comprehensive multilingual benchmark covering 42 languages that addresses cultural and linguistic biases in evaluation, with improved translation quality and culturally sensitive question subsets.", + "metric_id": "llm_stats.global-mmlu.score", + "metric_name": "Global-MMLU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "global-mmlu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Global-MMLU", + "raw_categories": "[\"language\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.551, + "details": { + "raw_score": "0.551", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e2b-it-litert-preview", + "raw_benchmark_id": "global-mmlu", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e2b-it-litert-preview\",\"https://llm-stats.com/benchmarks/global-mmlu\",\"https://api.llm-stats.com/leaderboard/benchmarks/global-mmlu\"]", + "raw_score_id": "global-mmlu::gemma-3n-e2b-it-litert-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hiddenmath::hiddenmath-gemma-3n-e2b-it-litert-preview", + "evaluation_name": "llm_stats.hiddenmath", + "source_data": { + "dataset_name": "HiddenMath", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e2b-it-litert-preview", + "https://llm-stats.com/benchmarks/hiddenmath", + "https://api.llm-stats.com/leaderboard/benchmarks/hiddenmath" + ], + "additional_details": { + "raw_benchmark_id": "hiddenmath", + "raw_model_id": "gemma-3n-e2b-it-litert-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Google DeepMind's internal mathematical reasoning benchmark that introduces novel problems not encountered during model training to evaluate true mathematical reasoning capabilities rather than memorization", + "metric_id": "llm_stats.hiddenmath.score", + "metric_name": "HiddenMath score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hiddenmath", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HiddenMath", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.277, + "details": { + "raw_score": "0.277", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e2b-it-litert-preview", + "raw_benchmark_id": "hiddenmath", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e2b-it-litert-preview\",\"https://llm-stats.com/benchmarks/hiddenmath\",\"https://api.llm-stats.com/leaderboard/benchmarks/hiddenmath\"]", + "raw_score_id": "hiddenmath::gemma-3n-e2b-it-litert-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "livecodebench-v5::livecodebench-v5-gemma-3n-e2b-it-litert-preview", + "evaluation_name": "llm_stats.livecodebench-v5", + "source_data": { + "dataset_name": "LiveCodeBench v5", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e2b-it-litert-preview", + "https://llm-stats.com/benchmarks/livecodebench-v5", + "https://api.llm-stats.com/leaderboard/benchmarks/livecodebench-v5" + ], + "additional_details": { + "raw_benchmark_id": "livecodebench-v5", + "raw_model_id": "gemma-3n-e2b-it-litert-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LiveCodeBench is a holistic and contamination-free evaluation benchmark for large language models for code. It continuously collects new problems from programming contests (LeetCode, AtCoder, CodeForces) and evaluates four different scenarios: code generation, self-repair, code execution, and test output prediction. Problems are annotated with release dates to enable evaluation on unseen problems released after a model's training cutoff.", + "metric_id": "llm_stats.livecodebench-v5.score", + "metric_name": "LiveCodeBench v5 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "livecodebench-v5", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LiveCodeBench v5", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.186, + "details": { + "raw_score": "0.186", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e2b-it-litert-preview", + "raw_benchmark_id": "livecodebench-v5", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e2b-it-litert-preview\",\"https://llm-stats.com/benchmarks/livecodebench-v5\",\"https://api.llm-stats.com/leaderboard/benchmarks/livecodebench-v5\"]", + "raw_score_id": "livecodebench-v5::gemma-3n-e2b-it-litert-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "natural-questions::natural-questions-gemma-3n-e2b-it-litert-preview", + "evaluation_name": "llm_stats.natural-questions", + "source_data": { + "dataset_name": "Natural Questions", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e2b-it-litert-preview", + "https://llm-stats.com/benchmarks/natural-questions", + "https://api.llm-stats.com/leaderboard/benchmarks/natural-questions" + ], + "additional_details": { + "raw_benchmark_id": "natural-questions", + "raw_model_id": "gemma-3n-e2b-it-litert-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Natural Questions is a question answering dataset featuring real anonymized queries issued to Google search engine. It contains 307,373 training examples where annotators provide long answers (passages) and short answers (entities) from Wikipedia pages, or mark them as unanswerable.", + "metric_id": "llm_stats.natural-questions.score", + "metric_name": "Natural Questions score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "natural-questions", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Natural Questions", + "raw_categories": "[\"reasoning\",\"search\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.155, + "details": { + "raw_score": "0.155", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e2b-it-litert-preview", + "raw_benchmark_id": "natural-questions", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e2b-it-litert-preview\",\"https://llm-stats.com/benchmarks/natural-questions\",\"https://api.llm-stats.com/leaderboard/benchmarks/natural-questions\"]", + "raw_score_id": "natural-questions::gemma-3n-e2b-it-litert-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "piqa::piqa-gemma-3n-e2b-it-litert-preview", + "evaluation_name": "llm_stats.piqa", + "source_data": { + "dataset_name": "PIQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e2b-it-litert-preview", + "https://llm-stats.com/benchmarks/piqa", + "https://api.llm-stats.com/leaderboard/benchmarks/piqa" + ], + "additional_details": { + "raw_benchmark_id": "piqa", + "raw_model_id": "gemma-3n-e2b-it-litert-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "PIQA (Physical Interaction: Question Answering) is a benchmark dataset for physical commonsense reasoning in natural language. It tests AI systems' ability to answer questions requiring physical world knowledge through multiple choice questions with everyday situations, focusing on atypical solutions inspired by instructables.com. The dataset contains 21,000 multiple choice questions where models must choose the most appropriate solution for physical interactions.", + "metric_id": "llm_stats.piqa.score", + "metric_name": "PIQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "piqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PIQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.789, + "details": { + "raw_score": "0.789", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e2b-it-litert-preview", + "raw_benchmark_id": "piqa", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e2b-it-litert-preview\",\"https://llm-stats.com/benchmarks/piqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/piqa\"]", + "raw_score_id": "piqa::gemma-3n-e2b-it-litert-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "social-iqa::social-iqa-gemma-3n-e2b-it-litert-preview", + "evaluation_name": "llm_stats.social-iqa", + "source_data": { + "dataset_name": "Social IQa", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e2b-it-litert-preview", + "https://llm-stats.com/benchmarks/social-iqa", + "https://api.llm-stats.com/leaderboard/benchmarks/social-iqa" + ], + "additional_details": { + "raw_benchmark_id": "social-iqa", + "raw_model_id": "gemma-3n-e2b-it-litert-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The first large-scale benchmark for commonsense reasoning about social situations. Contains 38,000 multiple choice questions probing emotional and social intelligence in everyday situations, testing commonsense understanding of social interactions and theory of mind reasoning about the implied emotions and behavior of others.", + "metric_id": "llm_stats.social-iqa.score", + "metric_name": "Social IQa score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "social-iqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Social IQa", + "raw_categories": "[\"psychology\",\"reasoning\",\"creativity\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.488, + "details": { + "raw_score": "0.488", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e2b-it-litert-preview", + "raw_benchmark_id": "social-iqa", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e2b-it-litert-preview\",\"https://llm-stats.com/benchmarks/social-iqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/social-iqa\"]", + "raw_score_id": "social-iqa::gemma-3n-e2b-it-litert-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "triviaqa::triviaqa-gemma-3n-e2b-it-litert-preview", + "evaluation_name": "llm_stats.triviaqa", + "source_data": { + "dataset_name": "TriviaQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e2b-it-litert-preview", + "https://llm-stats.com/benchmarks/triviaqa", + "https://api.llm-stats.com/leaderboard/benchmarks/triviaqa" + ], + "additional_details": { + "raw_benchmark_id": "triviaqa", + "raw_model_id": "gemma-3n-e2b-it-litert-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A large-scale reading comprehension dataset containing over 650K question-answer-evidence triples. TriviaQA includes 95K question-answer pairs authored by trivia enthusiasts and independently gathered evidence documents (six per question on average) that provide high quality distant supervision for answering the questions. The dataset features relatively complex, compositional questions with considerable syntactic and lexical variability, requiring cross-sentence reasoning to find answers.", + "metric_id": "llm_stats.triviaqa.score", + "metric_name": "TriviaQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "triviaqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TriviaQA", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.608, + "details": { + "raw_score": "0.608", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e2b-it-litert-preview", + "raw_benchmark_id": "triviaqa", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e2b-it-litert-preview\",\"https://llm-stats.com/benchmarks/triviaqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/triviaqa\"]", + "raw_score_id": "triviaqa::gemma-3n-e2b-it-litert-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/google/gemma-3n-e2b/48165fbc-35b1-44a0-b72f-ad215650216e.json b/data/llm_stats/google/gemma-3n-e2b/48165fbc-35b1-44a0-b72f-ad215650216e.json new file mode 100644 index 0000000000000000000000000000000000000000..b0128af2a019017f141426f56251dda032894462 --- /dev/null +++ b/data/llm_stats/google/gemma-3n-e2b/48165fbc-35b1-44a0-b72f-ad215650216e.json @@ -0,0 +1,463 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/google_gemma-3n-e2b/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Gemma 3n E2B", + "id": "google/gemma-3n-e2b", + "developer": "google", + "additional_details": { + "raw_id": "gemma-3n-e2b", + "raw_name": "Gemma 3n E2B", + "raw_model_id": "gemma-3n-e2b", + "raw_model_name": "Gemma 3n E2B", + "raw_organization_id": "google", + "raw_organization_name": "Google", + "raw_release_date": "2025-06-26", + "raw_announcement_date": "2025-06-26", + "raw_multimodal": "true", + "raw_param_count": "8000000000", + "raw_provider_slug": "google", + "raw_provider_name": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "arc-e::arc-e-gemma-3n-e2b", + "evaluation_name": "llm_stats.arc-e", + "source_data": { + "dataset_name": "ARC-E", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e2b", + "https://llm-stats.com/benchmarks/arc-e", + "https://api.llm-stats.com/leaderboard/benchmarks/arc-e" + ], + "additional_details": { + "raw_benchmark_id": "arc-e", + "raw_model_id": "gemma-3n-e2b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ARC-E (AI2 Reasoning Challenge - Easy Set) is a subset of grade-school level, multiple-choice science questions that requires knowledge and reasoning capabilities. Part of the AI2 Reasoning Challenge dataset containing 5,197 questions that test scientific reasoning and factual knowledge. The Easy Set contains questions that are answerable by retrieval-based and word co-occurrence algorithms, making them more accessible than the Challenge Set.", + "metric_id": "llm_stats.arc-e.score", + "metric_name": "ARC-E score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arc-e", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ARC-E", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.758, + "details": { + "raw_score": "0.758", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e2b", + "raw_benchmark_id": "arc-e", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e2b\",\"https://llm-stats.com/benchmarks/arc-e\",\"https://api.llm-stats.com/leaderboard/benchmarks/arc-e\"]", + "raw_score_id": "arc-e::gemma-3n-e2b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "big-bench-hard::big-bench-hard-gemma-3n-e2b", + "evaluation_name": "llm_stats.big-bench-hard", + "source_data": { + "dataset_name": "BIG-Bench Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e2b", + "https://llm-stats.com/benchmarks/big-bench-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard" + ], + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_model_id": "gemma-3n-e2b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIG-Bench Hard (BBH) is a subset of 23 challenging BIG-Bench tasks selected because prior language model evaluations did not outperform average human-rater performance. The benchmark contains 6,511 evaluation examples testing various forms of multi-step reasoning including arithmetic, logical reasoning (Boolean expressions, logical deduction), geometric reasoning, temporal reasoning, and language understanding. Tasks require capabilities such as causal judgment, object counting, navigation, pattern recognition, and complex problem solving.", + "metric_id": "llm_stats.big-bench-hard.score", + "metric_name": "BIG-Bench Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench Hard", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "21" + } + }, + "score_details": { + "score": 0.443, + "details": { + "raw_score": "0.443", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e2b", + "raw_benchmark_id": "big-bench-hard", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e2b\",\"https://llm-stats.com/benchmarks/big-bench-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard\"]", + "raw_score_id": "big-bench-hard::gemma-3n-e2b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "boolq::boolq-gemma-3n-e2b", + "evaluation_name": "llm_stats.boolq", + "source_data": { + "dataset_name": "BoolQ", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e2b", + "https://llm-stats.com/benchmarks/boolq", + "https://api.llm-stats.com/leaderboard/benchmarks/boolq" + ], + "additional_details": { + "raw_benchmark_id": "boolq", + "raw_model_id": "gemma-3n-e2b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BoolQ is a reading comprehension dataset for yes/no questions containing 15,942 naturally occurring examples. Each example consists of a question, passage, and boolean answer, where questions are generated in unprompted and unconstrained settings. The dataset challenges models with complex, non-factoid information requiring entailment-like inference to solve.", + "metric_id": "llm_stats.boolq.score", + "metric_name": "BoolQ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "boolq", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BoolQ", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.764, + "details": { + "raw_score": "0.764", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e2b", + "raw_benchmark_id": "boolq", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e2b\",\"https://llm-stats.com/benchmarks/boolq\",\"https://api.llm-stats.com/leaderboard/benchmarks/boolq\"]", + "raw_score_id": "boolq::gemma-3n-e2b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "natural-questions::natural-questions-gemma-3n-e2b", + "evaluation_name": "llm_stats.natural-questions", + "source_data": { + "dataset_name": "Natural Questions", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e2b", + "https://llm-stats.com/benchmarks/natural-questions", + "https://api.llm-stats.com/leaderboard/benchmarks/natural-questions" + ], + "additional_details": { + "raw_benchmark_id": "natural-questions", + "raw_model_id": "gemma-3n-e2b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Natural Questions is a question answering dataset featuring real anonymized queries issued to Google search engine. It contains 307,373 training examples where annotators provide long answers (passages) and short answers (entities) from Wikipedia pages, or mark them as unanswerable.", + "metric_id": "llm_stats.natural-questions.score", + "metric_name": "Natural Questions score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "natural-questions", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Natural Questions", + "raw_categories": "[\"reasoning\",\"search\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.155, + "details": { + "raw_score": "0.155", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e2b", + "raw_benchmark_id": "natural-questions", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e2b\",\"https://llm-stats.com/benchmarks/natural-questions\",\"https://api.llm-stats.com/leaderboard/benchmarks/natural-questions\"]", + "raw_score_id": "natural-questions::gemma-3n-e2b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "piqa::piqa-gemma-3n-e2b", + "evaluation_name": "llm_stats.piqa", + "source_data": { + "dataset_name": "PIQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e2b", + "https://llm-stats.com/benchmarks/piqa", + "https://api.llm-stats.com/leaderboard/benchmarks/piqa" + ], + "additional_details": { + "raw_benchmark_id": "piqa", + "raw_model_id": "gemma-3n-e2b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "PIQA (Physical Interaction: Question Answering) is a benchmark dataset for physical commonsense reasoning in natural language. It tests AI systems' ability to answer questions requiring physical world knowledge through multiple choice questions with everyday situations, focusing on atypical solutions inspired by instructables.com. The dataset contains 21,000 multiple choice questions where models must choose the most appropriate solution for physical interactions.", + "metric_id": "llm_stats.piqa.score", + "metric_name": "PIQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "piqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PIQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.789, + "details": { + "raw_score": "0.789", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e2b", + "raw_benchmark_id": "piqa", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e2b\",\"https://llm-stats.com/benchmarks/piqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/piqa\"]", + "raw_score_id": "piqa::gemma-3n-e2b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "social-iqa::social-iqa-gemma-3n-e2b", + "evaluation_name": "llm_stats.social-iqa", + "source_data": { + "dataset_name": "Social IQa", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e2b", + "https://llm-stats.com/benchmarks/social-iqa", + "https://api.llm-stats.com/leaderboard/benchmarks/social-iqa" + ], + "additional_details": { + "raw_benchmark_id": "social-iqa", + "raw_model_id": "gemma-3n-e2b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The first large-scale benchmark for commonsense reasoning about social situations. Contains 38,000 multiple choice questions probing emotional and social intelligence in everyday situations, testing commonsense understanding of social interactions and theory of mind reasoning about the implied emotions and behavior of others.", + "metric_id": "llm_stats.social-iqa.score", + "metric_name": "Social IQa score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "social-iqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Social IQa", + "raw_categories": "[\"psychology\",\"reasoning\",\"creativity\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.488, + "details": { + "raw_score": "0.488", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e2b", + "raw_benchmark_id": "social-iqa", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e2b\",\"https://llm-stats.com/benchmarks/social-iqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/social-iqa\"]", + "raw_score_id": "social-iqa::gemma-3n-e2b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "triviaqa::triviaqa-gemma-3n-e2b", + "evaluation_name": "llm_stats.triviaqa", + "source_data": { + "dataset_name": "TriviaQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e2b", + "https://llm-stats.com/benchmarks/triviaqa", + "https://api.llm-stats.com/leaderboard/benchmarks/triviaqa" + ], + "additional_details": { + "raw_benchmark_id": "triviaqa", + "raw_model_id": "gemma-3n-e2b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A large-scale reading comprehension dataset containing over 650K question-answer-evidence triples. TriviaQA includes 95K question-answer pairs authored by trivia enthusiasts and independently gathered evidence documents (six per question on average) that provide high quality distant supervision for answering the questions. The dataset features relatively complex, compositional questions with considerable syntactic and lexical variability, requiring cross-sentence reasoning to find answers.", + "metric_id": "llm_stats.triviaqa.score", + "metric_name": "TriviaQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "triviaqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TriviaQA", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.608, + "details": { + "raw_score": "0.608", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e2b", + "raw_benchmark_id": "triviaqa", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e2b\",\"https://llm-stats.com/benchmarks/triviaqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/triviaqa\"]", + "raw_score_id": "triviaqa::gemma-3n-e2b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "winogrande::winogrande-gemma-3n-e2b", + "evaluation_name": "llm_stats.winogrande", + "source_data": { + "dataset_name": "Winogrande", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e2b", + "https://llm-stats.com/benchmarks/winogrande", + "https://api.llm-stats.com/leaderboard/benchmarks/winogrande" + ], + "additional_details": { + "raw_benchmark_id": "winogrande", + "raw_model_id": "gemma-3n-e2b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WinoGrande: An Adversarial Winograd Schema Challenge at Scale. A large-scale dataset of 44,000 pronoun resolution problems designed to test machine commonsense reasoning. Uses adversarial filtering to reduce spurious biases and provides a more robust evaluation of whether AI systems truly understand commonsense or exploit statistical shortcuts. Current best AI methods achieve 59.4-79.1% accuracy, significantly below human performance of 94.0%.", + "metric_id": "llm_stats.winogrande.score", + "metric_name": "Winogrande score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "winogrande", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Winogrande", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "22" + } + }, + "score_details": { + "score": 0.668, + "details": { + "raw_score": "0.668", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e2b", + "raw_benchmark_id": "winogrande", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e2b\",\"https://llm-stats.com/benchmarks/winogrande\",\"https://api.llm-stats.com/leaderboard/benchmarks/winogrande\"]", + "raw_score_id": "winogrande::gemma-3n-e2b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/google/gemma-3n-e4b-it-litert-preview/d08921f8-d822-48ba-ba23-3da59d422812.json b/data/llm_stats/google/gemma-3n-e4b-it-litert-preview/d08921f8-d822-48ba-ba23-3da59d422812.json new file mode 100644 index 0000000000000000000000000000000000000000..d46776ee532816b1b0c2fb8fa5ceb782d1365d57 --- /dev/null +++ b/data/llm_stats/google/gemma-3n-e4b-it-litert-preview/d08921f8-d822-48ba-ba23-3da59d422812.json @@ -0,0 +1,879 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/google_gemma-3n-e4b-it-litert-preview/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Gemma 3n E4B Instructed LiteRT Preview", + "id": "google/gemma-3n-e4b-it-litert-preview", + "developer": "google", + "additional_details": { + "raw_id": "gemma-3n-e4b-it-litert-preview", + "raw_name": "Gemma 3n E4B Instructed LiteRT Preview", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "raw_model_name": "Gemma 3n E4B Instructed LiteRT Preview", + "raw_organization_id": "google", + "raw_organization_name": "Google", + "raw_release_date": "2025-05-20", + "raw_announcement_date": "2025-05-20", + "raw_multimodal": "true", + "raw_param_count": "1910000000", + "raw_provider_slug": "google", + "raw_provider_name": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "arc-e::arc-e-gemma-3n-e4b-it-litert-preview", + "evaluation_name": "llm_stats.arc-e", + "source_data": { + "dataset_name": "ARC-E", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview", + "https://llm-stats.com/benchmarks/arc-e", + "https://api.llm-stats.com/leaderboard/benchmarks/arc-e" + ], + "additional_details": { + "raw_benchmark_id": "arc-e", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ARC-E (AI2 Reasoning Challenge - Easy Set) is a subset of grade-school level, multiple-choice science questions that requires knowledge and reasoning capabilities. Part of the AI2 Reasoning Challenge dataset containing 5,197 questions that test scientific reasoning and factual knowledge. The Easy Set contains questions that are answerable by retrieval-based and word co-occurrence algorithms, making them more accessible than the Challenge Set.", + "metric_id": "llm_stats.arc-e.score", + "metric_name": "ARC-E score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arc-e", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ARC-E", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.816, + "details": { + "raw_score": "0.816", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "raw_benchmark_id": "arc-e", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview\",\"https://llm-stats.com/benchmarks/arc-e\",\"https://api.llm-stats.com/leaderboard/benchmarks/arc-e\"]", + "raw_score_id": "arc-e::gemma-3n-e4b-it-litert-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "big-bench-hard::big-bench-hard-gemma-3n-e4b-it-litert-preview", + "evaluation_name": "llm_stats.big-bench-hard", + "source_data": { + "dataset_name": "BIG-Bench Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview", + "https://llm-stats.com/benchmarks/big-bench-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard" + ], + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIG-Bench Hard (BBH) is a subset of 23 challenging BIG-Bench tasks selected because prior language model evaluations did not outperform average human-rater performance. The benchmark contains 6,511 evaluation examples testing various forms of multi-step reasoning including arithmetic, logical reasoning (Boolean expressions, logical deduction), geometric reasoning, temporal reasoning, and language understanding. Tasks require capabilities such as causal judgment, object counting, navigation, pattern recognition, and complex problem solving.", + "metric_id": "llm_stats.big-bench-hard.score", + "metric_name": "BIG-Bench Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench Hard", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "21" + } + }, + "score_details": { + "score": 0.529, + "details": { + "raw_score": "0.529", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "raw_benchmark_id": "big-bench-hard", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview\",\"https://llm-stats.com/benchmarks/big-bench-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard\"]", + "raw_score_id": "big-bench-hard::gemma-3n-e4b-it-litert-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "boolq::boolq-gemma-3n-e4b-it-litert-preview", + "evaluation_name": "llm_stats.boolq", + "source_data": { + "dataset_name": "BoolQ", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview", + "https://llm-stats.com/benchmarks/boolq", + "https://api.llm-stats.com/leaderboard/benchmarks/boolq" + ], + "additional_details": { + "raw_benchmark_id": "boolq", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BoolQ is a reading comprehension dataset for yes/no questions containing 15,942 naturally occurring examples. Each example consists of a question, passage, and boolean answer, where questions are generated in unprompted and unconstrained settings. The dataset challenges models with complex, non-factoid information requiring entailment-like inference to solve.", + "metric_id": "llm_stats.boolq.score", + "metric_name": "BoolQ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "boolq", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BoolQ", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.816, + "details": { + "raw_score": "0.816", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "raw_benchmark_id": "boolq", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview\",\"https://llm-stats.com/benchmarks/boolq\",\"https://api.llm-stats.com/leaderboard/benchmarks/boolq\"]", + "raw_score_id": "boolq::gemma-3n-e4b-it-litert-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "codegolf-v2.2::codegolf-v2.2-gemma-3n-e4b-it-litert-preview", + "evaluation_name": "llm_stats.codegolf-v2.2", + "source_data": { + "dataset_name": "Codegolf v2.2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview", + "https://llm-stats.com/benchmarks/codegolf-v2.2", + "https://api.llm-stats.com/leaderboard/benchmarks/codegolf-v2.2" + ], + "additional_details": { + "raw_benchmark_id": "codegolf-v2.2", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Codegolf v2.2 benchmark", + "metric_id": "llm_stats.codegolf-v2.2.score", + "metric_name": "Codegolf v2.2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "codegolf-v2.2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Codegolf v2.2", + "raw_categories": "[\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.168, + "details": { + "raw_score": "0.168", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "raw_benchmark_id": "codegolf-v2.2", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview\",\"https://llm-stats.com/benchmarks/codegolf-v2.2\",\"https://api.llm-stats.com/leaderboard/benchmarks/codegolf-v2.2\"]", + "raw_score_id": "codegolf-v2.2::gemma-3n-e4b-it-litert-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "eclektic::eclektic-gemma-3n-e4b-it-litert-preview", + "evaluation_name": "llm_stats.eclektic", + "source_data": { + "dataset_name": "ECLeKTic", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview", + "https://llm-stats.com/benchmarks/eclektic", + "https://api.llm-stats.com/leaderboard/benchmarks/eclektic" + ], + "additional_details": { + "raw_benchmark_id": "eclektic", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multilingual closed-book question answering dataset that evaluates cross-lingual knowledge transfer in large language models across 12 languages, using knowledge-seeking questions based on Wikipedia articles that exist only in one language", + "metric_id": "llm_stats.eclektic.score", + "metric_name": "ECLeKTic score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "eclektic", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ECLeKTic", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.019, + "details": { + "raw_score": "0.019", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "raw_benchmark_id": "eclektic", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview\",\"https://llm-stats.com/benchmarks/eclektic\",\"https://api.llm-stats.com/leaderboard/benchmarks/eclektic\"]", + "raw_score_id": "eclektic::gemma-3n-e4b-it-litert-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "global-mmlu-lite::global-mmlu-lite-gemma-3n-e4b-it-litert-preview", + "evaluation_name": "llm_stats.global-mmlu-lite", + "source_data": { + "dataset_name": "Global-MMLU-Lite", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview", + "https://llm-stats.com/benchmarks/global-mmlu-lite", + "https://api.llm-stats.com/leaderboard/benchmarks/global-mmlu-lite" + ], + "additional_details": { + "raw_benchmark_id": "global-mmlu-lite", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A lightweight version of Global MMLU benchmark that evaluates language models across multiple languages while addressing cultural and linguistic biases in multilingual evaluation.", + "metric_id": "llm_stats.global-mmlu-lite.score", + "metric_name": "Global-MMLU-Lite score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "global-mmlu-lite", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Global-MMLU-Lite", + "raw_categories": "[\"language\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "14" + } + }, + "score_details": { + "score": 0.645, + "details": { + "raw_score": "0.645", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "raw_benchmark_id": "global-mmlu-lite", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview\",\"https://llm-stats.com/benchmarks/global-mmlu-lite\",\"https://api.llm-stats.com/leaderboard/benchmarks/global-mmlu-lite\"]", + "raw_score_id": "global-mmlu-lite::gemma-3n-e4b-it-litert-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "global-mmlu::global-mmlu-gemma-3n-e4b-it-litert-preview", + "evaluation_name": "llm_stats.global-mmlu", + "source_data": { + "dataset_name": "Global-MMLU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview", + "https://llm-stats.com/benchmarks/global-mmlu", + "https://api.llm-stats.com/leaderboard/benchmarks/global-mmlu" + ], + "additional_details": { + "raw_benchmark_id": "global-mmlu", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A comprehensive multilingual benchmark covering 42 languages that addresses cultural and linguistic biases in evaluation, with improved translation quality and culturally sensitive question subsets.", + "metric_id": "llm_stats.global-mmlu.score", + "metric_name": "Global-MMLU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "global-mmlu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Global-MMLU", + "raw_categories": "[\"language\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.603, + "details": { + "raw_score": "0.603", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "raw_benchmark_id": "global-mmlu", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview\",\"https://llm-stats.com/benchmarks/global-mmlu\",\"https://api.llm-stats.com/leaderboard/benchmarks/global-mmlu\"]", + "raw_score_id": "global-mmlu::gemma-3n-e4b-it-litert-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hellaswag::hellaswag-gemma-3n-e4b-it-litert-preview", + "evaluation_name": "llm_stats.hellaswag", + "source_data": { + "dataset_name": "HellaSwag", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview", + "https://llm-stats.com/benchmarks/hellaswag", + "https://api.llm-stats.com/leaderboard/benchmarks/hellaswag" + ], + "additional_details": { + "raw_benchmark_id": "hellaswag", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A challenging commonsense natural language inference dataset that uses Adversarial Filtering to create questions trivial for humans (>95% accuracy) but difficult for state-of-the-art models, requiring completion of sentence endings based on physical situations and everyday activities", + "metric_id": "llm_stats.hellaswag.score", + "metric_name": "HellaSwag score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hellaswag", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HellaSwag", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.786, + "details": { + "raw_score": "0.786", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "raw_benchmark_id": "hellaswag", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview\",\"https://llm-stats.com/benchmarks/hellaswag\",\"https://api.llm-stats.com/leaderboard/benchmarks/hellaswag\"]", + "raw_score_id": "hellaswag::gemma-3n-e4b-it-litert-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hiddenmath::hiddenmath-gemma-3n-e4b-it-litert-preview", + "evaluation_name": "llm_stats.hiddenmath", + "source_data": { + "dataset_name": "HiddenMath", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview", + "https://llm-stats.com/benchmarks/hiddenmath", + "https://api.llm-stats.com/leaderboard/benchmarks/hiddenmath" + ], + "additional_details": { + "raw_benchmark_id": "hiddenmath", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Google DeepMind's internal mathematical reasoning benchmark that introduces novel problems not encountered during model training to evaluate true mathematical reasoning capabilities rather than memorization", + "metric_id": "llm_stats.hiddenmath.score", + "metric_name": "HiddenMath score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hiddenmath", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HiddenMath", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.377, + "details": { + "raw_score": "0.377", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "raw_benchmark_id": "hiddenmath", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview\",\"https://llm-stats.com/benchmarks/hiddenmath\",\"https://api.llm-stats.com/leaderboard/benchmarks/hiddenmath\"]", + "raw_score_id": "hiddenmath::gemma-3n-e4b-it-litert-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "livecodebench-v5::livecodebench-v5-gemma-3n-e4b-it-litert-preview", + "evaluation_name": "llm_stats.livecodebench-v5", + "source_data": { + "dataset_name": "LiveCodeBench v5", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview", + "https://llm-stats.com/benchmarks/livecodebench-v5", + "https://api.llm-stats.com/leaderboard/benchmarks/livecodebench-v5" + ], + "additional_details": { + "raw_benchmark_id": "livecodebench-v5", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LiveCodeBench is a holistic and contamination-free evaluation benchmark for large language models for code. It continuously collects new problems from programming contests (LeetCode, AtCoder, CodeForces) and evaluates four different scenarios: code generation, self-repair, code execution, and test output prediction. Problems are annotated with release dates to enable evaluation on unseen problems released after a model's training cutoff.", + "metric_id": "llm_stats.livecodebench-v5.score", + "metric_name": "LiveCodeBench v5 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "livecodebench-v5", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LiveCodeBench v5", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.257, + "details": { + "raw_score": "0.257", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "raw_benchmark_id": "livecodebench-v5", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview\",\"https://llm-stats.com/benchmarks/livecodebench-v5\",\"https://api.llm-stats.com/leaderboard/benchmarks/livecodebench-v5\"]", + "raw_score_id": "livecodebench-v5::gemma-3n-e4b-it-litert-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "natural-questions::natural-questions-gemma-3n-e4b-it-litert-preview", + "evaluation_name": "llm_stats.natural-questions", + "source_data": { + "dataset_name": "Natural Questions", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview", + "https://llm-stats.com/benchmarks/natural-questions", + "https://api.llm-stats.com/leaderboard/benchmarks/natural-questions" + ], + "additional_details": { + "raw_benchmark_id": "natural-questions", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Natural Questions is a question answering dataset featuring real anonymized queries issued to Google search engine. It contains 307,373 training examples where annotators provide long answers (passages) and short answers (entities) from Wikipedia pages, or mark them as unanswerable.", + "metric_id": "llm_stats.natural-questions.score", + "metric_name": "Natural Questions score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "natural-questions", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Natural Questions", + "raw_categories": "[\"reasoning\",\"search\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.209, + "details": { + "raw_score": "0.209", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "raw_benchmark_id": "natural-questions", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview\",\"https://llm-stats.com/benchmarks/natural-questions\",\"https://api.llm-stats.com/leaderboard/benchmarks/natural-questions\"]", + "raw_score_id": "natural-questions::gemma-3n-e4b-it-litert-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "piqa::piqa-gemma-3n-e4b-it-litert-preview", + "evaluation_name": "llm_stats.piqa", + "source_data": { + "dataset_name": "PIQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview", + "https://llm-stats.com/benchmarks/piqa", + "https://api.llm-stats.com/leaderboard/benchmarks/piqa" + ], + "additional_details": { + "raw_benchmark_id": "piqa", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "PIQA (Physical Interaction: Question Answering) is a benchmark dataset for physical commonsense reasoning in natural language. It tests AI systems' ability to answer questions requiring physical world knowledge through multiple choice questions with everyday situations, focusing on atypical solutions inspired by instructables.com. The dataset contains 21,000 multiple choice questions where models must choose the most appropriate solution for physical interactions.", + "metric_id": "llm_stats.piqa.score", + "metric_name": "PIQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "piqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PIQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.81, + "details": { + "raw_score": "0.81", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "raw_benchmark_id": "piqa", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview\",\"https://llm-stats.com/benchmarks/piqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/piqa\"]", + "raw_score_id": "piqa::gemma-3n-e4b-it-litert-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "social-iqa::social-iqa-gemma-3n-e4b-it-litert-preview", + "evaluation_name": "llm_stats.social-iqa", + "source_data": { + "dataset_name": "Social IQa", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview", + "https://llm-stats.com/benchmarks/social-iqa", + "https://api.llm-stats.com/leaderboard/benchmarks/social-iqa" + ], + "additional_details": { + "raw_benchmark_id": "social-iqa", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The first large-scale benchmark for commonsense reasoning about social situations. Contains 38,000 multiple choice questions probing emotional and social intelligence in everyday situations, testing commonsense understanding of social interactions and theory of mind reasoning about the implied emotions and behavior of others.", + "metric_id": "llm_stats.social-iqa.score", + "metric_name": "Social IQa score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "social-iqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Social IQa", + "raw_categories": "[\"psychology\",\"reasoning\",\"creativity\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.5, + "details": { + "raw_score": "0.5", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "raw_benchmark_id": "social-iqa", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview\",\"https://llm-stats.com/benchmarks/social-iqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/social-iqa\"]", + "raw_score_id": "social-iqa::gemma-3n-e4b-it-litert-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "triviaqa::triviaqa-gemma-3n-e4b-it-litert-preview", + "evaluation_name": "llm_stats.triviaqa", + "source_data": { + "dataset_name": "TriviaQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview", + "https://llm-stats.com/benchmarks/triviaqa", + "https://api.llm-stats.com/leaderboard/benchmarks/triviaqa" + ], + "additional_details": { + "raw_benchmark_id": "triviaqa", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A large-scale reading comprehension dataset containing over 650K question-answer-evidence triples. TriviaQA includes 95K question-answer pairs authored by trivia enthusiasts and independently gathered evidence documents (six per question on average) that provide high quality distant supervision for answering the questions. The dataset features relatively complex, compositional questions with considerable syntactic and lexical variability, requiring cross-sentence reasoning to find answers.", + "metric_id": "llm_stats.triviaqa.score", + "metric_name": "TriviaQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "triviaqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TriviaQA", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.702, + "details": { + "raw_score": "0.702", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "raw_benchmark_id": "triviaqa", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview\",\"https://llm-stats.com/benchmarks/triviaqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/triviaqa\"]", + "raw_score_id": "triviaqa::gemma-3n-e4b-it-litert-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "winogrande::winogrande-gemma-3n-e4b-it-litert-preview", + "evaluation_name": "llm_stats.winogrande", + "source_data": { + "dataset_name": "Winogrande", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview", + "https://llm-stats.com/benchmarks/winogrande", + "https://api.llm-stats.com/leaderboard/benchmarks/winogrande" + ], + "additional_details": { + "raw_benchmark_id": "winogrande", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WinoGrande: An Adversarial Winograd Schema Challenge at Scale. A large-scale dataset of 44,000 pronoun resolution problems designed to test machine commonsense reasoning. Uses adversarial filtering to reduce spurious biases and provides a more robust evaluation of whether AI systems truly understand commonsense or exploit statistical shortcuts. Current best AI methods achieve 59.4-79.1% accuracy, significantly below human performance of 94.0%.", + "metric_id": "llm_stats.winogrande.score", + "metric_name": "Winogrande score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "winogrande", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Winogrande", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "22" + } + }, + "score_details": { + "score": 0.717, + "details": { + "raw_score": "0.717", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "raw_benchmark_id": "winogrande", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview\",\"https://llm-stats.com/benchmarks/winogrande\",\"https://api.llm-stats.com/leaderboard/benchmarks/winogrande\"]", + "raw_score_id": "winogrande::gemma-3n-e4b-it-litert-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "wmt24::wmt24-gemma-3n-e4b-it-litert-preview", + "evaluation_name": "llm_stats.wmt24", + "source_data": { + "dataset_name": "WMT24++", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview", + "https://llm-stats.com/benchmarks/wmt24", + "https://api.llm-stats.com/leaderboard/benchmarks/wmt24++" + ], + "additional_details": { + "raw_benchmark_id": "wmt24++", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WMT24++ is a comprehensive multilingual machine translation benchmark that expands the WMT24 dataset to cover 55 languages and dialects. It includes human-written references and post-edits across four domains (literary, news, social, and speech) to evaluate machine translation systems and large language models across diverse linguistic contexts.", + "metric_id": "llm_stats.wmt24.score", + "metric_name": "WMT24++ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "wmt24++", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WMT24++", + "raw_categories": "[\"language\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.501, + "details": { + "raw_score": "0.501", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e4b-it-litert-preview", + "raw_benchmark_id": "wmt24++", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e4b-it-litert-preview\",\"https://llm-stats.com/benchmarks/wmt24\",\"https://api.llm-stats.com/leaderboard/benchmarks/wmt24++\"]", + "raw_score_id": "wmt24++::gemma-3n-e4b-it-litert-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/google/gemma-3n-e4b/58826b80-1ee0-4a56-9595-af34153a3292.json b/data/llm_stats/google/gemma-3n-e4b/58826b80-1ee0-4a56-9595-af34153a3292.json new file mode 100644 index 0000000000000000000000000000000000000000..9408518e0f4a48626fe57eff81f918d804a1fa58 --- /dev/null +++ b/data/llm_stats/google/gemma-3n-e4b/58826b80-1ee0-4a56-9595-af34153a3292.json @@ -0,0 +1,515 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/google_gemma-3n-e4b/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Gemma 3n E4B", + "id": "google/gemma-3n-e4b", + "developer": "google", + "additional_details": { + "raw_id": "gemma-3n-e4b", + "raw_name": "Gemma 3n E4B", + "raw_model_id": "gemma-3n-e4b", + "raw_model_name": "Gemma 3n E4B", + "raw_organization_id": "google", + "raw_organization_name": "Google", + "raw_release_date": "2025-06-26", + "raw_announcement_date": "2025-06-26", + "raw_multimodal": "true", + "raw_param_count": "8000000000", + "raw_provider_slug": "google", + "raw_provider_name": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "arc-e::arc-e-gemma-3n-e4b", + "evaluation_name": "llm_stats.arc-e", + "source_data": { + "dataset_name": "ARC-E", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e4b", + "https://llm-stats.com/benchmarks/arc-e", + "https://api.llm-stats.com/leaderboard/benchmarks/arc-e" + ], + "additional_details": { + "raw_benchmark_id": "arc-e", + "raw_model_id": "gemma-3n-e4b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ARC-E (AI2 Reasoning Challenge - Easy Set) is a subset of grade-school level, multiple-choice science questions that requires knowledge and reasoning capabilities. Part of the AI2 Reasoning Challenge dataset containing 5,197 questions that test scientific reasoning and factual knowledge. The Easy Set contains questions that are answerable by retrieval-based and word co-occurrence algorithms, making them more accessible than the Challenge Set.", + "metric_id": "llm_stats.arc-e.score", + "metric_name": "ARC-E score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arc-e", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ARC-E", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.816, + "details": { + "raw_score": "0.816", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e4b", + "raw_benchmark_id": "arc-e", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e4b\",\"https://llm-stats.com/benchmarks/arc-e\",\"https://api.llm-stats.com/leaderboard/benchmarks/arc-e\"]", + "raw_score_id": "arc-e::gemma-3n-e4b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "big-bench-hard::big-bench-hard-gemma-3n-e4b", + "evaluation_name": "llm_stats.big-bench-hard", + "source_data": { + "dataset_name": "BIG-Bench Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e4b", + "https://llm-stats.com/benchmarks/big-bench-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard" + ], + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_model_id": "gemma-3n-e4b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIG-Bench Hard (BBH) is a subset of 23 challenging BIG-Bench tasks selected because prior language model evaluations did not outperform average human-rater performance. The benchmark contains 6,511 evaluation examples testing various forms of multi-step reasoning including arithmetic, logical reasoning (Boolean expressions, logical deduction), geometric reasoning, temporal reasoning, and language understanding. Tasks require capabilities such as causal judgment, object counting, navigation, pattern recognition, and complex problem solving.", + "metric_id": "llm_stats.big-bench-hard.score", + "metric_name": "BIG-Bench Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench Hard", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "21" + } + }, + "score_details": { + "score": 0.529, + "details": { + "raw_score": "0.529", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e4b", + "raw_benchmark_id": "big-bench-hard", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e4b\",\"https://llm-stats.com/benchmarks/big-bench-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard\"]", + "raw_score_id": "big-bench-hard::gemma-3n-e4b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "boolq::boolq-gemma-3n-e4b", + "evaluation_name": "llm_stats.boolq", + "source_data": { + "dataset_name": "BoolQ", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e4b", + "https://llm-stats.com/benchmarks/boolq", + "https://api.llm-stats.com/leaderboard/benchmarks/boolq" + ], + "additional_details": { + "raw_benchmark_id": "boolq", + "raw_model_id": "gemma-3n-e4b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BoolQ is a reading comprehension dataset for yes/no questions containing 15,942 naturally occurring examples. Each example consists of a question, passage, and boolean answer, where questions are generated in unprompted and unconstrained settings. The dataset challenges models with complex, non-factoid information requiring entailment-like inference to solve.", + "metric_id": "llm_stats.boolq.score", + "metric_name": "BoolQ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "boolq", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BoolQ", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.816, + "details": { + "raw_score": "0.816", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e4b", + "raw_benchmark_id": "boolq", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e4b\",\"https://llm-stats.com/benchmarks/boolq\",\"https://api.llm-stats.com/leaderboard/benchmarks/boolq\"]", + "raw_score_id": "boolq::gemma-3n-e4b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hellaswag::hellaswag-gemma-3n-e4b", + "evaluation_name": "llm_stats.hellaswag", + "source_data": { + "dataset_name": "HellaSwag", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e4b", + "https://llm-stats.com/benchmarks/hellaswag", + "https://api.llm-stats.com/leaderboard/benchmarks/hellaswag" + ], + "additional_details": { + "raw_benchmark_id": "hellaswag", + "raw_model_id": "gemma-3n-e4b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A challenging commonsense natural language inference dataset that uses Adversarial Filtering to create questions trivial for humans (>95% accuracy) but difficult for state-of-the-art models, requiring completion of sentence endings based on physical situations and everyday activities", + "metric_id": "llm_stats.hellaswag.score", + "metric_name": "HellaSwag score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hellaswag", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HellaSwag", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.786, + "details": { + "raw_score": "0.786", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e4b", + "raw_benchmark_id": "hellaswag", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e4b\",\"https://llm-stats.com/benchmarks/hellaswag\",\"https://api.llm-stats.com/leaderboard/benchmarks/hellaswag\"]", + "raw_score_id": "hellaswag::gemma-3n-e4b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "natural-questions::natural-questions-gemma-3n-e4b", + "evaluation_name": "llm_stats.natural-questions", + "source_data": { + "dataset_name": "Natural Questions", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e4b", + "https://llm-stats.com/benchmarks/natural-questions", + "https://api.llm-stats.com/leaderboard/benchmarks/natural-questions" + ], + "additional_details": { + "raw_benchmark_id": "natural-questions", + "raw_model_id": "gemma-3n-e4b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Natural Questions is a question answering dataset featuring real anonymized queries issued to Google search engine. It contains 307,373 training examples where annotators provide long answers (passages) and short answers (entities) from Wikipedia pages, or mark them as unanswerable.", + "metric_id": "llm_stats.natural-questions.score", + "metric_name": "Natural Questions score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "natural-questions", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Natural Questions", + "raw_categories": "[\"reasoning\",\"search\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.209, + "details": { + "raw_score": "0.209", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e4b", + "raw_benchmark_id": "natural-questions", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e4b\",\"https://llm-stats.com/benchmarks/natural-questions\",\"https://api.llm-stats.com/leaderboard/benchmarks/natural-questions\"]", + "raw_score_id": "natural-questions::gemma-3n-e4b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "piqa::piqa-gemma-3n-e4b", + "evaluation_name": "llm_stats.piqa", + "source_data": { + "dataset_name": "PIQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e4b", + "https://llm-stats.com/benchmarks/piqa", + "https://api.llm-stats.com/leaderboard/benchmarks/piqa" + ], + "additional_details": { + "raw_benchmark_id": "piqa", + "raw_model_id": "gemma-3n-e4b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "PIQA (Physical Interaction: Question Answering) is a benchmark dataset for physical commonsense reasoning in natural language. It tests AI systems' ability to answer questions requiring physical world knowledge through multiple choice questions with everyday situations, focusing on atypical solutions inspired by instructables.com. The dataset contains 21,000 multiple choice questions where models must choose the most appropriate solution for physical interactions.", + "metric_id": "llm_stats.piqa.score", + "metric_name": "PIQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "piqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PIQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.81, + "details": { + "raw_score": "0.81", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e4b", + "raw_benchmark_id": "piqa", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e4b\",\"https://llm-stats.com/benchmarks/piqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/piqa\"]", + "raw_score_id": "piqa::gemma-3n-e4b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "social-iqa::social-iqa-gemma-3n-e4b", + "evaluation_name": "llm_stats.social-iqa", + "source_data": { + "dataset_name": "Social IQa", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e4b", + "https://llm-stats.com/benchmarks/social-iqa", + "https://api.llm-stats.com/leaderboard/benchmarks/social-iqa" + ], + "additional_details": { + "raw_benchmark_id": "social-iqa", + "raw_model_id": "gemma-3n-e4b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The first large-scale benchmark for commonsense reasoning about social situations. Contains 38,000 multiple choice questions probing emotional and social intelligence in everyday situations, testing commonsense understanding of social interactions and theory of mind reasoning about the implied emotions and behavior of others.", + "metric_id": "llm_stats.social-iqa.score", + "metric_name": "Social IQa score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "social-iqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Social IQa", + "raw_categories": "[\"psychology\",\"reasoning\",\"creativity\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.5, + "details": { + "raw_score": "0.5", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e4b", + "raw_benchmark_id": "social-iqa", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e4b\",\"https://llm-stats.com/benchmarks/social-iqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/social-iqa\"]", + "raw_score_id": "social-iqa::gemma-3n-e4b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "triviaqa::triviaqa-gemma-3n-e4b", + "evaluation_name": "llm_stats.triviaqa", + "source_data": { + "dataset_name": "TriviaQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e4b", + "https://llm-stats.com/benchmarks/triviaqa", + "https://api.llm-stats.com/leaderboard/benchmarks/triviaqa" + ], + "additional_details": { + "raw_benchmark_id": "triviaqa", + "raw_model_id": "gemma-3n-e4b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A large-scale reading comprehension dataset containing over 650K question-answer-evidence triples. TriviaQA includes 95K question-answer pairs authored by trivia enthusiasts and independently gathered evidence documents (six per question on average) that provide high quality distant supervision for answering the questions. The dataset features relatively complex, compositional questions with considerable syntactic and lexical variability, requiring cross-sentence reasoning to find answers.", + "metric_id": "llm_stats.triviaqa.score", + "metric_name": "TriviaQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "triviaqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TriviaQA", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.702, + "details": { + "raw_score": "0.702", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e4b", + "raw_benchmark_id": "triviaqa", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e4b\",\"https://llm-stats.com/benchmarks/triviaqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/triviaqa\"]", + "raw_score_id": "triviaqa::gemma-3n-e4b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "winogrande::winogrande-gemma-3n-e4b", + "evaluation_name": "llm_stats.winogrande", + "source_data": { + "dataset_name": "Winogrande", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-3n-e4b", + "https://llm-stats.com/benchmarks/winogrande", + "https://api.llm-stats.com/leaderboard/benchmarks/winogrande" + ], + "additional_details": { + "raw_benchmark_id": "winogrande", + "raw_model_id": "gemma-3n-e4b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WinoGrande: An Adversarial Winograd Schema Challenge at Scale. A large-scale dataset of 44,000 pronoun resolution problems designed to test machine commonsense reasoning. Uses adversarial filtering to reduce spurious biases and provides a more robust evaluation of whether AI systems truly understand commonsense or exploit statistical shortcuts. Current best AI methods achieve 59.4-79.1% accuracy, significantly below human performance of 94.0%.", + "metric_id": "llm_stats.winogrande.score", + "metric_name": "Winogrande score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "winogrande", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Winogrande", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "22" + } + }, + "score_details": { + "score": 0.717, + "details": { + "raw_score": "0.717", + "raw_score_field": "score", + "raw_model_id": "gemma-3n-e4b", + "raw_benchmark_id": "winogrande", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-3n-e4b\",\"https://llm-stats.com/benchmarks/winogrande\",\"https://api.llm-stats.com/leaderboard/benchmarks/winogrande\"]", + "raw_score_id": "winogrande::gemma-3n-e4b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/google/gemma-4-12b-it/e2a33bfa-e779-40a2-93a7-09a7af786bbf.json b/data/llm_stats/google/gemma-4-12b-it/e2a33bfa-e779-40a2-93a7-09a7af786bbf.json new file mode 100644 index 0000000000000000000000000000000000000000..a42ea2fa50e9ea10ded21cea9d02484c79348ada --- /dev/null +++ b/data/llm_stats/google/gemma-4-12b-it/e2a33bfa-e779-40a2-93a7-09a7af786bbf.json @@ -0,0 +1,515 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/google_gemma-4-12b-it/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Gemma 4 12B", + "id": "google/gemma-4-12b-it", + "developer": "google", + "additional_details": { + "raw_id": "gemma-4-12b-it", + "raw_name": "Gemma 4 12B", + "raw_model_id": "gemma-4-12b-it", + "raw_model_name": "Gemma 4 12B", + "raw_organization_id": "google", + "raw_organization_name": "Google", + "raw_release_date": "2026-05-23", + "raw_announcement_date": "2026-05-23", + "raw_multimodal": "true", + "raw_param_count": "11959730224", + "raw_provider_slug": "google", + "raw_provider_name": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aime-2026::aime-2026-gemma-4-12b-it", + "evaluation_name": "llm_stats.aime-2026", + "source_data": { + "dataset_name": "AIME 2026", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-12b-it", + "https://llm-stats.com/benchmarks/aime-2026", + "https://api.llm-stats.com/leaderboard/benchmarks/aime-2026" + ], + "additional_details": { + "raw_benchmark_id": "aime-2026", + "raw_model_id": "gemma-4-12b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "All 30 problems from the 2026 American Invitational Mathematics Examination (AIME I and AIME II), testing olympiad-level mathematical reasoning with integer answers from 000-999. Used as an AI benchmark to evaluate large language models' ability to solve complex mathematical problems requiring multi-step logical deductions and structured symbolic reasoning.", + "metric_id": "llm_stats.aime-2026.score", + "metric_name": "AIME 2026 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aime-2026", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AIME 2026", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "17" + } + }, + "score_details": { + "score": 0.775, + "details": { + "raw_score": "0.775", + "raw_score_field": "score", + "raw_model_id": "gemma-4-12b-it", + "raw_benchmark_id": "aime-2026", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-12b-it\",\"https://llm-stats.com/benchmarks/aime-2026\",\"https://api.llm-stats.com/leaderboard/benchmarks/aime-2026\"]", + "raw_score_id": "aime-2026::gemma-4-12b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "big-bench-extra-hard::big-bench-extra-hard-gemma-4-12b-it", + "evaluation_name": "llm_stats.big-bench-extra-hard", + "source_data": { + "dataset_name": "BIG-Bench Extra Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-12b-it", + "https://llm-stats.com/benchmarks/big-bench-extra-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench-extra-hard" + ], + "additional_details": { + "raw_benchmark_id": "big-bench-extra-hard", + "raw_model_id": "gemma-4-12b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIG-Bench Extra Hard (BBEH) is a challenging benchmark that replaces each task in BIG-Bench Hard with a novel task that probes similar reasoning capabilities but exhibits significantly increased difficulty. The benchmark contains 23 tasks testing diverse reasoning skills including many-hop reasoning, causal understanding, spatial reasoning, temporal arithmetic, geometric reasoning, linguistic reasoning, logic puzzles, and humor understanding. Designed to address saturation on existing benchmarks where state-of-the-art models achieve near-perfect scores, BBEH shows substantial room for improvement with best models achieving only 9.8-44.8% average accuracy.", + "metric_id": "llm_stats.big-bench-extra-hard.score", + "metric_name": "BIG-Bench Extra Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench-extra-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench Extra Hard", + "raw_categories": "[\"language\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.53, + "details": { + "raw_score": "0.53", + "raw_score_field": "score", + "raw_model_id": "gemma-4-12b-it", + "raw_benchmark_id": "big-bench-extra-hard", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-12b-it\",\"https://llm-stats.com/benchmarks/big-bench-extra-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench-extra-hard\"]", + "raw_score_id": "big-bench-extra-hard::gemma-4-12b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "codeforces::codeforces-gemma-4-12b-it", + "evaluation_name": "llm_stats.codeforces", + "source_data": { + "dataset_name": "CodeForces", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-12b-it", + "https://llm-stats.com/benchmarks/codeforces", + "https://api.llm-stats.com/leaderboard/benchmarks/codeforces" + ], + "additional_details": { + "raw_benchmark_id": "codeforces", + "raw_model_id": "gemma-4-12b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A competitive programming benchmark using problems from the CodeForces platform. The benchmark evaluates code generation capabilities of LLMs on algorithmic problems with difficulty ratings ranging from 800 to 2400. Problems cover diverse algorithmic categories including dynamic programming, graph algorithms, data structures, and mathematical problems with standardized evaluation through direct platform submission.", + "metric_id": "llm_stats.codeforces.score", + "metric_name": "CodeForces score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "codeforces", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CodeForces", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.553, + "details": { + "raw_score": "0.553", + "raw_score_field": "score", + "raw_model_id": "gemma-4-12b-it", + "raw_benchmark_id": "codeforces", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-12b-it\",\"https://llm-stats.com/benchmarks/codeforces\",\"https://api.llm-stats.com/leaderboard/benchmarks/codeforces\"]", + "raw_score_id": "codeforces::gemma-4-12b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "covost2::covost2-gemma-4-12b-it", + "evaluation_name": "llm_stats.covost2", + "source_data": { + "dataset_name": "CoVoST2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-12b-it", + "https://llm-stats.com/benchmarks/covost2", + "https://api.llm-stats.com/leaderboard/benchmarks/covost2" + ], + "additional_details": { + "raw_benchmark_id": "covost2", + "raw_model_id": "gemma-4-12b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CoVoST 2 is a large-scale multilingual speech translation corpus derived from Common Voice, covering translations from 21 languages into English and from English into 15 languages. The dataset contains 2,880 hours of speech with 78K speakers for speech translation research.", + "metric_id": "llm_stats.covost2.score", + "metric_name": "CoVoST2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "covost2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CoVoST2", + "raw_categories": "[\"language\",\"speech_to_text\",\"audio\"]", + "raw_modality": "audio", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.385, + "details": { + "raw_score": "0.385", + "raw_score_field": "score", + "raw_model_id": "gemma-4-12b-it", + "raw_benchmark_id": "covost2", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-12b-it\",\"https://llm-stats.com/benchmarks/covost2\",\"https://api.llm-stats.com/leaderboard/benchmarks/covost2\"]", + "raw_score_id": "covost2::gemma-4-12b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "fleurs::fleurs-gemma-4-12b-it", + "evaluation_name": "llm_stats.fleurs", + "source_data": { + "dataset_name": "FLEURS", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-12b-it", + "https://llm-stats.com/benchmarks/fleurs", + "https://api.llm-stats.com/leaderboard/benchmarks/fleurs" + ], + "additional_details": { + "raw_benchmark_id": "fleurs", + "raw_model_id": "gemma-4-12b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Few-shot Learning Evaluation of Universal Representations of Speech - a parallel speech dataset in 102 languages built on FLoRes-101 with approximately 12 hours of speech supervision per language for tasks including ASR, speech language identification, translation and retrieval. Scores are shown as speech recognition accuracy (1 - word error rate), so higher is better.", + "metric_id": "llm_stats.fleurs.score", + "metric_name": "FLEURS score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "fleurs", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FLEURS", + "raw_categories": "[\"language\",\"speech_to_text\"]", + "raw_modality": "audio", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.931, + "details": { + "raw_score": "0.931", + "raw_score_field": "score", + "raw_model_id": "gemma-4-12b-it", + "raw_benchmark_id": "fleurs", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-12b-it\",\"https://llm-stats.com/benchmarks/fleurs\",\"https://api.llm-stats.com/leaderboard/benchmarks/fleurs\"]", + "raw_score_id": "fleurs::gemma-4-12b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mathvision::mathvision-gemma-4-12b-it", + "evaluation_name": "llm_stats.mathvision", + "source_data": { + "dataset_name": "MathVision", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-12b-it", + "https://llm-stats.com/benchmarks/mathvision", + "https://api.llm-stats.com/leaderboard/benchmarks/mathvision" + ], + "additional_details": { + "raw_benchmark_id": "mathvision", + "raw_model_id": "gemma-4-12b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MATH-Vision is a dataset designed to measure multimodal mathematical reasoning capabilities. It focuses on evaluating how well models can solve mathematical problems that require both visual understanding and mathematical reasoning, bridging the gap between visual and mathematical domains.", + "metric_id": "llm_stats.mathvision.score", + "metric_name": "MathVision score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mathvision", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MathVision", + "raw_categories": "[\"math\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "31" + } + }, + "score_details": { + "score": 0.797, + "details": { + "raw_score": "0.797", + "raw_score_field": "score", + "raw_model_id": "gemma-4-12b-it", + "raw_benchmark_id": "mathvision", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-12b-it\",\"https://llm-stats.com/benchmarks/mathvision\",\"https://api.llm-stats.com/leaderboard/benchmarks/mathvision\"]", + "raw_score_id": "mathvision::gemma-4-12b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "medxpertqa::medxpertqa-gemma-4-12b-it", + "evaluation_name": "llm_stats.medxpertqa", + "source_data": { + "dataset_name": "MedXpertQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-12b-it", + "https://llm-stats.com/benchmarks/medxpertqa", + "https://api.llm-stats.com/leaderboard/benchmarks/medxpertqa" + ], + "additional_details": { + "raw_benchmark_id": "medxpertqa", + "raw_model_id": "gemma-4-12b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A comprehensive benchmark to evaluate expert-level medical knowledge and advanced reasoning, featuring 4,460 questions spanning 17 specialties and 11 body systems. Includes both text-only and multimodal subsets with expert-level exam questions incorporating diverse medical images and rich clinical information.", + "metric_id": "llm_stats.medxpertqa.score", + "metric_name": "MedXpertQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "medxpertqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MedXpertQA", + "raw_categories": "[\"multimodal\",\"reasoning\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.487, + "details": { + "raw_score": "0.487", + "raw_score_field": "score", + "raw_model_id": "gemma-4-12b-it", + "raw_benchmark_id": "medxpertqa", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-12b-it\",\"https://llm-stats.com/benchmarks/medxpertqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/medxpertqa\"]", + "raw_score_id": "medxpertqa::gemma-4-12b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mrcr-v2::mrcr-v2-gemma-4-12b-it", + "evaluation_name": "llm_stats.mrcr-v2", + "source_data": { + "dataset_name": "MRCR v2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-12b-it", + "https://llm-stats.com/benchmarks/mrcr-v2", + "https://api.llm-stats.com/leaderboard/benchmarks/mrcr-v2" + ], + "additional_details": { + "raw_benchmark_id": "mrcr-v2", + "raw_model_id": "gemma-4-12b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MRCR v2 (Multi-Round Coreference Resolution version 2) is an enhanced version of the synthetic long-context reasoning task. It extends the original MRCR framework with improved evaluation criteria and additional complexity for testing models' ability to maintain attention and reasoning across extended contexts.", + "metric_id": "llm_stats.mrcr-v2.score", + "metric_name": "MRCR v2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mrcr-v2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MRCR v2", + "raw_categories": "[\"long_context\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.434, + "details": { + "raw_score": "0.434", + "raw_score_field": "score", + "raw_model_id": "gemma-4-12b-it", + "raw_benchmark_id": "mrcr-v2", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-12b-it\",\"https://llm-stats.com/benchmarks/mrcr-v2\",\"https://api.llm-stats.com/leaderboard/benchmarks/mrcr-v2\"]", + "raw_score_id": "mrcr-v2::gemma-4-12b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "omnidocbench-1.5::omnidocbench-1.5-gemma-4-12b-it", + "evaluation_name": "llm_stats.omnidocbench-1.5", + "source_data": { + "dataset_name": "OmniDocBench 1.5", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-12b-it", + "https://llm-stats.com/benchmarks/omnidocbench-1.5", + "https://api.llm-stats.com/leaderboard/benchmarks/omnidocbench-1.5" + ], + "additional_details": { + "raw_benchmark_id": "omnidocbench-1.5", + "raw_model_id": "gemma-4-12b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OmniDocBench 1.5 is a comprehensive benchmark for evaluating multimodal large language models on document understanding tasks, including OCR, document parsing, information extraction, and visual question answering across diverse document types. Lower Overall Edit Distance scores are better.", + "metric_id": "llm_stats.omnidocbench-1.5.score", + "metric_name": "OmniDocBench 1.5 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "omnidocbench-1.5", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OmniDocBench 1.5", + "raw_categories": "[\"multimodal\",\"reasoning\",\"structured_output\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.164, + "details": { + "raw_score": "0.164", + "raw_score_field": "score", + "raw_model_id": "gemma-4-12b-it", + "raw_benchmark_id": "omnidocbench-1.5", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-12b-it\",\"https://llm-stats.com/benchmarks/omnidocbench-1.5\",\"https://api.llm-stats.com/leaderboard/benchmarks/omnidocbench-1.5\"]", + "raw_score_id": "omnidocbench-1.5::gemma-4-12b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/google/gemma-4-26b-a4b-it/912f49ed-cee4-4188-a4ea-7647e058d061.json b/data/llm_stats/google/gemma-4-26b-a4b-it/912f49ed-cee4-4188-a4ea-7647e058d061.json new file mode 100644 index 0000000000000000000000000000000000000000..d1acbd102b2331e56accd0ae19838651b8715d1e --- /dev/null +++ b/data/llm_stats/google/gemma-4-26b-a4b-it/912f49ed-cee4-4188-a4ea-7647e058d061.json @@ -0,0 +1,360 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/google_gemma-4-26b-a4b-it/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "gemma-4-26b-a4b-it", + "id": "google/gemma-4-26b-a4b-it", + "developer": "google", + "additional_details": { + "raw_id": "gemma-4-26b-a4b-it", + "raw_model_name": "gemma-4-26b-a4b-it", + "raw_display_name": "Gemma 4 26B-A4B", + "raw_organization_id": "google", + "raw_organization_name": "Google", + "raw_context_length": "262144", + "raw_input_modalities": "[\"image\",\"text\"]", + "raw_output_modalities": "[\"text\"]", + "raw_input_price": "0.13", + "raw_output_price": "0.4", + "raw_model_id": "gemma-4-26b-a4b-it", + "raw_provider_slug": "google", + "raw_provider_name": "ZeroEval" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aime-2026::aime-2026-gemma-4-26b-a4b-it", + "evaluation_name": "llm_stats.aime-2026", + "source_data": { + "dataset_name": "AIME 2026", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-26b-a4b-it", + "https://llm-stats.com/benchmarks/aime-2026", + "https://api.llm-stats.com/leaderboard/benchmarks/aime-2026" + ], + "additional_details": { + "raw_benchmark_id": "aime-2026", + "raw_model_id": "gemma-4-26b-a4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "All 30 problems from the 2026 American Invitational Mathematics Examination (AIME I and AIME II), testing olympiad-level mathematical reasoning with integer answers from 000-999. Used as an AI benchmark to evaluate large language models' ability to solve complex mathematical problems requiring multi-step logical deductions and structured symbolic reasoning.", + "metric_id": "llm_stats.aime-2026.score", + "metric_name": "AIME 2026 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aime-2026", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AIME 2026", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "17" + } + }, + "score_details": { + "score": 0.883, + "details": { + "raw_score": "0.883", + "raw_score_field": "score", + "raw_model_id": "gemma-4-26b-a4b-it", + "raw_benchmark_id": "aime-2026", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-26b-a4b-it\",\"https://llm-stats.com/benchmarks/aime-2026\",\"https://api.llm-stats.com/leaderboard/benchmarks/aime-2026\"]", + "raw_score_id": "aime-2026::gemma-4-26b-a4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "big-bench-extra-hard::big-bench-extra-hard-gemma-4-26b-a4b-it", + "evaluation_name": "llm_stats.big-bench-extra-hard", + "source_data": { + "dataset_name": "BIG-Bench Extra Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-26b-a4b-it", + "https://llm-stats.com/benchmarks/big-bench-extra-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench-extra-hard" + ], + "additional_details": { + "raw_benchmark_id": "big-bench-extra-hard", + "raw_model_id": "gemma-4-26b-a4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIG-Bench Extra Hard (BBEH) is a challenging benchmark that replaces each task in BIG-Bench Hard with a novel task that probes similar reasoning capabilities but exhibits significantly increased difficulty. The benchmark contains 23 tasks testing diverse reasoning skills including many-hop reasoning, causal understanding, spatial reasoning, temporal arithmetic, geometric reasoning, linguistic reasoning, logic puzzles, and humor understanding. Designed to address saturation on existing benchmarks where state-of-the-art models achieve near-perfect scores, BBEH shows substantial room for improvement with best models achieving only 9.8-44.8% average accuracy.", + "metric_id": "llm_stats.big-bench-extra-hard.score", + "metric_name": "BIG-Bench Extra Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench-extra-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench Extra Hard", + "raw_categories": "[\"language\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.648, + "details": { + "raw_score": "0.648", + "raw_score_field": "score", + "raw_model_id": "gemma-4-26b-a4b-it", + "raw_benchmark_id": "big-bench-extra-hard", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-26b-a4b-it\",\"https://llm-stats.com/benchmarks/big-bench-extra-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench-extra-hard\"]", + "raw_score_id": "big-bench-extra-hard::gemma-4-26b-a4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mathvision::mathvision-gemma-4-26b-a4b-it", + "evaluation_name": "llm_stats.mathvision", + "source_data": { + "dataset_name": "MathVision", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-26b-a4b-it", + "https://llm-stats.com/benchmarks/mathvision", + "https://api.llm-stats.com/leaderboard/benchmarks/mathvision" + ], + "additional_details": { + "raw_benchmark_id": "mathvision", + "raw_model_id": "gemma-4-26b-a4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MATH-Vision is a dataset designed to measure multimodal mathematical reasoning capabilities. It focuses on evaluating how well models can solve mathematical problems that require both visual understanding and mathematical reasoning, bridging the gap between visual and mathematical domains.", + "metric_id": "llm_stats.mathvision.score", + "metric_name": "MathVision score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mathvision", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MathVision", + "raw_categories": "[\"math\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "31" + } + }, + "score_details": { + "score": 0.824, + "details": { + "raw_score": "0.824", + "raw_score_field": "score", + "raw_model_id": "gemma-4-26b-a4b-it", + "raw_benchmark_id": "mathvision", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-26b-a4b-it\",\"https://llm-stats.com/benchmarks/mathvision\",\"https://api.llm-stats.com/leaderboard/benchmarks/mathvision\"]", + "raw_score_id": "mathvision::gemma-4-26b-a4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "medxpertqa::medxpertqa-gemma-4-26b-a4b-it", + "evaluation_name": "llm_stats.medxpertqa", + "source_data": { + "dataset_name": "MedXpertQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-26b-a4b-it", + "https://llm-stats.com/benchmarks/medxpertqa", + "https://api.llm-stats.com/leaderboard/benchmarks/medxpertqa" + ], + "additional_details": { + "raw_benchmark_id": "medxpertqa", + "raw_model_id": "gemma-4-26b-a4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A comprehensive benchmark to evaluate expert-level medical knowledge and advanced reasoning, featuring 4,460 questions spanning 17 specialties and 11 body systems. Includes both text-only and multimodal subsets with expert-level exam questions incorporating diverse medical images and rich clinical information.", + "metric_id": "llm_stats.medxpertqa.score", + "metric_name": "MedXpertQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "medxpertqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MedXpertQA", + "raw_categories": "[\"multimodal\",\"reasoning\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.581, + "details": { + "raw_score": "0.581", + "raw_score_field": "score", + "raw_model_id": "gemma-4-26b-a4b-it", + "raw_benchmark_id": "medxpertqa", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-26b-a4b-it\",\"https://llm-stats.com/benchmarks/medxpertqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/medxpertqa\"]", + "raw_score_id": "medxpertqa::gemma-4-26b-a4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mrcr-v2::mrcr-v2-gemma-4-26b-a4b-it", + "evaluation_name": "llm_stats.mrcr-v2", + "source_data": { + "dataset_name": "MRCR v2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-26b-a4b-it", + "https://llm-stats.com/benchmarks/mrcr-v2", + "https://api.llm-stats.com/leaderboard/benchmarks/mrcr-v2" + ], + "additional_details": { + "raw_benchmark_id": "mrcr-v2", + "raw_model_id": "gemma-4-26b-a4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MRCR v2 (Multi-Round Coreference Resolution version 2) is an enhanced version of the synthetic long-context reasoning task. It extends the original MRCR framework with improved evaluation criteria and additional complexity for testing models' ability to maintain attention and reasoning across extended contexts.", + "metric_id": "llm_stats.mrcr-v2.score", + "metric_name": "MRCR v2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mrcr-v2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MRCR v2", + "raw_categories": "[\"long_context\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.441, + "details": { + "raw_score": "0.441", + "raw_score_field": "score", + "raw_model_id": "gemma-4-26b-a4b-it", + "raw_benchmark_id": "mrcr-v2", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-26b-a4b-it\",\"https://llm-stats.com/benchmarks/mrcr-v2\",\"https://api.llm-stats.com/leaderboard/benchmarks/mrcr-v2\"]", + "raw_score_id": "mrcr-v2::gemma-4-26b-a4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "t2-bench::t2-bench-gemma-4-26b-a4b-it", + "evaluation_name": "llm_stats.t2-bench", + "source_data": { + "dataset_name": "t2-bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-26b-a4b-it", + "https://llm-stats.com/benchmarks/t2-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/t2-bench" + ], + "additional_details": { + "raw_benchmark_id": "t2-bench", + "raw_model_id": "gemma-4-26b-a4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "t2-bench is a benchmark for evaluating agentic tool use capabilities, measuring how well models can select, sequence, and utilize tools to solve complex tasks. It tests autonomous planning and execution in multi-step scenarios.", + "metric_id": "llm_stats.t2-bench.score", + "metric_name": "t2-bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "t2-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "t2-bench", + "raw_categories": "[\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.855, + "details": { + "raw_score": "0.855", + "raw_score_field": "score", + "raw_model_id": "gemma-4-26b-a4b-it", + "raw_benchmark_id": "t2-bench", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-26b-a4b-it\",\"https://llm-stats.com/benchmarks/t2-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/t2-bench\"]", + "raw_score_id": "t2-bench::gemma-4-26b-a4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/google/gemma-4-31b-it/36a50e9b-825c-4ba5-91c7-5aa2f0837e23.json b/data/llm_stats/google/gemma-4-31b-it/36a50e9b-825c-4ba5-91c7-5aa2f0837e23.json new file mode 100644 index 0000000000000000000000000000000000000000..319454fa528043d54b43feb323f80af1bdd26281 --- /dev/null +++ b/data/llm_stats/google/gemma-4-31b-it/36a50e9b-825c-4ba5-91c7-5aa2f0837e23.json @@ -0,0 +1,516 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/google_gemma-4-31b-it/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "gemma-4-31b-it", + "id": "google/gemma-4-31b-it", + "developer": "google", + "additional_details": { + "raw_id": "gemma-4-31b-it", + "raw_model_name": "gemma-4-31b-it", + "raw_display_name": "Gemma 4 31B", + "raw_organization_id": "google", + "raw_organization_name": "Google", + "raw_context_length": "262144", + "raw_input_modalities": "[\"image\",\"text\"]", + "raw_output_modalities": "[\"text\"]", + "raw_input_price": "0.13", + "raw_output_price": "0.38", + "raw_model_id": "gemma-4-31b-it", + "raw_provider_slug": "google", + "raw_provider_name": "ZeroEval" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aime-2026::aime-2026-gemma-4-31b-it", + "evaluation_name": "llm_stats.aime-2026", + "source_data": { + "dataset_name": "AIME 2026", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-31b-it", + "https://llm-stats.com/benchmarks/aime-2026", + "https://api.llm-stats.com/leaderboard/benchmarks/aime-2026" + ], + "additional_details": { + "raw_benchmark_id": "aime-2026", + "raw_model_id": "gemma-4-31b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "All 30 problems from the 2026 American Invitational Mathematics Examination (AIME I and AIME II), testing olympiad-level mathematical reasoning with integer answers from 000-999. Used as an AI benchmark to evaluate large language models' ability to solve complex mathematical problems requiring multi-step logical deductions and structured symbolic reasoning.", + "metric_id": "llm_stats.aime-2026.score", + "metric_name": "AIME 2026 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aime-2026", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AIME 2026", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "17" + } + }, + "score_details": { + "score": 0.892, + "details": { + "raw_score": "0.892", + "raw_score_field": "score", + "raw_model_id": "gemma-4-31b-it", + "raw_benchmark_id": "aime-2026", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-31b-it\",\"https://llm-stats.com/benchmarks/aime-2026\",\"https://api.llm-stats.com/leaderboard/benchmarks/aime-2026\"]", + "raw_score_id": "aime-2026::gemma-4-31b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "big-bench-extra-hard::big-bench-extra-hard-gemma-4-31b-it", + "evaluation_name": "llm_stats.big-bench-extra-hard", + "source_data": { + "dataset_name": "BIG-Bench Extra Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-31b-it", + "https://llm-stats.com/benchmarks/big-bench-extra-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench-extra-hard" + ], + "additional_details": { + "raw_benchmark_id": "big-bench-extra-hard", + "raw_model_id": "gemma-4-31b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIG-Bench Extra Hard (BBEH) is a challenging benchmark that replaces each task in BIG-Bench Hard with a novel task that probes similar reasoning capabilities but exhibits significantly increased difficulty. The benchmark contains 23 tasks testing diverse reasoning skills including many-hop reasoning, causal understanding, spatial reasoning, temporal arithmetic, geometric reasoning, linguistic reasoning, logic puzzles, and humor understanding. Designed to address saturation on existing benchmarks where state-of-the-art models achieve near-perfect scores, BBEH shows substantial room for improvement with best models achieving only 9.8-44.8% average accuracy.", + "metric_id": "llm_stats.big-bench-extra-hard.score", + "metric_name": "BIG-Bench Extra Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench-extra-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench Extra Hard", + "raw_categories": "[\"language\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.744, + "details": { + "raw_score": "0.744", + "raw_score_field": "score", + "raw_model_id": "gemma-4-31b-it", + "raw_benchmark_id": "big-bench-extra-hard", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-31b-it\",\"https://llm-stats.com/benchmarks/big-bench-extra-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench-extra-hard\"]", + "raw_score_id": "big-bench-extra-hard::gemma-4-31b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mathvision::mathvision-gemma-4-31b-it", + "evaluation_name": "llm_stats.mathvision", + "source_data": { + "dataset_name": "MathVision", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-31b-it", + "https://llm-stats.com/benchmarks/mathvision", + "https://api.llm-stats.com/leaderboard/benchmarks/mathvision" + ], + "additional_details": { + "raw_benchmark_id": "mathvision", + "raw_model_id": "gemma-4-31b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MATH-Vision is a dataset designed to measure multimodal mathematical reasoning capabilities. It focuses on evaluating how well models can solve mathematical problems that require both visual understanding and mathematical reasoning, bridging the gap between visual and mathematical domains.", + "metric_id": "llm_stats.mathvision.score", + "metric_name": "MathVision score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mathvision", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MathVision", + "raw_categories": "[\"math\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "31" + } + }, + "score_details": { + "score": 0.856, + "details": { + "raw_score": "0.856", + "raw_score_field": "score", + "raw_model_id": "gemma-4-31b-it", + "raw_benchmark_id": "mathvision", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-31b-it\",\"https://llm-stats.com/benchmarks/mathvision\",\"https://api.llm-stats.com/leaderboard/benchmarks/mathvision\"]", + "raw_score_id": "mathvision::gemma-4-31b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "medxpertqa::medxpertqa-gemma-4-31b-it", + "evaluation_name": "llm_stats.medxpertqa", + "source_data": { + "dataset_name": "MedXpertQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-31b-it", + "https://llm-stats.com/benchmarks/medxpertqa", + "https://api.llm-stats.com/leaderboard/benchmarks/medxpertqa" + ], + "additional_details": { + "raw_benchmark_id": "medxpertqa", + "raw_model_id": "gemma-4-31b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A comprehensive benchmark to evaluate expert-level medical knowledge and advanced reasoning, featuring 4,460 questions spanning 17 specialties and 11 body systems. Includes both text-only and multimodal subsets with expert-level exam questions incorporating diverse medical images and rich clinical information.", + "metric_id": "llm_stats.medxpertqa.score", + "metric_name": "MedXpertQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "medxpertqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MedXpertQA", + "raw_categories": "[\"multimodal\",\"reasoning\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.613, + "details": { + "raw_score": "0.613", + "raw_score_field": "score", + "raw_model_id": "gemma-4-31b-it", + "raw_benchmark_id": "medxpertqa", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-31b-it\",\"https://llm-stats.com/benchmarks/medxpertqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/medxpertqa\"]", + "raw_score_id": "medxpertqa::gemma-4-31b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu-pro::mmlu-pro-gemma-4-31b-it", + "evaluation_name": "llm_stats.mmlu-pro", + "source_data": { + "dataset_name": "MMLU-Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-31b-it", + "https://llm-stats.com/benchmarks/mmlu-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu-pro" + ], + "additional_details": { + "raw_benchmark_id": "mmlu-pro", + "raw_model_id": "gemma-4-31b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A more robust and challenging multi-task language understanding benchmark that extends MMLU by expanding multiple-choice options from 4 to 10, eliminating trivial questions, and focusing on reasoning-intensive tasks. Features over 12,000 curated questions across 14 domains and causes a 16-33% accuracy drop compared to original MMLU.", + "metric_id": "llm_stats.mmlu-pro.score", + "metric_name": "MMLU-Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU-Pro", + "raw_categories": "[\"language\",\"legal\",\"math\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "129" + } + }, + "score_details": { + "score": 0.852, + "details": { + "raw_score": "0.852", + "raw_score_field": "score", + "raw_model_id": "gemma-4-31b-it", + "raw_benchmark_id": "mmlu-pro", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-31b-it\",\"https://llm-stats.com/benchmarks/mmlu-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu-pro\"]", + "raw_score_id": "mmlu-pro::gemma-4-31b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmmlu::mmmlu-gemma-4-31b-it", + "evaluation_name": "llm_stats.mmmlu", + "source_data": { + "dataset_name": "MMMLU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-31b-it", + "https://llm-stats.com/benchmarks/mmmlu", + "https://api.llm-stats.com/leaderboard/benchmarks/mmmlu" + ], + "additional_details": { + "raw_benchmark_id": "mmmlu", + "raw_model_id": "gemma-4-31b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Multilingual Massive Multitask Language Understanding dataset released by OpenAI, featuring professionally translated MMLU test questions across 14 languages including Arabic, Bengali, German, Spanish, French, Hindi, Indonesian, Italian, Japanese, Korean, Portuguese, Swahili, Yoruba, and Chinese. Contains approximately 15,908 multiple-choice questions per language covering 57 subjects.", + "metric_id": "llm_stats.mmmlu.score", + "metric_name": "MMMLU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmmlu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMMLU", + "raw_categories": "[\"language\",\"math\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "49" + } + }, + "score_details": { + "score": 0.884, + "details": { + "raw_score": "0.884", + "raw_score_field": "score", + "raw_model_id": "gemma-4-31b-it", + "raw_benchmark_id": "mmmlu", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-31b-it\",\"https://llm-stats.com/benchmarks/mmmlu\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmmlu\"]", + "raw_score_id": "mmmlu::gemma-4-31b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmmu-pro::mmmu-pro-gemma-4-31b-it", + "evaluation_name": "llm_stats.mmmu-pro", + "source_data": { + "dataset_name": "MMMU-Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-31b-it", + "https://llm-stats.com/benchmarks/mmmu-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/mmmu-pro" + ], + "additional_details": { + "raw_benchmark_id": "mmmu-pro", + "raw_model_id": "gemma-4-31b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A more robust multi-discipline multimodal understanding benchmark that enhances MMMU through a three-step process: filtering text-only answerable questions, augmenting candidate options, and introducing vision-only input settings. Achieves significantly lower model performance (16.8-26.9%) compared to original MMMU, providing more rigorous evaluation that closely mimics real-world scenarios.", + "metric_id": "llm_stats.mmmu-pro.score", + "metric_name": "MMMU-Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmmu-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMMU-Pro", + "raw_categories": "[\"multimodal\",\"reasoning\",\"general\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "60" + } + }, + "score_details": { + "score": 0.769, + "details": { + "raw_score": "0.769", + "raw_score_field": "score", + "raw_model_id": "gemma-4-31b-it", + "raw_benchmark_id": "mmmu-pro", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-31b-it\",\"https://llm-stats.com/benchmarks/mmmu-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmmu-pro\"]", + "raw_score_id": "mmmu-pro::gemma-4-31b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mrcr-v2::mrcr-v2-gemma-4-31b-it", + "evaluation_name": "llm_stats.mrcr-v2", + "source_data": { + "dataset_name": "MRCR v2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-31b-it", + "https://llm-stats.com/benchmarks/mrcr-v2", + "https://api.llm-stats.com/leaderboard/benchmarks/mrcr-v2" + ], + "additional_details": { + "raw_benchmark_id": "mrcr-v2", + "raw_model_id": "gemma-4-31b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MRCR v2 (Multi-Round Coreference Resolution version 2) is an enhanced version of the synthetic long-context reasoning task. It extends the original MRCR framework with improved evaluation criteria and additional complexity for testing models' ability to maintain attention and reasoning across extended contexts.", + "metric_id": "llm_stats.mrcr-v2.score", + "metric_name": "MRCR v2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mrcr-v2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MRCR v2", + "raw_categories": "[\"long_context\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.664, + "details": { + "raw_score": "0.664", + "raw_score_field": "score", + "raw_model_id": "gemma-4-31b-it", + "raw_benchmark_id": "mrcr-v2", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-31b-it\",\"https://llm-stats.com/benchmarks/mrcr-v2\",\"https://api.llm-stats.com/leaderboard/benchmarks/mrcr-v2\"]", + "raw_score_id": "mrcr-v2::gemma-4-31b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "t2-bench::t2-bench-gemma-4-31b-it", + "evaluation_name": "llm_stats.t2-bench", + "source_data": { + "dataset_name": "t2-bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-31b-it", + "https://llm-stats.com/benchmarks/t2-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/t2-bench" + ], + "additional_details": { + "raw_benchmark_id": "t2-bench", + "raw_model_id": "gemma-4-31b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "t2-bench is a benchmark for evaluating agentic tool use capabilities, measuring how well models can select, sequence, and utilize tools to solve complex tasks. It tests autonomous planning and execution in multi-step scenarios.", + "metric_id": "llm_stats.t2-bench.score", + "metric_name": "t2-bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "t2-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "t2-bench", + "raw_categories": "[\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.864, + "details": { + "raw_score": "0.864", + "raw_score_field": "score", + "raw_model_id": "gemma-4-31b-it", + "raw_benchmark_id": "t2-bench", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-31b-it\",\"https://llm-stats.com/benchmarks/t2-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/t2-bench\"]", + "raw_score_id": "t2-bench::gemma-4-31b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/google/gemma-4-e2b-it/07806b8f-abf3-444f-bd5e-dd5a75639a56.json b/data/llm_stats/google/gemma-4-e2b-it/07806b8f-abf3-444f-bd5e-dd5a75639a56.json new file mode 100644 index 0000000000000000000000000000000000000000..6509e2e6b90a174d800e6847cdd012eefafd816a --- /dev/null +++ b/data/llm_stats/google/gemma-4-e2b-it/07806b8f-abf3-444f-bd5e-dd5a75639a56.json @@ -0,0 +1,255 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/google_gemma-4-e2b-it/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Gemma 4 E2B", + "id": "google/gemma-4-e2b-it", + "developer": "google", + "additional_details": { + "raw_id": "gemma-4-e2b-it", + "raw_name": "Gemma 4 E2B", + "raw_model_id": "gemma-4-e2b-it", + "raw_model_name": "Gemma 4 E2B", + "raw_organization_id": "google", + "raw_organization_name": "Google", + "raw_release_date": "2026-04-02", + "raw_announcement_date": "2026-04-02", + "raw_multimodal": "true", + "raw_param_count": "5100000000", + "raw_provider_slug": "google", + "raw_provider_name": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aime-2026::aime-2026-gemma-4-e2b-it", + "evaluation_name": "llm_stats.aime-2026", + "source_data": { + "dataset_name": "AIME 2026", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-e2b-it", + "https://llm-stats.com/benchmarks/aime-2026", + "https://api.llm-stats.com/leaderboard/benchmarks/aime-2026" + ], + "additional_details": { + "raw_benchmark_id": "aime-2026", + "raw_model_id": "gemma-4-e2b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "All 30 problems from the 2026 American Invitational Mathematics Examination (AIME I and AIME II), testing olympiad-level mathematical reasoning with integer answers from 000-999. Used as an AI benchmark to evaluate large language models' ability to solve complex mathematical problems requiring multi-step logical deductions and structured symbolic reasoning.", + "metric_id": "llm_stats.aime-2026.score", + "metric_name": "AIME 2026 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aime-2026", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AIME 2026", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "17" + } + }, + "score_details": { + "score": 0.375, + "details": { + "raw_score": "0.375", + "raw_score_field": "score", + "raw_model_id": "gemma-4-e2b-it", + "raw_benchmark_id": "aime-2026", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-e2b-it\",\"https://llm-stats.com/benchmarks/aime-2026\",\"https://api.llm-stats.com/leaderboard/benchmarks/aime-2026\"]", + "raw_score_id": "aime-2026::gemma-4-e2b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "big-bench-extra-hard::big-bench-extra-hard-gemma-4-e2b-it", + "evaluation_name": "llm_stats.big-bench-extra-hard", + "source_data": { + "dataset_name": "BIG-Bench Extra Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-e2b-it", + "https://llm-stats.com/benchmarks/big-bench-extra-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench-extra-hard" + ], + "additional_details": { + "raw_benchmark_id": "big-bench-extra-hard", + "raw_model_id": "gemma-4-e2b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIG-Bench Extra Hard (BBEH) is a challenging benchmark that replaces each task in BIG-Bench Hard with a novel task that probes similar reasoning capabilities but exhibits significantly increased difficulty. The benchmark contains 23 tasks testing diverse reasoning skills including many-hop reasoning, causal understanding, spatial reasoning, temporal arithmetic, geometric reasoning, linguistic reasoning, logic puzzles, and humor understanding. Designed to address saturation on existing benchmarks where state-of-the-art models achieve near-perfect scores, BBEH shows substantial room for improvement with best models achieving only 9.8-44.8% average accuracy.", + "metric_id": "llm_stats.big-bench-extra-hard.score", + "metric_name": "BIG-Bench Extra Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench-extra-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench Extra Hard", + "raw_categories": "[\"language\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.219, + "details": { + "raw_score": "0.219", + "raw_score_field": "score", + "raw_model_id": "gemma-4-e2b-it", + "raw_benchmark_id": "big-bench-extra-hard", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-e2b-it\",\"https://llm-stats.com/benchmarks/big-bench-extra-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench-extra-hard\"]", + "raw_score_id": "big-bench-extra-hard::gemma-4-e2b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "medxpertqa::medxpertqa-gemma-4-e2b-it", + "evaluation_name": "llm_stats.medxpertqa", + "source_data": { + "dataset_name": "MedXpertQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-e2b-it", + "https://llm-stats.com/benchmarks/medxpertqa", + "https://api.llm-stats.com/leaderboard/benchmarks/medxpertqa" + ], + "additional_details": { + "raw_benchmark_id": "medxpertqa", + "raw_model_id": "gemma-4-e2b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A comprehensive benchmark to evaluate expert-level medical knowledge and advanced reasoning, featuring 4,460 questions spanning 17 specialties and 11 body systems. Includes both text-only and multimodal subsets with expert-level exam questions incorporating diverse medical images and rich clinical information.", + "metric_id": "llm_stats.medxpertqa.score", + "metric_name": "MedXpertQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "medxpertqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MedXpertQA", + "raw_categories": "[\"multimodal\",\"reasoning\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.235, + "details": { + "raw_score": "0.235", + "raw_score_field": "score", + "raw_model_id": "gemma-4-e2b-it", + "raw_benchmark_id": "medxpertqa", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-e2b-it\",\"https://llm-stats.com/benchmarks/medxpertqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/medxpertqa\"]", + "raw_score_id": "medxpertqa::gemma-4-e2b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mrcr-v2::mrcr-v2-gemma-4-e2b-it", + "evaluation_name": "llm_stats.mrcr-v2", + "source_data": { + "dataset_name": "MRCR v2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-e2b-it", + "https://llm-stats.com/benchmarks/mrcr-v2", + "https://api.llm-stats.com/leaderboard/benchmarks/mrcr-v2" + ], + "additional_details": { + "raw_benchmark_id": "mrcr-v2", + "raw_model_id": "gemma-4-e2b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MRCR v2 (Multi-Round Coreference Resolution version 2) is an enhanced version of the synthetic long-context reasoning task. It extends the original MRCR framework with improved evaluation criteria and additional complexity for testing models' ability to maintain attention and reasoning across extended contexts.", + "metric_id": "llm_stats.mrcr-v2.score", + "metric_name": "MRCR v2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mrcr-v2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MRCR v2", + "raw_categories": "[\"long_context\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.191, + "details": { + "raw_score": "0.191", + "raw_score_field": "score", + "raw_model_id": "gemma-4-e2b-it", + "raw_benchmark_id": "mrcr-v2", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-e2b-it\",\"https://llm-stats.com/benchmarks/mrcr-v2\",\"https://api.llm-stats.com/leaderboard/benchmarks/mrcr-v2\"]", + "raw_score_id": "mrcr-v2::gemma-4-e2b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/google/gemma-4-e4b-it/272d9080-4dea-400e-b14e-b43059f588f9.json b/data/llm_stats/google/gemma-4-e4b-it/272d9080-4dea-400e-b14e-b43059f588f9.json new file mode 100644 index 0000000000000000000000000000000000000000..6ed07e6c2d6231af245a85b36526e7ee7d4042b0 --- /dev/null +++ b/data/llm_stats/google/gemma-4-e4b-it/272d9080-4dea-400e-b14e-b43059f588f9.json @@ -0,0 +1,307 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/google_gemma-4-e4b-it/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Gemma 4 E4B", + "id": "google/gemma-4-e4b-it", + "developer": "google", + "additional_details": { + "raw_id": "gemma-4-e4b-it", + "raw_name": "Gemma 4 E4B", + "raw_model_id": "gemma-4-e4b-it", + "raw_model_name": "Gemma 4 E4B", + "raw_organization_id": "google", + "raw_organization_name": "Google", + "raw_release_date": "2026-04-02", + "raw_announcement_date": "2026-04-02", + "raw_multimodal": "true", + "raw_param_count": "8000000000", + "raw_provider_slug": "google", + "raw_provider_name": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aime-2026::aime-2026-gemma-4-e4b-it", + "evaluation_name": "llm_stats.aime-2026", + "source_data": { + "dataset_name": "AIME 2026", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-e4b-it", + "https://llm-stats.com/benchmarks/aime-2026", + "https://api.llm-stats.com/leaderboard/benchmarks/aime-2026" + ], + "additional_details": { + "raw_benchmark_id": "aime-2026", + "raw_model_id": "gemma-4-e4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "All 30 problems from the 2026 American Invitational Mathematics Examination (AIME I and AIME II), testing olympiad-level mathematical reasoning with integer answers from 000-999. Used as an AI benchmark to evaluate large language models' ability to solve complex mathematical problems requiring multi-step logical deductions and structured symbolic reasoning.", + "metric_id": "llm_stats.aime-2026.score", + "metric_name": "AIME 2026 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aime-2026", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AIME 2026", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "17" + } + }, + "score_details": { + "score": 0.425, + "details": { + "raw_score": "0.425", + "raw_score_field": "score", + "raw_model_id": "gemma-4-e4b-it", + "raw_benchmark_id": "aime-2026", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-e4b-it\",\"https://llm-stats.com/benchmarks/aime-2026\",\"https://api.llm-stats.com/leaderboard/benchmarks/aime-2026\"]", + "raw_score_id": "aime-2026::gemma-4-e4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "big-bench-extra-hard::big-bench-extra-hard-gemma-4-e4b-it", + "evaluation_name": "llm_stats.big-bench-extra-hard", + "source_data": { + "dataset_name": "BIG-Bench Extra Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-e4b-it", + "https://llm-stats.com/benchmarks/big-bench-extra-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench-extra-hard" + ], + "additional_details": { + "raw_benchmark_id": "big-bench-extra-hard", + "raw_model_id": "gemma-4-e4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIG-Bench Extra Hard (BBEH) is a challenging benchmark that replaces each task in BIG-Bench Hard with a novel task that probes similar reasoning capabilities but exhibits significantly increased difficulty. The benchmark contains 23 tasks testing diverse reasoning skills including many-hop reasoning, causal understanding, spatial reasoning, temporal arithmetic, geometric reasoning, linguistic reasoning, logic puzzles, and humor understanding. Designed to address saturation on existing benchmarks where state-of-the-art models achieve near-perfect scores, BBEH shows substantial room for improvement with best models achieving only 9.8-44.8% average accuracy.", + "metric_id": "llm_stats.big-bench-extra-hard.score", + "metric_name": "BIG-Bench Extra Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench-extra-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench Extra Hard", + "raw_categories": "[\"language\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.331, + "details": { + "raw_score": "0.331", + "raw_score_field": "score", + "raw_model_id": "gemma-4-e4b-it", + "raw_benchmark_id": "big-bench-extra-hard", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-e4b-it\",\"https://llm-stats.com/benchmarks/big-bench-extra-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench-extra-hard\"]", + "raw_score_id": "big-bench-extra-hard::gemma-4-e4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "medxpertqa::medxpertqa-gemma-4-e4b-it", + "evaluation_name": "llm_stats.medxpertqa", + "source_data": { + "dataset_name": "MedXpertQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-e4b-it", + "https://llm-stats.com/benchmarks/medxpertqa", + "https://api.llm-stats.com/leaderboard/benchmarks/medxpertqa" + ], + "additional_details": { + "raw_benchmark_id": "medxpertqa", + "raw_model_id": "gemma-4-e4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A comprehensive benchmark to evaluate expert-level medical knowledge and advanced reasoning, featuring 4,460 questions spanning 17 specialties and 11 body systems. Includes both text-only and multimodal subsets with expert-level exam questions incorporating diverse medical images and rich clinical information.", + "metric_id": "llm_stats.medxpertqa.score", + "metric_name": "MedXpertQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "medxpertqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MedXpertQA", + "raw_categories": "[\"multimodal\",\"reasoning\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.287, + "details": { + "raw_score": "0.287", + "raw_score_field": "score", + "raw_model_id": "gemma-4-e4b-it", + "raw_benchmark_id": "medxpertqa", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-e4b-it\",\"https://llm-stats.com/benchmarks/medxpertqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/medxpertqa\"]", + "raw_score_id": "medxpertqa::gemma-4-e4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mrcr-v2::mrcr-v2-gemma-4-e4b-it", + "evaluation_name": "llm_stats.mrcr-v2", + "source_data": { + "dataset_name": "MRCR v2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-e4b-it", + "https://llm-stats.com/benchmarks/mrcr-v2", + "https://api.llm-stats.com/leaderboard/benchmarks/mrcr-v2" + ], + "additional_details": { + "raw_benchmark_id": "mrcr-v2", + "raw_model_id": "gemma-4-e4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MRCR v2 (Multi-Round Coreference Resolution version 2) is an enhanced version of the synthetic long-context reasoning task. It extends the original MRCR framework with improved evaluation criteria and additional complexity for testing models' ability to maintain attention and reasoning across extended contexts.", + "metric_id": "llm_stats.mrcr-v2.score", + "metric_name": "MRCR v2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mrcr-v2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MRCR v2", + "raw_categories": "[\"long_context\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.254, + "details": { + "raw_score": "0.254", + "raw_score_field": "score", + "raw_model_id": "gemma-4-e4b-it", + "raw_benchmark_id": "mrcr-v2", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-e4b-it\",\"https://llm-stats.com/benchmarks/mrcr-v2\",\"https://api.llm-stats.com/leaderboard/benchmarks/mrcr-v2\"]", + "raw_score_id": "mrcr-v2::gemma-4-e4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "t2-bench::t2-bench-gemma-4-e4b-it", + "evaluation_name": "llm_stats.t2-bench", + "source_data": { + "dataset_name": "t2-bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gemma-4-e4b-it", + "https://llm-stats.com/benchmarks/t2-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/t2-bench" + ], + "additional_details": { + "raw_benchmark_id": "t2-bench", + "raw_model_id": "gemma-4-e4b-it", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "t2-bench is a benchmark for evaluating agentic tool use capabilities, measuring how well models can select, sequence, and utilize tools to solve complex tasks. It tests autonomous planning and execution in multi-step scenarios.", + "metric_id": "llm_stats.t2-bench.score", + "metric_name": "t2-bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "t2-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "t2-bench", + "raw_categories": "[\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.575, + "details": { + "raw_score": "0.575", + "raw_score_field": "score", + "raw_model_id": "gemma-4-e4b-it", + "raw_benchmark_id": "t2-bench", + "source_urls_json": "[\"https://llm-stats.com/models/gemma-4-e4b-it\",\"https://llm-stats.com/benchmarks/t2-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/t2-bench\"]", + "raw_score_id": "t2-bench::gemma-4-e4b-it", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/ibm/granite-3.3-8b-base/89742b4e-3e8b-4da0-b25a-e8f6d09a69a6.json b/data/llm_stats/ibm/granite-3.3-8b-base/89742b4e-3e8b-4da0-b25a-e8f6d09a69a6.json new file mode 100644 index 0000000000000000000000000000000000000000..69300afac27aad020068fe77fa4617e5a6f99d48 --- /dev/null +++ b/data/llm_stats/ibm/granite-3.3-8b-base/89742b4e-3e8b-4da0-b25a-e8f6d09a69a6.json @@ -0,0 +1,723 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/ibm_granite-3.3-8b-base/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Granite 3.3 8B Base", + "id": "ibm/granite-3.3-8b-base", + "developer": "ibm", + "additional_details": { + "raw_id": "granite-3.3-8b-base", + "raw_name": "Granite 3.3 8B Base", + "raw_model_id": "granite-3.3-8b-base", + "raw_model_name": "Granite 3.3 8B Base", + "raw_organization_id": "ibm", + "raw_organization_name": "IBM", + "raw_release_date": "2025-04-16", + "raw_announcement_date": "2025-04-16", + "raw_multimodal": "true", + "raw_param_count": "8170000000", + "raw_provider_slug": "ibm", + "raw_provider_name": "IBM" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "agieval::agieval-granite-3.3-8b-base", + "evaluation_name": "llm_stats.agieval", + "source_data": { + "dataset_name": "AGIEval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/granite-3.3-8b-base", + "https://llm-stats.com/benchmarks/agieval", + "https://api.llm-stats.com/leaderboard/benchmarks/agieval" + ], + "additional_details": { + "raw_benchmark_id": "agieval", + "raw_model_id": "granite-3.3-8b-base", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A human-centric benchmark for evaluating foundation models on standardized exams including college entrance exams (Gaokao, SAT), law school admission tests (LSAT), math competitions, lawyer qualification tests, and civil service exams. Contains 20 tasks (18 multiple-choice, 2 cloze) designed to assess understanding, knowledge, reasoning, and calculation abilities in real-world academic and professional contexts.", + "metric_id": "llm_stats.agieval.score", + "metric_name": "AGIEval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "agieval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AGIEval", + "raw_categories": "[\"legal\",\"math\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.493, + "details": { + "raw_score": "0.493", + "raw_score_field": "score", + "raw_model_id": "granite-3.3-8b-base", + "raw_benchmark_id": "agieval", + "source_urls_json": "[\"https://llm-stats.com/models/granite-3.3-8b-base\",\"https://llm-stats.com/benchmarks/agieval\",\"https://api.llm-stats.com/leaderboard/benchmarks/agieval\"]", + "raw_score_id": "agieval::granite-3.3-8b-base", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "alpacaeval-2.0::alpacaeval-2.0-granite-3.3-8b-base", + "evaluation_name": "llm_stats.alpacaeval-2.0", + "source_data": { + "dataset_name": "AlpacaEval 2.0", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/granite-3.3-8b-base", + "https://llm-stats.com/benchmarks/alpacaeval-2.0", + "https://api.llm-stats.com/leaderboard/benchmarks/alpacaeval-2.0" + ], + "additional_details": { + "raw_benchmark_id": "alpacaeval-2.0", + "raw_model_id": "granite-3.3-8b-base", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "AlpacaEval 2.0 is a length-controlled automatic evaluator for instruction-following language models that uses GPT-4 Turbo to assess model responses against a baseline. It evaluates models on 805 diverse instruction-following tasks including creative writing, classification, programming, and general knowledge questions. The benchmark achieves 0.98 Spearman correlation with ChatBot Arena while being fast (< 3 minutes) and affordable (< $10 in OpenAI credits). It addresses length bias in automatic evaluation through length-controlled win-rates and uses weighted scoring based on response quality.", + "metric_id": "llm_stats.alpacaeval-2.0.score", + "metric_name": "AlpacaEval 2.0 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "alpacaeval-2.0", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AlpacaEval 2.0", + "raw_categories": "[\"reasoning\",\"general\",\"creativity\",\"writing\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.6268, + "details": { + "raw_score": "0.6268", + "raw_score_field": "score", + "raw_model_id": "granite-3.3-8b-base", + "raw_benchmark_id": "alpacaeval-2.0", + "source_urls_json": "[\"https://llm-stats.com/models/granite-3.3-8b-base\",\"https://llm-stats.com/benchmarks/alpacaeval-2.0\",\"https://api.llm-stats.com/leaderboard/benchmarks/alpacaeval-2.0\"]", + "raw_score_id": "alpacaeval-2.0::granite-3.3-8b-base", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "arena-hard::arena-hard-granite-3.3-8b-base", + "evaluation_name": "llm_stats.arena-hard", + "source_data": { + "dataset_name": "Arena Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/granite-3.3-8b-base", + "https://llm-stats.com/benchmarks/arena-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/arena-hard" + ], + "additional_details": { + "raw_benchmark_id": "arena-hard", + "raw_model_id": "granite-3.3-8b-base", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Arena-Hard-Auto is an automatic evaluation benchmark for instruction-tuned LLMs consisting of 500 challenging real-world prompts curated by BenchBuilder. It includes open-ended software engineering problems, mathematical questions, and creative writing tasks. The benchmark uses LLM-as-a-Judge methodology with GPT-4.1 and Gemini-2.5 as automatic judges to approximate human preference. Arena-Hard achieves 98.6% correlation with human preference rankings and provides 3x higher separation of model performances compared to MT-Bench, making it highly effective for distinguishing between models of similar quality.", + "metric_id": "llm_stats.arena-hard.score", + "metric_name": "Arena Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arena-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Arena Hard", + "raw_categories": "[\"reasoning\",\"general\",\"creativity\",\"writing\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "26" + } + }, + "score_details": { + "score": 0.5756, + "details": { + "raw_score": "0.5756", + "raw_score_field": "score", + "raw_model_id": "granite-3.3-8b-base", + "raw_benchmark_id": "arena-hard", + "source_urls_json": "[\"https://llm-stats.com/models/granite-3.3-8b-base\",\"https://llm-stats.com/benchmarks/arena-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/arena-hard\"]", + "raw_score_id": "arena-hard::granite-3.3-8b-base", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "attaq::attaq-granite-3.3-8b-base", + "evaluation_name": "llm_stats.attaq", + "source_data": { + "dataset_name": "AttaQ", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/granite-3.3-8b-base", + "https://llm-stats.com/benchmarks/attaq", + "https://api.llm-stats.com/leaderboard/benchmarks/attaq" + ], + "additional_details": { + "raw_benchmark_id": "attaq", + "raw_model_id": "granite-3.3-8b-base", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "AttaQ is a unique dataset containing adversarial examples in the form of questions designed to provoke harmful or inappropriate responses from large language models. The benchmark evaluates safety vulnerabilities by using specialized clustering techniques that analyze both the semantic similarity of input attacks and the harmfulness of model responses, facilitating targeted improvements to model safety mechanisms.", + "metric_id": "llm_stats.attaq.score", + "metric_name": "AttaQ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "attaq", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AttaQ", + "raw_categories": "[\"safety\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.885, + "details": { + "raw_score": "0.885", + "raw_score_field": "score", + "raw_model_id": "granite-3.3-8b-base", + "raw_benchmark_id": "attaq", + "source_urls_json": "[\"https://llm-stats.com/models/granite-3.3-8b-base\",\"https://llm-stats.com/benchmarks/attaq\",\"https://api.llm-stats.com/leaderboard/benchmarks/attaq\"]", + "raw_score_id": "attaq::granite-3.3-8b-base", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "big-bench-hard::big-bench-hard-granite-3.3-8b-base", + "evaluation_name": "llm_stats.big-bench-hard", + "source_data": { + "dataset_name": "BIG-Bench Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/granite-3.3-8b-base", + "https://llm-stats.com/benchmarks/big-bench-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard" + ], + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_model_id": "granite-3.3-8b-base", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIG-Bench Hard (BBH) is a subset of 23 challenging BIG-Bench tasks selected because prior language model evaluations did not outperform average human-rater performance. The benchmark contains 6,511 evaluation examples testing various forms of multi-step reasoning including arithmetic, logical reasoning (Boolean expressions, logical deduction), geometric reasoning, temporal reasoning, and language understanding. Tasks require capabilities such as causal judgment, object counting, navigation, pattern recognition, and complex problem solving.", + "metric_id": "llm_stats.big-bench-hard.score", + "metric_name": "BIG-Bench Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench Hard", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "21" + } + }, + "score_details": { + "score": 0.6913, + "details": { + "raw_score": "0.6913", + "raw_score_field": "score", + "raw_model_id": "granite-3.3-8b-base", + "raw_benchmark_id": "big-bench-hard", + "source_urls_json": "[\"https://llm-stats.com/models/granite-3.3-8b-base\",\"https://llm-stats.com/benchmarks/big-bench-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard\"]", + "raw_score_id": "big-bench-hard::granite-3.3-8b-base", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hellaswag::hellaswag-granite-3.3-8b-base", + "evaluation_name": "llm_stats.hellaswag", + "source_data": { + "dataset_name": "HellaSwag", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/granite-3.3-8b-base", + "https://llm-stats.com/benchmarks/hellaswag", + "https://api.llm-stats.com/leaderboard/benchmarks/hellaswag" + ], + "additional_details": { + "raw_benchmark_id": "hellaswag", + "raw_model_id": "granite-3.3-8b-base", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A challenging commonsense natural language inference dataset that uses Adversarial Filtering to create questions trivial for humans (>95% accuracy) but difficult for state-of-the-art models, requiring completion of sentence endings based on physical situations and everyday activities", + "metric_id": "llm_stats.hellaswag.score", + "metric_name": "HellaSwag score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hellaswag", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HellaSwag", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.801, + "details": { + "raw_score": "0.801", + "raw_score_field": "score", + "raw_model_id": "granite-3.3-8b-base", + "raw_benchmark_id": "hellaswag", + "source_urls_json": "[\"https://llm-stats.com/models/granite-3.3-8b-base\",\"https://llm-stats.com/benchmarks/hellaswag\",\"https://api.llm-stats.com/leaderboard/benchmarks/hellaswag\"]", + "raw_score_id": "hellaswag::granite-3.3-8b-base", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "humaneval::humaneval-granite-3.3-8b-base", + "evaluation_name": "llm_stats.humaneval", + "source_data": { + "dataset_name": "HumanEval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/granite-3.3-8b-base", + "https://llm-stats.com/benchmarks/humaneval", + "https://api.llm-stats.com/leaderboard/benchmarks/humaneval" + ], + "additional_details": { + "raw_benchmark_id": "humaneval", + "raw_model_id": "granite-3.3-8b-base", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A benchmark that measures functional correctness for synthesizing programs from docstrings, consisting of 164 original programming problems assessing language comprehension, algorithms, and simple mathematics", + "metric_id": "llm_stats.humaneval.score", + "metric_name": "HumanEval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "humaneval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HumanEval", + "raw_categories": "[\"reasoning\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "66" + } + }, + "score_details": { + "score": 0.8973, + "details": { + "raw_score": "0.8973", + "raw_score_field": "score", + "raw_model_id": "granite-3.3-8b-base", + "raw_benchmark_id": "humaneval", + "source_urls_json": "[\"https://llm-stats.com/models/granite-3.3-8b-base\",\"https://llm-stats.com/benchmarks/humaneval\",\"https://api.llm-stats.com/leaderboard/benchmarks/humaneval\"]", + "raw_score_id": "humaneval::granite-3.3-8b-base", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "humaneval::humaneval-granite-3.3-8b-base", + "evaluation_name": "llm_stats.humaneval", + "source_data": { + "dataset_name": "HumanEval+", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/granite-3.3-8b-base", + "https://llm-stats.com/benchmarks/humaneval", + "https://api.llm-stats.com/leaderboard/benchmarks/humaneval+" + ], + "additional_details": { + "raw_benchmark_id": "humaneval+", + "raw_model_id": "granite-3.3-8b-base", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Enhanced version of HumanEval that extends the original test cases by 80x using EvalPlus framework for rigorous evaluation of LLM-synthesized code functional correctness, detecting previously undetected wrong code", + "metric_id": "llm_stats.humaneval.score", + "metric_name": "HumanEval+ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "humaneval+", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HumanEval+", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.8609, + "details": { + "raw_score": "0.8609", + "raw_score_field": "score", + "raw_model_id": "granite-3.3-8b-base", + "raw_benchmark_id": "humaneval+", + "source_urls_json": "[\"https://llm-stats.com/models/granite-3.3-8b-base\",\"https://llm-stats.com/benchmarks/humaneval\",\"https://api.llm-stats.com/leaderboard/benchmarks/humaneval+\"]", + "raw_score_id": "humaneval+::granite-3.3-8b-base", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "nq::nq-granite-3.3-8b-base", + "evaluation_name": "llm_stats.nq", + "source_data": { + "dataset_name": "NQ", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/granite-3.3-8b-base", + "https://llm-stats.com/benchmarks/nq", + "https://api.llm-stats.com/leaderboard/benchmarks/nq" + ], + "additional_details": { + "raw_benchmark_id": "nq", + "raw_model_id": "granite-3.3-8b-base", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Natural Questions (NQ) benchmark containing real user questions issued to Google search with answers found from Wikipedia, designed for training and evaluation of automatic question answering systems", + "metric_id": "llm_stats.nq.score", + "metric_name": "NQ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "nq", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "NQ", + "raw_categories": "[\"reasoning\",\"search\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.365, + "details": { + "raw_score": "0.365", + "raw_score_field": "score", + "raw_model_id": "granite-3.3-8b-base", + "raw_benchmark_id": "nq", + "source_urls_json": "[\"https://llm-stats.com/models/granite-3.3-8b-base\",\"https://llm-stats.com/benchmarks/nq\",\"https://api.llm-stats.com/leaderboard/benchmarks/nq\"]", + "raw_score_id": "nq::granite-3.3-8b-base", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "popqa::popqa-granite-3.3-8b-base", + "evaluation_name": "llm_stats.popqa", + "source_data": { + "dataset_name": "PopQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/granite-3.3-8b-base", + "https://llm-stats.com/benchmarks/popqa", + "https://api.llm-stats.com/leaderboard/benchmarks/popqa" + ], + "additional_details": { + "raw_benchmark_id": "popqa", + "raw_model_id": "granite-3.3-8b-base", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "PopQA is an entity-centric open-domain question-answering dataset consisting of 14,000 QA pairs designed to evaluate language models' ability to memorize and recall factual knowledge across entities with varying popularity levels. The dataset probes both parametric memory (stored in model parameters) and non-parametric memory effectiveness, with questions covering 16 diverse relationship types from Wikidata converted to natural language using templates. Created by sampling knowledge triples from Wikidata and converting them to natural language questions, focusing on long-tail entities to understand LMs' strengths and limitations in memorizing factual knowledge.", + "metric_id": "llm_stats.popqa.score", + "metric_name": "PopQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "popqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PopQA", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.2617, + "details": { + "raw_score": "0.2617", + "raw_score_field": "score", + "raw_model_id": "granite-3.3-8b-base", + "raw_benchmark_id": "popqa", + "source_urls_json": "[\"https://llm-stats.com/models/granite-3.3-8b-base\",\"https://llm-stats.com/benchmarks/popqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/popqa\"]", + "raw_score_id": "popqa::granite-3.3-8b-base", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "triviaqa::triviaqa-granite-3.3-8b-base", + "evaluation_name": "llm_stats.triviaqa", + "source_data": { + "dataset_name": "TriviaQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/granite-3.3-8b-base", + "https://llm-stats.com/benchmarks/triviaqa", + "https://api.llm-stats.com/leaderboard/benchmarks/triviaqa" + ], + "additional_details": { + "raw_benchmark_id": "triviaqa", + "raw_model_id": "granite-3.3-8b-base", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A large-scale reading comprehension dataset containing over 650K question-answer-evidence triples. TriviaQA includes 95K question-answer pairs authored by trivia enthusiasts and independently gathered evidence documents (six per question on average) that provide high quality distant supervision for answering the questions. The dataset features relatively complex, compositional questions with considerable syntactic and lexical variability, requiring cross-sentence reasoning to find answers.", + "metric_id": "llm_stats.triviaqa.score", + "metric_name": "TriviaQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "triviaqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TriviaQA", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.7818, + "details": { + "raw_score": "0.7818", + "raw_score_field": "score", + "raw_model_id": "granite-3.3-8b-base", + "raw_benchmark_id": "triviaqa", + "source_urls_json": "[\"https://llm-stats.com/models/granite-3.3-8b-base\",\"https://llm-stats.com/benchmarks/triviaqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/triviaqa\"]", + "raw_score_id": "triviaqa::granite-3.3-8b-base", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "truthfulqa::truthfulqa-granite-3.3-8b-base", + "evaluation_name": "llm_stats.truthfulqa", + "source_data": { + "dataset_name": "TruthfulQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/granite-3.3-8b-base", + "https://llm-stats.com/benchmarks/truthfulqa", + "https://api.llm-stats.com/leaderboard/benchmarks/truthfulqa" + ], + "additional_details": { + "raw_benchmark_id": "truthfulqa", + "raw_model_id": "granite-3.3-8b-base", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "TruthfulQA is a benchmark to measure whether language models are truthful in generating answers to questions. It comprises 817 questions that span 38 categories, including health, law, finance and politics. The questions are crafted such that some humans would answer falsely due to a false belief or misconception, testing models' ability to avoid generating false answers learned from human texts.", + "metric_id": "llm_stats.truthfulqa.score", + "metric_name": "TruthfulQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "truthfulqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TruthfulQA", + "raw_categories": "[\"legal\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.5215, + "details": { + "raw_score": "0.5215", + "raw_score_field": "score", + "raw_model_id": "granite-3.3-8b-base", + "raw_benchmark_id": "truthfulqa", + "source_urls_json": "[\"https://llm-stats.com/models/granite-3.3-8b-base\",\"https://llm-stats.com/benchmarks/truthfulqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/truthfulqa\"]", + "raw_score_id": "truthfulqa::granite-3.3-8b-base", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "winogrande::winogrande-granite-3.3-8b-base", + "evaluation_name": "llm_stats.winogrande", + "source_data": { + "dataset_name": "Winogrande", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/granite-3.3-8b-base", + "https://llm-stats.com/benchmarks/winogrande", + "https://api.llm-stats.com/leaderboard/benchmarks/winogrande" + ], + "additional_details": { + "raw_benchmark_id": "winogrande", + "raw_model_id": "granite-3.3-8b-base", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WinoGrande: An Adversarial Winograd Schema Challenge at Scale. A large-scale dataset of 44,000 pronoun resolution problems designed to test machine commonsense reasoning. Uses adversarial filtering to reduce spurious biases and provides a more robust evaluation of whether AI systems truly understand commonsense or exploit statistical shortcuts. Current best AI methods achieve 59.4-79.1% accuracy, significantly below human performance of 94.0%.", + "metric_id": "llm_stats.winogrande.score", + "metric_name": "Winogrande score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "winogrande", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Winogrande", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "22" + } + }, + "score_details": { + "score": 0.744, + "details": { + "raw_score": "0.744", + "raw_score_field": "score", + "raw_model_id": "granite-3.3-8b-base", + "raw_benchmark_id": "winogrande", + "source_urls_json": "[\"https://llm-stats.com/models/granite-3.3-8b-base\",\"https://llm-stats.com/benchmarks/winogrande\",\"https://api.llm-stats.com/leaderboard/benchmarks/winogrande\"]", + "raw_score_id": "winogrande::granite-3.3-8b-base", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/ibm/granite-3.3-8b-instruct/e33f5b82-c5b0-46ea-9e9c-91b788294520.json b/data/llm_stats/ibm/granite-3.3-8b-instruct/e33f5b82-c5b0-46ea-9e9c-91b788294520.json new file mode 100644 index 0000000000000000000000000000000000000000..7e661777f727c1c68eec2041f7759991e5a7f86b --- /dev/null +++ b/data/llm_stats/ibm/granite-3.3-8b-instruct/e33f5b82-c5b0-46ea-9e9c-91b788294520.json @@ -0,0 +1,463 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/ibm_granite-3.3-8b-instruct/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Granite 3.3 8B Instruct", + "id": "ibm/granite-3.3-8b-instruct", + "developer": "ibm", + "additional_details": { + "raw_id": "granite-3.3-8b-instruct", + "raw_name": "Granite 3.3 8B Instruct", + "raw_model_id": "granite-3.3-8b-instruct", + "raw_model_name": "Granite 3.3 8B Instruct", + "raw_organization_id": "ibm", + "raw_organization_name": "IBM", + "raw_release_date": "2025-04-16", + "raw_announcement_date": "2025-04-16", + "raw_multimodal": "true", + "raw_param_count": "8000000000", + "raw_provider_slug": "ibm", + "raw_provider_name": "IBM" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "alpacaeval-2.0::alpacaeval-2.0-granite-3.3-8b-instruct", + "evaluation_name": "llm_stats.alpacaeval-2.0", + "source_data": { + "dataset_name": "AlpacaEval 2.0", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/granite-3.3-8b-instruct", + "https://llm-stats.com/benchmarks/alpacaeval-2.0", + "https://api.llm-stats.com/leaderboard/benchmarks/alpacaeval-2.0" + ], + "additional_details": { + "raw_benchmark_id": "alpacaeval-2.0", + "raw_model_id": "granite-3.3-8b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "AlpacaEval 2.0 is a length-controlled automatic evaluator for instruction-following language models that uses GPT-4 Turbo to assess model responses against a baseline. It evaluates models on 805 diverse instruction-following tasks including creative writing, classification, programming, and general knowledge questions. The benchmark achieves 0.98 Spearman correlation with ChatBot Arena while being fast (< 3 minutes) and affordable (< $10 in OpenAI credits). It addresses length bias in automatic evaluation through length-controlled win-rates and uses weighted scoring based on response quality.", + "metric_id": "llm_stats.alpacaeval-2.0.score", + "metric_name": "AlpacaEval 2.0 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "alpacaeval-2.0", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AlpacaEval 2.0", + "raw_categories": "[\"reasoning\",\"general\",\"creativity\",\"writing\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.6268, + "details": { + "raw_score": "0.6268", + "raw_score_field": "score", + "raw_model_id": "granite-3.3-8b-instruct", + "raw_benchmark_id": "alpacaeval-2.0", + "source_urls_json": "[\"https://llm-stats.com/models/granite-3.3-8b-instruct\",\"https://llm-stats.com/benchmarks/alpacaeval-2.0\",\"https://api.llm-stats.com/leaderboard/benchmarks/alpacaeval-2.0\"]", + "raw_score_id": "alpacaeval-2.0::granite-3.3-8b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "arena-hard::arena-hard-granite-3.3-8b-instruct", + "evaluation_name": "llm_stats.arena-hard", + "source_data": { + "dataset_name": "Arena Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/granite-3.3-8b-instruct", + "https://llm-stats.com/benchmarks/arena-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/arena-hard" + ], + "additional_details": { + "raw_benchmark_id": "arena-hard", + "raw_model_id": "granite-3.3-8b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Arena-Hard-Auto is an automatic evaluation benchmark for instruction-tuned LLMs consisting of 500 challenging real-world prompts curated by BenchBuilder. It includes open-ended software engineering problems, mathematical questions, and creative writing tasks. The benchmark uses LLM-as-a-Judge methodology with GPT-4.1 and Gemini-2.5 as automatic judges to approximate human preference. Arena-Hard achieves 98.6% correlation with human preference rankings and provides 3x higher separation of model performances compared to MT-Bench, making it highly effective for distinguishing between models of similar quality.", + "metric_id": "llm_stats.arena-hard.score", + "metric_name": "Arena Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arena-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Arena Hard", + "raw_categories": "[\"reasoning\",\"general\",\"creativity\",\"writing\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "26" + } + }, + "score_details": { + "score": 0.5756, + "details": { + "raw_score": "0.5756", + "raw_score_field": "score", + "raw_model_id": "granite-3.3-8b-instruct", + "raw_benchmark_id": "arena-hard", + "source_urls_json": "[\"https://llm-stats.com/models/granite-3.3-8b-instruct\",\"https://llm-stats.com/benchmarks/arena-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/arena-hard\"]", + "raw_score_id": "arena-hard::granite-3.3-8b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "attaq::attaq-granite-3.3-8b-instruct", + "evaluation_name": "llm_stats.attaq", + "source_data": { + "dataset_name": "AttaQ", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/granite-3.3-8b-instruct", + "https://llm-stats.com/benchmarks/attaq", + "https://api.llm-stats.com/leaderboard/benchmarks/attaq" + ], + "additional_details": { + "raw_benchmark_id": "attaq", + "raw_model_id": "granite-3.3-8b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "AttaQ is a unique dataset containing adversarial examples in the form of questions designed to provoke harmful or inappropriate responses from large language models. The benchmark evaluates safety vulnerabilities by using specialized clustering techniques that analyze both the semantic similarity of input attacks and the harmfulness of model responses, facilitating targeted improvements to model safety mechanisms.", + "metric_id": "llm_stats.attaq.score", + "metric_name": "AttaQ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "attaq", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AttaQ", + "raw_categories": "[\"safety\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.885, + "details": { + "raw_score": "0.885", + "raw_score_field": "score", + "raw_model_id": "granite-3.3-8b-instruct", + "raw_benchmark_id": "attaq", + "source_urls_json": "[\"https://llm-stats.com/models/granite-3.3-8b-instruct\",\"https://llm-stats.com/benchmarks/attaq\",\"https://api.llm-stats.com/leaderboard/benchmarks/attaq\"]", + "raw_score_id": "attaq::granite-3.3-8b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "big-bench-hard::big-bench-hard-granite-3.3-8b-instruct", + "evaluation_name": "llm_stats.big-bench-hard", + "source_data": { + "dataset_name": "BIG-Bench Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/granite-3.3-8b-instruct", + "https://llm-stats.com/benchmarks/big-bench-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard" + ], + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_model_id": "granite-3.3-8b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIG-Bench Hard (BBH) is a subset of 23 challenging BIG-Bench tasks selected because prior language model evaluations did not outperform average human-rater performance. The benchmark contains 6,511 evaluation examples testing various forms of multi-step reasoning including arithmetic, logical reasoning (Boolean expressions, logical deduction), geometric reasoning, temporal reasoning, and language understanding. Tasks require capabilities such as causal judgment, object counting, navigation, pattern recognition, and complex problem solving.", + "metric_id": "llm_stats.big-bench-hard.score", + "metric_name": "BIG-Bench Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench Hard", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "21" + } + }, + "score_details": { + "score": 0.6913, + "details": { + "raw_score": "0.6913", + "raw_score_field": "score", + "raw_model_id": "granite-3.3-8b-instruct", + "raw_benchmark_id": "big-bench-hard", + "source_urls_json": "[\"https://llm-stats.com/models/granite-3.3-8b-instruct\",\"https://llm-stats.com/benchmarks/big-bench-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard\"]", + "raw_score_id": "big-bench-hard::granite-3.3-8b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "humaneval::humaneval-granite-3.3-8b-instruct", + "evaluation_name": "llm_stats.humaneval", + "source_data": { + "dataset_name": "HumanEval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/granite-3.3-8b-instruct", + "https://llm-stats.com/benchmarks/humaneval", + "https://api.llm-stats.com/leaderboard/benchmarks/humaneval" + ], + "additional_details": { + "raw_benchmark_id": "humaneval", + "raw_model_id": "granite-3.3-8b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A benchmark that measures functional correctness for synthesizing programs from docstrings, consisting of 164 original programming problems assessing language comprehension, algorithms, and simple mathematics", + "metric_id": "llm_stats.humaneval.score", + "metric_name": "HumanEval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "humaneval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HumanEval", + "raw_categories": "[\"reasoning\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "66" + } + }, + "score_details": { + "score": 0.8973, + "details": { + "raw_score": "0.8973", + "raw_score_field": "score", + "raw_model_id": "granite-3.3-8b-instruct", + "raw_benchmark_id": "humaneval", + "source_urls_json": "[\"https://llm-stats.com/models/granite-3.3-8b-instruct\",\"https://llm-stats.com/benchmarks/humaneval\",\"https://api.llm-stats.com/leaderboard/benchmarks/humaneval\"]", + "raw_score_id": "humaneval::granite-3.3-8b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "humaneval::humaneval-granite-3.3-8b-instruct", + "evaluation_name": "llm_stats.humaneval", + "source_data": { + "dataset_name": "HumanEval+", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/granite-3.3-8b-instruct", + "https://llm-stats.com/benchmarks/humaneval", + "https://api.llm-stats.com/leaderboard/benchmarks/humaneval+" + ], + "additional_details": { + "raw_benchmark_id": "humaneval+", + "raw_model_id": "granite-3.3-8b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Enhanced version of HumanEval that extends the original test cases by 80x using EvalPlus framework for rigorous evaluation of LLM-synthesized code functional correctness, detecting previously undetected wrong code", + "metric_id": "llm_stats.humaneval.score", + "metric_name": "HumanEval+ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "humaneval+", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HumanEval+", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.8609, + "details": { + "raw_score": "0.8609", + "raw_score_field": "score", + "raw_model_id": "granite-3.3-8b-instruct", + "raw_benchmark_id": "humaneval+", + "source_urls_json": "[\"https://llm-stats.com/models/granite-3.3-8b-instruct\",\"https://llm-stats.com/benchmarks/humaneval\",\"https://api.llm-stats.com/leaderboard/benchmarks/humaneval+\"]", + "raw_score_id": "humaneval+::granite-3.3-8b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "popqa::popqa-granite-3.3-8b-instruct", + "evaluation_name": "llm_stats.popqa", + "source_data": { + "dataset_name": "PopQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/granite-3.3-8b-instruct", + "https://llm-stats.com/benchmarks/popqa", + "https://api.llm-stats.com/leaderboard/benchmarks/popqa" + ], + "additional_details": { + "raw_benchmark_id": "popqa", + "raw_model_id": "granite-3.3-8b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "PopQA is an entity-centric open-domain question-answering dataset consisting of 14,000 QA pairs designed to evaluate language models' ability to memorize and recall factual knowledge across entities with varying popularity levels. The dataset probes both parametric memory (stored in model parameters) and non-parametric memory effectiveness, with questions covering 16 diverse relationship types from Wikidata converted to natural language using templates. Created by sampling knowledge triples from Wikidata and converting them to natural language questions, focusing on long-tail entities to understand LMs' strengths and limitations in memorizing factual knowledge.", + "metric_id": "llm_stats.popqa.score", + "metric_name": "PopQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "popqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PopQA", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.2617, + "details": { + "raw_score": "0.2617", + "raw_score_field": "score", + "raw_model_id": "granite-3.3-8b-instruct", + "raw_benchmark_id": "popqa", + "source_urls_json": "[\"https://llm-stats.com/models/granite-3.3-8b-instruct\",\"https://llm-stats.com/benchmarks/popqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/popqa\"]", + "raw_score_id": "popqa::granite-3.3-8b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "truthfulqa::truthfulqa-granite-3.3-8b-instruct", + "evaluation_name": "llm_stats.truthfulqa", + "source_data": { + "dataset_name": "TruthfulQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/granite-3.3-8b-instruct", + "https://llm-stats.com/benchmarks/truthfulqa", + "https://api.llm-stats.com/leaderboard/benchmarks/truthfulqa" + ], + "additional_details": { + "raw_benchmark_id": "truthfulqa", + "raw_model_id": "granite-3.3-8b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "TruthfulQA is a benchmark to measure whether language models are truthful in generating answers to questions. It comprises 817 questions that span 38 categories, including health, law, finance and politics. The questions are crafted such that some humans would answer falsely due to a false belief or misconception, testing models' ability to avoid generating false answers learned from human texts.", + "metric_id": "llm_stats.truthfulqa.score", + "metric_name": "TruthfulQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "truthfulqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TruthfulQA", + "raw_categories": "[\"legal\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.6686, + "details": { + "raw_score": "0.6686", + "raw_score_field": "score", + "raw_model_id": "granite-3.3-8b-instruct", + "raw_benchmark_id": "truthfulqa", + "source_urls_json": "[\"https://llm-stats.com/models/granite-3.3-8b-instruct\",\"https://llm-stats.com/benchmarks/truthfulqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/truthfulqa\"]", + "raw_score_id": "truthfulqa::granite-3.3-8b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/ibm/granite-4.0-tiny-preview/65e01c5e-45f9-4005-870a-a471afed257d.json b/data/llm_stats/ibm/granite-4.0-tiny-preview/65e01c5e-45f9-4005-870a-a471afed257d.json new file mode 100644 index 0000000000000000000000000000000000000000..22d417825d15823e647f23cd4f63f36e6902499d --- /dev/null +++ b/data/llm_stats/ibm/granite-4.0-tiny-preview/65e01c5e-45f9-4005-870a-a471afed257d.json @@ -0,0 +1,359 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/ibm_granite-4.0-tiny-preview/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "IBM Granite 4.0 Tiny Preview", + "id": "ibm/granite-4.0-tiny-preview", + "developer": "ibm", + "additional_details": { + "raw_id": "granite-4.0-tiny-preview", + "raw_name": "IBM Granite 4.0 Tiny Preview", + "raw_model_id": "granite-4.0-tiny-preview", + "raw_model_name": "IBM Granite 4.0 Tiny Preview", + "raw_organization_id": "ibm", + "raw_organization_name": "IBM", + "raw_release_date": "2025-05-02", + "raw_announcement_date": "2025-05-02", + "raw_multimodal": "false", + "raw_param_count": "7000000000", + "raw_provider_slug": "ibm", + "raw_provider_name": "IBM" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "alpacaeval-2.0::alpacaeval-2.0-granite-4.0-tiny-preview", + "evaluation_name": "llm_stats.alpacaeval-2.0", + "source_data": { + "dataset_name": "AlpacaEval 2.0", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/granite-4.0-tiny-preview", + "https://llm-stats.com/benchmarks/alpacaeval-2.0", + "https://api.llm-stats.com/leaderboard/benchmarks/alpacaeval-2.0" + ], + "additional_details": { + "raw_benchmark_id": "alpacaeval-2.0", + "raw_model_id": "granite-4.0-tiny-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "AlpacaEval 2.0 is a length-controlled automatic evaluator for instruction-following language models that uses GPT-4 Turbo to assess model responses against a baseline. It evaluates models on 805 diverse instruction-following tasks including creative writing, classification, programming, and general knowledge questions. The benchmark achieves 0.98 Spearman correlation with ChatBot Arena while being fast (< 3 minutes) and affordable (< $10 in OpenAI credits). It addresses length bias in automatic evaluation through length-controlled win-rates and uses weighted scoring based on response quality.", + "metric_id": "llm_stats.alpacaeval-2.0.score", + "metric_name": "AlpacaEval 2.0 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "alpacaeval-2.0", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AlpacaEval 2.0", + "raw_categories": "[\"reasoning\",\"general\",\"creativity\",\"writing\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.3516, + "details": { + "raw_score": "0.3516", + "raw_score_field": "score", + "raw_model_id": "granite-4.0-tiny-preview", + "raw_benchmark_id": "alpacaeval-2.0", + "source_urls_json": "[\"https://llm-stats.com/models/granite-4.0-tiny-preview\",\"https://llm-stats.com/benchmarks/alpacaeval-2.0\",\"https://api.llm-stats.com/leaderboard/benchmarks/alpacaeval-2.0\"]", + "raw_score_id": "alpacaeval-2.0::granite-4.0-tiny-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "attaq::attaq-granite-4.0-tiny-preview", + "evaluation_name": "llm_stats.attaq", + "source_data": { + "dataset_name": "AttaQ", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/granite-4.0-tiny-preview", + "https://llm-stats.com/benchmarks/attaq", + "https://api.llm-stats.com/leaderboard/benchmarks/attaq" + ], + "additional_details": { + "raw_benchmark_id": "attaq", + "raw_model_id": "granite-4.0-tiny-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "AttaQ is a unique dataset containing adversarial examples in the form of questions designed to provoke harmful or inappropriate responses from large language models. The benchmark evaluates safety vulnerabilities by using specialized clustering techniques that analyze both the semantic similarity of input attacks and the harmfulness of model responses, facilitating targeted improvements to model safety mechanisms.", + "metric_id": "llm_stats.attaq.score", + "metric_name": "AttaQ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "attaq", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AttaQ", + "raw_categories": "[\"safety\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.861, + "details": { + "raw_score": "0.861", + "raw_score_field": "score", + "raw_model_id": "granite-4.0-tiny-preview", + "raw_benchmark_id": "attaq", + "source_urls_json": "[\"https://llm-stats.com/models/granite-4.0-tiny-preview\",\"https://llm-stats.com/benchmarks/attaq\",\"https://api.llm-stats.com/leaderboard/benchmarks/attaq\"]", + "raw_score_id": "attaq::granite-4.0-tiny-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "big-bench-hard::big-bench-hard-granite-4.0-tiny-preview", + "evaluation_name": "llm_stats.big-bench-hard", + "source_data": { + "dataset_name": "BIG-Bench Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/granite-4.0-tiny-preview", + "https://llm-stats.com/benchmarks/big-bench-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard" + ], + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_model_id": "granite-4.0-tiny-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIG-Bench Hard (BBH) is a subset of 23 challenging BIG-Bench tasks selected because prior language model evaluations did not outperform average human-rater performance. The benchmark contains 6,511 evaluation examples testing various forms of multi-step reasoning including arithmetic, logical reasoning (Boolean expressions, logical deduction), geometric reasoning, temporal reasoning, and language understanding. Tasks require capabilities such as causal judgment, object counting, navigation, pattern recognition, and complex problem solving.", + "metric_id": "llm_stats.big-bench-hard.score", + "metric_name": "BIG-Bench Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench Hard", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "21" + } + }, + "score_details": { + "score": 0.557, + "details": { + "raw_score": "0.557", + "raw_score_field": "score", + "raw_model_id": "granite-4.0-tiny-preview", + "raw_benchmark_id": "big-bench-hard", + "source_urls_json": "[\"https://llm-stats.com/models/granite-4.0-tiny-preview\",\"https://llm-stats.com/benchmarks/big-bench-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard\"]", + "raw_score_id": "big-bench-hard::granite-4.0-tiny-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "humaneval::humaneval-granite-4.0-tiny-preview", + "evaluation_name": "llm_stats.humaneval", + "source_data": { + "dataset_name": "HumanEval+", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/granite-4.0-tiny-preview", + "https://llm-stats.com/benchmarks/humaneval", + "https://api.llm-stats.com/leaderboard/benchmarks/humaneval+" + ], + "additional_details": { + "raw_benchmark_id": "humaneval+", + "raw_model_id": "granite-4.0-tiny-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Enhanced version of HumanEval that extends the original test cases by 80x using EvalPlus framework for rigorous evaluation of LLM-synthesized code functional correctness, detecting previously undetected wrong code", + "metric_id": "llm_stats.humaneval.score", + "metric_name": "HumanEval+ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "humaneval+", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HumanEval+", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.783, + "details": { + "raw_score": "0.783", + "raw_score_field": "score", + "raw_model_id": "granite-4.0-tiny-preview", + "raw_benchmark_id": "humaneval+", + "source_urls_json": "[\"https://llm-stats.com/models/granite-4.0-tiny-preview\",\"https://llm-stats.com/benchmarks/humaneval\",\"https://api.llm-stats.com/leaderboard/benchmarks/humaneval+\"]", + "raw_score_id": "humaneval+::granite-4.0-tiny-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "popqa::popqa-granite-4.0-tiny-preview", + "evaluation_name": "llm_stats.popqa", + "source_data": { + "dataset_name": "PopQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/granite-4.0-tiny-preview", + "https://llm-stats.com/benchmarks/popqa", + "https://api.llm-stats.com/leaderboard/benchmarks/popqa" + ], + "additional_details": { + "raw_benchmark_id": "popqa", + "raw_model_id": "granite-4.0-tiny-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "PopQA is an entity-centric open-domain question-answering dataset consisting of 14,000 QA pairs designed to evaluate language models' ability to memorize and recall factual knowledge across entities with varying popularity levels. The dataset probes both parametric memory (stored in model parameters) and non-parametric memory effectiveness, with questions covering 16 diverse relationship types from Wikidata converted to natural language using templates. Created by sampling knowledge triples from Wikidata and converting them to natural language questions, focusing on long-tail entities to understand LMs' strengths and limitations in memorizing factual knowledge.", + "metric_id": "llm_stats.popqa.score", + "metric_name": "PopQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "popqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PopQA", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.229, + "details": { + "raw_score": "0.229", + "raw_score_field": "score", + "raw_model_id": "granite-4.0-tiny-preview", + "raw_benchmark_id": "popqa", + "source_urls_json": "[\"https://llm-stats.com/models/granite-4.0-tiny-preview\",\"https://llm-stats.com/benchmarks/popqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/popqa\"]", + "raw_score_id": "popqa::granite-4.0-tiny-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "truthfulqa::truthfulqa-granite-4.0-tiny-preview", + "evaluation_name": "llm_stats.truthfulqa", + "source_data": { + "dataset_name": "TruthfulQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/granite-4.0-tiny-preview", + "https://llm-stats.com/benchmarks/truthfulqa", + "https://api.llm-stats.com/leaderboard/benchmarks/truthfulqa" + ], + "additional_details": { + "raw_benchmark_id": "truthfulqa", + "raw_model_id": "granite-4.0-tiny-preview", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "TruthfulQA is a benchmark to measure whether language models are truthful in generating answers to questions. It comprises 817 questions that span 38 categories, including health, law, finance and politics. The questions are crafted such that some humans would answer falsely due to a false belief or misconception, testing models' ability to avoid generating false answers learned from human texts.", + "metric_id": "llm_stats.truthfulqa.score", + "metric_name": "TruthfulQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "truthfulqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TruthfulQA", + "raw_categories": "[\"legal\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.581, + "details": { + "raw_score": "0.581", + "raw_score_field": "score", + "raw_model_id": "granite-4.0-tiny-preview", + "raw_benchmark_id": "truthfulqa", + "source_urls_json": "[\"https://llm-stats.com/models/granite-4.0-tiny-preview\",\"https://llm-stats.com/benchmarks/truthfulqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/truthfulqa\"]", + "raw_score_id": "truthfulqa::granite-4.0-tiny-preview", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/microsoft/mai-code-1-flash/413a6465-071d-4067-909f-c2e998650554.json b/data/llm_stats/microsoft/mai-code-1-flash/413a6465-071d-4067-909f-c2e998650554.json new file mode 100644 index 0000000000000000000000000000000000000000..dbced47f1b23cdf7d0fd384723d3ae824c9f514e --- /dev/null +++ b/data/llm_stats/microsoft/mai-code-1-flash/413a6465-071d-4067-909f-c2e998650554.json @@ -0,0 +1,410 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/microsoft_mai-code-1-flash/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "MAI-Code-1-Flash", + "id": "microsoft/mai-code-1-flash", + "developer": "microsoft", + "additional_details": { + "raw_id": "mai-code-1-flash", + "raw_name": "MAI-Code-1-Flash", + "raw_model_id": "mai-code-1-flash", + "raw_model_name": "MAI-Code-1-Flash", + "raw_organization_id": "microsoft", + "raw_organization_name": "Microsoft", + "raw_release_date": "2026-06-02", + "raw_announcement_date": "2026-06-02", + "raw_multimodal": "false", + "raw_provider_slug": "microsoft", + "raw_provider_name": "Microsoft" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "advancedif::advancedif-mai-code-1-flash", + "evaluation_name": "llm_stats.advancedif", + "source_data": { + "dataset_name": "AdvancedIF", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mai-code-1-flash", + "https://llm-stats.com/benchmarks/advancedif", + "https://api.llm-stats.com/leaderboard/benchmarks/advancedif" + ], + "additional_details": { + "raw_benchmark_id": "advancedif", + "raw_model_id": "mai-code-1-flash", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "AdvancedIF is a rubric-based benchmark measuring complex, multi-turn, and system-prompted instruction following ability, scored with a calibrated LLM judge against per-instruction rubrics.", + "metric_id": "llm_stats.advancedif.score", + "metric_name": "AdvancedIF score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "advancedif", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AdvancedIF", + "raw_categories": "[\"instruction_following\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.714, + "details": { + "raw_score": "0.714", + "raw_score_field": "score", + "raw_model_id": "mai-code-1-flash", + "raw_benchmark_id": "advancedif", + "source_urls_json": "[\"https://llm-stats.com/models/mai-code-1-flash\",\"https://llm-stats.com/benchmarks/advancedif\",\"https://api.llm-stats.com/leaderboard/benchmarks/advancedif\"]", + "raw_score_id": "advancedif::mai-code-1-flash", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "aime-2026::aime-2026-mai-code-1-flash", + "evaluation_name": "llm_stats.aime-2026", + "source_data": { + "dataset_name": "AIME 2026", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mai-code-1-flash", + "https://llm-stats.com/benchmarks/aime-2026", + "https://api.llm-stats.com/leaderboard/benchmarks/aime-2026" + ], + "additional_details": { + "raw_benchmark_id": "aime-2026", + "raw_model_id": "mai-code-1-flash", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "All 30 problems from the 2026 American Invitational Mathematics Examination (AIME I and AIME II), testing olympiad-level mathematical reasoning with integer answers from 000-999. Used as an AI benchmark to evaluate large language models' ability to solve complex mathematical problems requiring multi-step logical deductions and structured symbolic reasoning.", + "metric_id": "llm_stats.aime-2026.score", + "metric_name": "AIME 2026 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aime-2026", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AIME 2026", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "17" + } + }, + "score_details": { + "score": 0.925, + "details": { + "raw_score": "0.925", + "raw_score_field": "score", + "raw_model_id": "mai-code-1-flash", + "raw_benchmark_id": "aime-2026", + "source_urls_json": "[\"https://llm-stats.com/models/mai-code-1-flash\",\"https://llm-stats.com/benchmarks/aime-2026\",\"https://api.llm-stats.com/leaderboard/benchmarks/aime-2026\"]", + "raw_score_id": "aime-2026::mai-code-1-flash", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "amo-bench::amo-bench-mai-code-1-flash", + "evaluation_name": "llm_stats.amo-bench", + "source_data": { + "dataset_name": "AMO Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mai-code-1-flash", + "https://llm-stats.com/benchmarks/amo-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/amo-bench" + ], + "additional_details": { + "raw_benchmark_id": "amo-bench", + "raw_model_id": "mai-code-1-flash", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "AMO Bench is an olympiad-level mathematics benchmark that evaluates advanced mathematical problem-solving and multi-step reasoning on competition-style problems.", + "metric_id": "llm_stats.amo-bench.score", + "metric_name": "AMO Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "amo-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AMO Bench", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.4, + "details": { + "raw_score": "0.4", + "raw_score_field": "score", + "raw_model_id": "mai-code-1-flash", + "raw_benchmark_id": "amo-bench", + "source_urls_json": "[\"https://llm-stats.com/models/mai-code-1-flash\",\"https://llm-stats.com/benchmarks/amo-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/amo-bench\"]", + "raw_score_id": "amo-bench::mai-code-1-flash", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "artifacts-bench::artifacts-bench-mai-code-1-flash", + "evaluation_name": "llm_stats.artifacts-bench", + "source_data": { + "dataset_name": "Artifacts Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mai-code-1-flash", + "https://llm-stats.com/benchmarks/artifacts-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/artifacts-bench" + ], + "additional_details": { + "raw_benchmark_id": "artifacts-bench", + "raw_model_id": "mai-code-1-flash", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Artifacts Bench evaluates a model's ability to generate visual code artifacts, measuring the quality of generated interactive and visual front-end outputs from natural-language requests.", + "metric_id": "llm_stats.artifacts-bench.score", + "metric_name": "Artifacts Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "artifacts-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Artifacts Bench", + "raw_categories": "[\"frontend_development\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.364, + "details": { + "raw_score": "0.364", + "raw_score_field": "score", + "raw_model_id": "mai-code-1-flash", + "raw_benchmark_id": "artifacts-bench", + "source_urls_json": "[\"https://llm-stats.com/models/mai-code-1-flash\",\"https://llm-stats.com/benchmarks/artifacts-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/artifacts-bench\"]", + "raw_score_id": "artifacts-bench::mai-code-1-flash", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "frontier-science::frontier-science-mai-code-1-flash", + "evaluation_name": "llm_stats.frontier-science", + "source_data": { + "dataset_name": "Frontier Science", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mai-code-1-flash", + "https://llm-stats.com/benchmarks/frontier-science", + "https://api.llm-stats.com/leaderboard/benchmarks/frontier-science" + ], + "additional_details": { + "raw_benchmark_id": "frontier-science", + "raw_model_id": "mai-code-1-flash", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Frontier Science is a benchmark of exceptionally challenging scientific reasoning problems spanning advanced natural-science domains, designed to test expert-level scientific understanding and multi-step reasoning.", + "metric_id": "llm_stats.frontier-science.score", + "metric_name": "Frontier Science score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "frontier-science", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Frontier Science", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.582, + "details": { + "raw_score": "0.582", + "raw_score_field": "score", + "raw_model_id": "mai-code-1-flash", + "raw_benchmark_id": "frontier-science", + "source_urls_json": "[\"https://llm-stats.com/models/mai-code-1-flash\",\"https://llm-stats.com/benchmarks/frontier-science\",\"https://api.llm-stats.com/leaderboard/benchmarks/frontier-science\"]", + "raw_score_id": "frontier-science::mai-code-1-flash", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "frontiermath::frontiermath-mai-code-1-flash", + "evaluation_name": "llm_stats.frontiermath", + "source_data": { + "dataset_name": "FrontierMath", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mai-code-1-flash", + "https://llm-stats.com/benchmarks/frontiermath", + "https://api.llm-stats.com/leaderboard/benchmarks/frontiermath" + ], + "additional_details": { + "raw_benchmark_id": "frontiermath", + "raw_model_id": "mai-code-1-flash", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A benchmark of hundreds of original, exceptionally challenging mathematics problems crafted and vetted by expert mathematicians, covering most major branches of modern mathematics from number theory and real analysis to algebraic geometry and category theory.", + "metric_id": "llm_stats.frontiermath.score", + "metric_name": "FrontierMath score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "frontiermath", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FrontierMath", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "14" + } + }, + "score_details": { + "score": 0.063, + "details": { + "raw_score": "0.063", + "raw_score_field": "score", + "raw_model_id": "mai-code-1-flash", + "raw_benchmark_id": "frontiermath", + "source_urls_json": "[\"https://llm-stats.com/models/mai-code-1-flash\",\"https://llm-stats.com/benchmarks/frontiermath\",\"https://api.llm-stats.com/leaderboard/benchmarks/frontiermath\"]", + "raw_score_id": "frontiermath::mai-code-1-flash", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ifbench::ifbench-mai-code-1-flash", + "evaluation_name": "llm_stats.ifbench", + "source_data": { + "dataset_name": "IFBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mai-code-1-flash", + "https://llm-stats.com/benchmarks/ifbench", + "https://api.llm-stats.com/leaderboard/benchmarks/ifbench" + ], + "additional_details": { + "raw_benchmark_id": "ifbench", + "raw_model_id": "mai-code-1-flash", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Instruction Following Benchmark evaluating model's ability to follow complex instructions", + "metric_id": "llm_stats.ifbench.score", + "metric_name": "IFBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ifbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IFBench", + "raw_categories": "[\"instruction_following\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.75, + "details": { + "raw_score": "0.75", + "raw_score_field": "score", + "raw_model_id": "mai-code-1-flash", + "raw_benchmark_id": "ifbench", + "source_urls_json": "[\"https://llm-stats.com/models/mai-code-1-flash\",\"https://llm-stats.com/benchmarks/ifbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/ifbench\"]", + "raw_score_id": "ifbench::mai-code-1-flash", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/microsoft/mai-thinking-1/86733211-b38f-4ad3-9a0a-974b30c765c5.json b/data/llm_stats/microsoft/mai-thinking-1/86733211-b38f-4ad3-9a0a-974b30c765c5.json new file mode 100644 index 0000000000000000000000000000000000000000..9ac13743430d7202f00659914e99153f113fc9a7 --- /dev/null +++ b/data/llm_stats/microsoft/mai-thinking-1/86733211-b38f-4ad3-9a0a-974b30c765c5.json @@ -0,0 +1,1035 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/microsoft_mai-thinking-1/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "MAI-Thinking-1", + "id": "microsoft/mai-thinking-1", + "developer": "microsoft", + "additional_details": { + "raw_id": "mai-thinking-1", + "raw_name": "MAI-Thinking-1", + "raw_model_id": "mai-thinking-1", + "raw_model_name": "MAI-Thinking-1", + "raw_organization_id": "microsoft", + "raw_organization_name": "Microsoft", + "raw_release_date": "2026-06-02", + "raw_announcement_date": "2026-06-02", + "raw_multimodal": "false", + "raw_param_count": "1000000000000", + "raw_provider_slug": "microsoft", + "raw_provider_name": "Microsoft" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "advancedif::advancedif-mai-thinking-1", + "evaluation_name": "llm_stats.advancedif", + "source_data": { + "dataset_name": "AdvancedIF", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mai-thinking-1", + "https://llm-stats.com/benchmarks/advancedif", + "https://api.llm-stats.com/leaderboard/benchmarks/advancedif" + ], + "additional_details": { + "raw_benchmark_id": "advancedif", + "raw_model_id": "mai-thinking-1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "AdvancedIF is a rubric-based benchmark measuring complex, multi-turn, and system-prompted instruction following ability, scored with a calibrated LLM judge against per-instruction rubrics.", + "metric_id": "llm_stats.advancedif.score", + "metric_name": "AdvancedIF score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "advancedif", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AdvancedIF", + "raw_categories": "[\"instruction_following\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.85, + "details": { + "raw_score": "0.85", + "raw_score_field": "score", + "raw_model_id": "mai-thinking-1", + "raw_benchmark_id": "advancedif", + "source_urls_json": "[\"https://llm-stats.com/models/mai-thinking-1\",\"https://llm-stats.com/benchmarks/advancedif\",\"https://api.llm-stats.com/leaderboard/benchmarks/advancedif\"]", + "raw_score_id": "advancedif::mai-thinking-1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "aime-2025::aime-2025-mai-thinking-1", + "evaluation_name": "llm_stats.aime-2025", + "source_data": { + "dataset_name": "AIME 2025", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mai-thinking-1", + "https://llm-stats.com/benchmarks/aime-2025", + "https://api.llm-stats.com/leaderboard/benchmarks/aime-2025" + ], + "additional_details": { + "raw_benchmark_id": "aime-2025", + "raw_model_id": "mai-thinking-1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "All 30 problems from the 2025 American Invitational Mathematics Examination (AIME I and AIME II), testing olympiad-level mathematical reasoning with integer answers from 000-999. Used as an AI benchmark to evaluate large language models' ability to solve complex mathematical problems requiring multi-step logical deductions and structured symbolic reasoning.", + "metric_id": "llm_stats.aime-2025.score", + "metric_name": "AIME 2025 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aime-2025", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AIME 2025", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "114" + } + }, + "score_details": { + "score": 0.97, + "details": { + "raw_score": "0.97", + "raw_score_field": "score", + "raw_model_id": "mai-thinking-1", + "raw_benchmark_id": "aime-2025", + "source_urls_json": "[\"https://llm-stats.com/models/mai-thinking-1\",\"https://llm-stats.com/benchmarks/aime-2025\",\"https://api.llm-stats.com/leaderboard/benchmarks/aime-2025\"]", + "raw_score_id": "aime-2025::mai-thinking-1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "aime-2026::aime-2026-mai-thinking-1", + "evaluation_name": "llm_stats.aime-2026", + "source_data": { + "dataset_name": "AIME 2026", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mai-thinking-1", + "https://llm-stats.com/benchmarks/aime-2026", + "https://api.llm-stats.com/leaderboard/benchmarks/aime-2026" + ], + "additional_details": { + "raw_benchmark_id": "aime-2026", + "raw_model_id": "mai-thinking-1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "All 30 problems from the 2026 American Invitational Mathematics Examination (AIME I and AIME II), testing olympiad-level mathematical reasoning with integer answers from 000-999. Used as an AI benchmark to evaluate large language models' ability to solve complex mathematical problems requiring multi-step logical deductions and structured symbolic reasoning.", + "metric_id": "llm_stats.aime-2026.score", + "metric_name": "AIME 2026 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aime-2026", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AIME 2026", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "17" + } + }, + "score_details": { + "score": 0.945, + "details": { + "raw_score": "0.945", + "raw_score_field": "score", + "raw_model_id": "mai-thinking-1", + "raw_benchmark_id": "aime-2026", + "source_urls_json": "[\"https://llm-stats.com/models/mai-thinking-1\",\"https://llm-stats.com/benchmarks/aime-2026\",\"https://api.llm-stats.com/leaderboard/benchmarks/aime-2026\"]", + "raw_score_id": "aime-2026::mai-thinking-1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "air-bench::air-bench-mai-thinking-1", + "evaluation_name": "llm_stats.air-bench", + "source_data": { + "dataset_name": "AIR-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mai-thinking-1", + "https://llm-stats.com/benchmarks/air-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/air-bench" + ], + "additional_details": { + "raw_benchmark_id": "air-bench", + "raw_model_id": "mai-thinking-1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "AIR-Bench 2024 is a safety benchmark grounded in risk categories derived from government regulations and company policies. It evaluates policy-grounded refusal across a broad regulatory and policy-derived harm taxonomy, using category-specific LLM-judge prompts that reward safe engagement rather than only penalizing unsafe responses.", + "metric_id": "llm_stats.air-bench.score", + "metric_name": "AIR-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "air-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AIR-Bench", + "raw_categories": "[\"safety\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.88, + "details": { + "raw_score": "0.88", + "raw_score_field": "score", + "raw_model_id": "mai-thinking-1", + "raw_benchmark_id": "air-bench", + "source_urls_json": "[\"https://llm-stats.com/models/mai-thinking-1\",\"https://llm-stats.com/benchmarks/air-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/air-bench\"]", + "raw_score_id": "air-bench::mai-thinking-1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bfcl-v3::bfcl-v3-mai-thinking-1", + "evaluation_name": "llm_stats.bfcl-v3", + "source_data": { + "dataset_name": "BFCL-v3", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mai-thinking-1", + "https://llm-stats.com/benchmarks/bfcl-v3", + "https://api.llm-stats.com/leaderboard/benchmarks/bfcl-v3" + ], + "additional_details": { + "raw_benchmark_id": "bfcl-v3", + "raw_model_id": "mai-thinking-1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Berkeley Function Calling Leaderboard v3 (BFCL-v3) is an advanced benchmark that evaluates large language models' function calling capabilities through multi-turn and multi-step interactions. It introduces extended conversational exchanges where models must retain contextual information across turns and execute multiple internal function calls for complex user requests. The benchmark includes 1000 test cases across domains like vehicle control, trading bots, travel booking, and file system management, using state-based evaluation to verify both system state changes and execution path correctness.", + "metric_id": "llm_stats.bfcl-v3.score", + "metric_name": "BFCL-v3 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bfcl-v3", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BFCL-v3", + "raw_categories": "[\"reasoning\",\"structured_output\",\"finance\",\"general\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "19" + } + }, + "score_details": { + "score": 0.72, + "details": { + "raw_score": "0.72", + "raw_score_field": "score", + "raw_model_id": "mai-thinking-1", + "raw_benchmark_id": "bfcl-v3", + "source_urls_json": "[\"https://llm-stats.com/models/mai-thinking-1\",\"https://llm-stats.com/benchmarks/bfcl-v3\",\"https://api.llm-stats.com/leaderboard/benchmarks/bfcl-v3\"]", + "raw_score_id": "bfcl-v3::mai-thinking-1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "corpusqa::corpusqa-mai-thinking-1", + "evaluation_name": "llm_stats.corpusqa", + "source_data": { + "dataset_name": "CorpusQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mai-thinking-1", + "https://llm-stats.com/benchmarks/corpusqa", + "https://api.llm-stats.com/leaderboard/benchmarks/corpusqa" + ], + "additional_details": { + "raw_benchmark_id": "corpusqa", + "raw_model_id": "mai-thinking-1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CorpusQA is a multi-document, free-form long-context question answering benchmark in which a model must retrieve and reason over information distributed across a large corpus to produce open-ended answers that are scored by an LLM judge.", + "metric_id": "llm_stats.corpusqa.score", + "metric_name": "CorpusQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "corpusqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CorpusQA", + "raw_categories": "[\"long_context\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.82, + "details": { + "raw_score": "0.82", + "raw_score_field": "score", + "raw_model_id": "mai-thinking-1", + "raw_benchmark_id": "corpusqa", + "source_urls_json": "[\"https://llm-stats.com/models/mai-thinking-1\",\"https://llm-stats.com/benchmarks/corpusqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/corpusqa\"]", + "raw_score_id": "corpusqa::mai-thinking-1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "cyberseceval-4::cyberseceval-4-mai-thinking-1", + "evaluation_name": "llm_stats.cyberseceval-4", + "source_data": { + "dataset_name": "CyberSecEval 4", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mai-thinking-1", + "https://llm-stats.com/benchmarks/cyberseceval-4", + "https://api.llm-stats.com/leaderboard/benchmarks/cyberseceval-4" + ], + "additional_details": { + "raw_benchmark_id": "cyberseceval-4", + "raw_model_id": "mai-thinking-1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CyberSecEval 4 is an evaluation suite covering cybersecurity-related capabilities and risks of large language models. The insecure-code-generation tracks measure whether a model produces vulnerable code: the Instruct track presents coding requests designed to elicit known insecure patterns, while the Autocomplete track prompts the model with code context leading up to a known insecure pattern, with vulnerabilities detected via static analysis.", + "metric_id": "llm_stats.cyberseceval-4.score", + "metric_name": "CyberSecEval 4 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "cyberseceval-4", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CyberSecEval 4", + "raw_categories": "[\"safety\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.63, + "details": { + "raw_score": "0.63", + "raw_score_field": "score", + "raw_model_id": "mai-thinking-1", + "raw_benchmark_id": "cyberseceval-4", + "source_urls_json": "[\"https://llm-stats.com/models/mai-thinking-1\",\"https://llm-stats.com/benchmarks/cyberseceval-4\",\"https://api.llm-stats.com/leaderboard/benchmarks/cyberseceval-4\"]", + "raw_score_id": "cyberseceval-4::mai-thinking-1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "graphwalks::graphwalks-mai-thinking-1", + "evaluation_name": "llm_stats.graphwalks", + "source_data": { + "dataset_name": "GraphWalks", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mai-thinking-1", + "https://llm-stats.com/benchmarks/graphwalks", + "https://api.llm-stats.com/leaderboard/benchmarks/graphwalks" + ], + "additional_details": { + "raw_benchmark_id": "graphwalks", + "raw_model_id": "mai-thinking-1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "GraphWalks is a synthetic multi-hop long-context reasoning benchmark in which a model is given an edge-list representation of a graph and must traverse it to find neighboring nodes (via breadth-first search) or parent nodes for a given start node. Performance is reported as F1 of the model-predicted answer set versus the ground truth.", + "metric_id": "llm_stats.graphwalks.score", + "metric_name": "GraphWalks score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "graphwalks", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GraphWalks", + "raw_categories": "[\"long_context\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.9, + "details": { + "raw_score": "0.9", + "raw_score_field": "score", + "raw_model_id": "mai-thinking-1", + "raw_benchmark_id": "graphwalks", + "source_urls_json": "[\"https://llm-stats.com/models/mai-thinking-1\",\"https://llm-stats.com/benchmarks/graphwalks\",\"https://api.llm-stats.com/leaderboard/benchmarks/graphwalks\"]", + "raw_score_id": "graphwalks::mai-thinking-1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "healthbench-professional::healthbench-professional-mai-thinking-1", + "evaluation_name": "llm_stats.healthbench-professional", + "source_data": { + "dataset_name": "HealthBench Professional", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mai-thinking-1", + "https://llm-stats.com/benchmarks/healthbench-professional", + "https://api.llm-stats.com/leaderboard/benchmarks/healthbench-professional" + ], + "additional_details": { + "raw_benchmark_id": "healthbench-professional", + "raw_model_id": "mai-thinking-1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "HealthBench Professional evaluates model capability and safety for clinician use cases using real clinician-style chats and physician-authored grading rubrics.", + "metric_id": "llm_stats.healthbench-professional.score", + "metric_name": "HealthBench Professional score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "healthbench-professional", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HealthBench Professional", + "raw_categories": "[\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.35, + "details": { + "raw_score": "0.35", + "raw_score_field": "score", + "raw_model_id": "mai-thinking-1", + "raw_benchmark_id": "healthbench-professional", + "source_urls_json": "[\"https://llm-stats.com/models/mai-thinking-1\",\"https://llm-stats.com/benchmarks/healthbench-professional\",\"https://api.llm-stats.com/leaderboard/benchmarks/healthbench-professional\"]", + "raw_score_id": "healthbench-professional::mai-thinking-1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hmmt-feb-26::hmmt-feb-26-mai-thinking-1", + "evaluation_name": "llm_stats.hmmt-feb-26", + "source_data": { + "dataset_name": "HMMT Feb 26", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mai-thinking-1", + "https://llm-stats.com/benchmarks/hmmt-feb-26", + "https://api.llm-stats.com/leaderboard/benchmarks/hmmt-feb-26" + ], + "additional_details": { + "raw_benchmark_id": "hmmt-feb-26", + "raw_model_id": "mai-thinking-1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "HMMT February 2026 is a math competition benchmark based on problems from the Harvard-MIT Mathematics Tournament, testing advanced mathematical problem-solving and reasoning.", + "metric_id": "llm_stats.hmmt-feb-26.score", + "metric_name": "HMMT Feb 26 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hmmt-feb-26", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HMMT Feb 26", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.849, + "details": { + "raw_score": "0.849", + "raw_score_field": "score", + "raw_model_id": "mai-thinking-1", + "raw_benchmark_id": "hmmt-feb-26", + "source_urls_json": "[\"https://llm-stats.com/models/mai-thinking-1\",\"https://llm-stats.com/benchmarks/hmmt-feb-26\",\"https://api.llm-stats.com/leaderboard/benchmarks/hmmt-feb-26\"]", + "raw_score_id": "hmmt-feb-26::mai-thinking-1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ifbench::ifbench-mai-thinking-1", + "evaluation_name": "llm_stats.ifbench", + "source_data": { + "dataset_name": "IFBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mai-thinking-1", + "https://llm-stats.com/benchmarks/ifbench", + "https://api.llm-stats.com/leaderboard/benchmarks/ifbench" + ], + "additional_details": { + "raw_benchmark_id": "ifbench", + "raw_model_id": "mai-thinking-1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Instruction Following Benchmark evaluating model's ability to follow complex instructions", + "metric_id": "llm_stats.ifbench.score", + "metric_name": "IFBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ifbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IFBench", + "raw_categories": "[\"instruction_following\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.69, + "details": { + "raw_score": "0.69", + "raw_score_field": "score", + "raw_model_id": "mai-thinking-1", + "raw_benchmark_id": "ifbench", + "source_urls_json": "[\"https://llm-stats.com/models/mai-thinking-1\",\"https://llm-stats.com/benchmarks/ifbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/ifbench\"]", + "raw_score_id": "ifbench::mai-thinking-1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "livecodebench-v6::livecodebench-v6-mai-thinking-1", + "evaluation_name": "llm_stats.livecodebench-v6", + "source_data": { + "dataset_name": "LiveCodeBench v6", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mai-thinking-1", + "https://llm-stats.com/benchmarks/livecodebench-v6", + "https://api.llm-stats.com/leaderboard/benchmarks/livecodebench-v6" + ], + "additional_details": { + "raw_benchmark_id": "livecodebench-v6", + "raw_model_id": "mai-thinking-1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LiveCodeBench is a holistic and contamination-free evaluation benchmark for large language models for code. It continuously collects new problems from programming contests (LeetCode, AtCoder, CodeForces) and evaluates four different scenarios: code generation, self-repair, code execution, and test output prediction. Problems are annotated with release dates to enable evaluation on unseen problems released after a model's training cutoff.", + "metric_id": "llm_stats.livecodebench-v6.score", + "metric_name": "LiveCodeBench v6 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "livecodebench-v6", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LiveCodeBench v6", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "53" + } + }, + "score_details": { + "score": 0.877, + "details": { + "raw_score": "0.877", + "raw_score_field": "score", + "raw_model_id": "mai-thinking-1", + "raw_benchmark_id": "livecodebench-v6", + "source_urls_json": "[\"https://llm-stats.com/models/mai-thinking-1\",\"https://llm-stats.com/benchmarks/livecodebench-v6\",\"https://api.llm-stats.com/leaderboard/benchmarks/livecodebench-v6\"]", + "raw_score_id": "livecodebench-v6::mai-thinking-1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "longbench-v2::longbench-v2-mai-thinking-1", + "evaluation_name": "llm_stats.longbench-v2", + "source_data": { + "dataset_name": "LongBench v2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mai-thinking-1", + "https://llm-stats.com/benchmarks/longbench-v2", + "https://api.llm-stats.com/leaderboard/benchmarks/longbench-v2" + ], + "additional_details": { + "raw_benchmark_id": "longbench-v2", + "raw_model_id": "mai-thinking-1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LongBench v2 is a benchmark designed to assess the ability of LLMs to handle long-context problems requiring deep understanding and reasoning across real-world multitasks. It consists of 503 challenging multiple-choice questions with contexts ranging from 8k to 2M words across six major task categories: single-document QA, multi-document QA, long in-context learning, long-dialogue history understanding, code repository understanding, and long structured data understanding.", + "metric_id": "llm_stats.longbench-v2.score", + "metric_name": "LongBench v2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "longbench-v2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LongBench v2", + "raw_categories": "[\"long_context\",\"reasoning\",\"structured_output\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.61, + "details": { + "raw_score": "0.61", + "raw_score_field": "score", + "raw_model_id": "mai-thinking-1", + "raw_benchmark_id": "longbench-v2", + "source_urls_json": "[\"https://llm-stats.com/models/mai-thinking-1\",\"https://llm-stats.com/benchmarks/longbench-v2\",\"https://api.llm-stats.com/leaderboard/benchmarks/longbench-v2\"]", + "raw_score_id": "longbench-v2::mai-thinking-1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "longfact::longfact-mai-thinking-1", + "evaluation_name": "llm_stats.longfact", + "source_data": { + "dataset_name": "LongFact", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mai-thinking-1", + "https://llm-stats.com/benchmarks/longfact", + "https://api.llm-stats.com/leaderboard/benchmarks/longfact" + ], + "additional_details": { + "raw_benchmark_id": "longfact", + "raw_model_id": "mai-thinking-1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LongFact evaluates factual precision over long-form generations containing many individual claims. Each claim is extracted and verified, and the model is scored on claim-level precision, measuring whether extended responses introduce unsupported or false statements.", + "metric_id": "llm_stats.longfact.score", + "metric_name": "LongFact score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "longfact", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LongFact", + "raw_categories": "[\"factuality\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.98, + "details": { + "raw_score": "0.98", + "raw_score_field": "score", + "raw_model_id": "mai-thinking-1", + "raw_benchmark_id": "longfact", + "source_urls_json": "[\"https://llm-stats.com/models/mai-thinking-1\",\"https://llm-stats.com/benchmarks/longfact\",\"https://api.llm-stats.com/leaderboard/benchmarks/longfact\"]", + "raw_score_id": "longfact::mai-thinking-1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "medxpertqa::medxpertqa-mai-thinking-1", + "evaluation_name": "llm_stats.medxpertqa", + "source_data": { + "dataset_name": "MedXpertQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mai-thinking-1", + "https://llm-stats.com/benchmarks/medxpertqa", + "https://api.llm-stats.com/leaderboard/benchmarks/medxpertqa" + ], + "additional_details": { + "raw_benchmark_id": "medxpertqa", + "raw_model_id": "mai-thinking-1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A comprehensive benchmark to evaluate expert-level medical knowledge and advanced reasoning, featuring 4,460 questions spanning 17 specialties and 11 body systems. Includes both text-only and multimodal subsets with expert-level exam questions incorporating diverse medical images and rich clinical information.", + "metric_id": "llm_stats.medxpertqa.score", + "metric_name": "MedXpertQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "medxpertqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MedXpertQA", + "raw_categories": "[\"multimodal\",\"reasoning\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.43, + "details": { + "raw_score": "0.43", + "raw_score_field": "score", + "raw_model_id": "mai-thinking-1", + "raw_benchmark_id": "medxpertqa", + "source_urls_json": "[\"https://llm-stats.com/models/mai-thinking-1\",\"https://llm-stats.com/benchmarks/medxpertqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/medxpertqa\"]", + "raw_score_id": "medxpertqa::mai-thinking-1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu-pro::mmlu-pro-mai-thinking-1", + "evaluation_name": "llm_stats.mmlu-pro", + "source_data": { + "dataset_name": "MMLU-Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mai-thinking-1", + "https://llm-stats.com/benchmarks/mmlu-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu-pro" + ], + "additional_details": { + "raw_benchmark_id": "mmlu-pro", + "raw_model_id": "mai-thinking-1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A more robust and challenging multi-task language understanding benchmark that extends MMLU by expanding multiple-choice options from 4 to 10, eliminating trivial questions, and focusing on reasoning-intensive tasks. Features over 12,000 curated questions across 14 domains and causes a 16-33% accuracy drop compared to original MMLU.", + "metric_id": "llm_stats.mmlu-pro.score", + "metric_name": "MMLU-Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU-Pro", + "raw_categories": "[\"language\",\"legal\",\"math\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "129" + } + }, + "score_details": { + "score": 0.85, + "details": { + "raw_score": "0.85", + "raw_score_field": "score", + "raw_model_id": "mai-thinking-1", + "raw_benchmark_id": "mmlu-pro", + "source_urls_json": "[\"https://llm-stats.com/models/mai-thinking-1\",\"https://llm-stats.com/benchmarks/mmlu-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu-pro\"]", + "raw_score_id": "mmlu-pro::mai-thinking-1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "multichallenge::multichallenge-mai-thinking-1", + "evaluation_name": "llm_stats.multichallenge", + "source_data": { + "dataset_name": "Multi-Challenge", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mai-thinking-1", + "https://llm-stats.com/benchmarks/multichallenge", + "https://api.llm-stats.com/leaderboard/benchmarks/multichallenge" + ], + "additional_details": { + "raw_benchmark_id": "multichallenge", + "raw_model_id": "mai-thinking-1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MultiChallenge is a realistic multi-turn conversation evaluation benchmark that challenges frontier LLMs across four key categories: instruction retention (maintaining instructions throughout conversations), inference memory (recalling and connecting details from previous turns), reliable versioned editing (adapting to evolving instructions during collaborative editing), and self-coherence (avoiding contradictions in responses). The benchmark evaluates models on sustained, contextually complex dialogues across diverse topics including travel planning, technical documentation, and professional communication.", + "metric_id": "llm_stats.multichallenge.score", + "metric_name": "Multi-Challenge score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "multichallenge", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Multi-Challenge", + "raw_categories": "[\"reasoning\",\"communication\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "29" + } + }, + "score_details": { + "score": 0.53, + "details": { + "raw_score": "0.53", + "raw_score_field": "score", + "raw_model_id": "mai-thinking-1", + "raw_benchmark_id": "multichallenge", + "source_urls_json": "[\"https://llm-stats.com/models/mai-thinking-1\",\"https://llm-stats.com/benchmarks/multichallenge\",\"https://api.llm-stats.com/leaderboard/benchmarks/multichallenge\"]", + "raw_score_id": "multichallenge::mai-thinking-1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "simpleqa-verified::simpleqa-verified-mai-thinking-1", + "evaluation_name": "llm_stats.simpleqa-verified", + "source_data": { + "dataset_name": "SimpleQA Verified", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mai-thinking-1", + "https://llm-stats.com/benchmarks/simpleqa-verified", + "https://api.llm-stats.com/leaderboard/benchmarks/simpleqa-verified" + ], + "additional_details": { + "raw_benchmark_id": "simpleqa-verified", + "raw_model_id": "mai-thinking-1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SimpleQA Verified is a curated, reliability-focused subset of SimpleQA that addresses label noise and redundancy in the original benchmark, measuring short-form parametric factual accuracy of large language models on fact-seeking questions with single, indisputable answers.", + "metric_id": "llm_stats.simpleqa-verified.score", + "metric_name": "SimpleQA Verified score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "simpleqa-verified", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SimpleQA Verified", + "raw_categories": "[\"reasoning\",\"factuality\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.31, + "details": { + "raw_score": "0.31", + "raw_score_field": "score", + "raw_model_id": "mai-thinking-1", + "raw_benchmark_id": "simpleqa-verified", + "source_urls_json": "[\"https://llm-stats.com/models/mai-thinking-1\",\"https://llm-stats.com/benchmarks/simpleqa-verified\",\"https://api.llm-stats.com/leaderboard/benchmarks/simpleqa-verified\"]", + "raw_score_id": "simpleqa-verified::mai-thinking-1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "truthfulqa::truthfulqa-mai-thinking-1", + "evaluation_name": "llm_stats.truthfulqa", + "source_data": { + "dataset_name": "TruthfulQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mai-thinking-1", + "https://llm-stats.com/benchmarks/truthfulqa", + "https://api.llm-stats.com/leaderboard/benchmarks/truthfulqa" + ], + "additional_details": { + "raw_benchmark_id": "truthfulqa", + "raw_model_id": "mai-thinking-1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "TruthfulQA is a benchmark to measure whether language models are truthful in generating answers to questions. It comprises 817 questions that span 38 categories, including health, law, finance and politics. The questions are crafted such that some humans would answer falsely due to a false belief or misconception, testing models' ability to avoid generating false answers learned from human texts.", + "metric_id": "llm_stats.truthfulqa.score", + "metric_name": "TruthfulQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "truthfulqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TruthfulQA", + "raw_categories": "[\"legal\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.88, + "details": { + "raw_score": "0.88", + "raw_score_field": "score", + "raw_model_id": "mai-thinking-1", + "raw_benchmark_id": "truthfulqa", + "source_urls_json": "[\"https://llm-stats.com/models/mai-thinking-1\",\"https://llm-stats.com/benchmarks/truthfulqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/truthfulqa\"]", + "raw_score_id": "truthfulqa::mai-thinking-1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/microsoft/phi-3.5-mini-instruct/e05442cd-3345-44a1-8b1c-170befaf712a.json b/data/llm_stats/microsoft/phi-3.5-mini-instruct/e05442cd-3345-44a1-8b1c-170befaf712a.json new file mode 100644 index 0000000000000000000000000000000000000000..acf674315709fd345f1d4dcfc17b109d916ef694 --- /dev/null +++ b/data/llm_stats/microsoft/phi-3.5-mini-instruct/e05442cd-3345-44a1-8b1c-170befaf712a.json @@ -0,0 +1,1087 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/microsoft_phi-3.5-mini-instruct/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Phi-3.5-mini-instruct", + "id": "microsoft/phi-3.5-mini-instruct", + "developer": "microsoft", + "additional_details": { + "raw_id": "phi-3.5-mini-instruct", + "raw_name": "Phi-3.5-mini-instruct", + "raw_model_id": "phi-3.5-mini-instruct", + "raw_model_name": "Phi-3.5-mini-instruct", + "raw_organization_id": "microsoft", + "raw_organization_name": "Microsoft", + "raw_release_date": "2024-08-23", + "raw_announcement_date": "2024-08-23", + "raw_multimodal": "false", + "raw_param_count": "3800000000", + "raw_provider_slug": "microsoft", + "raw_provider_name": "Microsoft" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "arc-c::arc-c-phi-3.5-mini-instruct", + "evaluation_name": "llm_stats.arc-c", + "source_data": { + "dataset_name": "ARC-C", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-mini-instruct", + "https://llm-stats.com/benchmarks/arc-c", + "https://api.llm-stats.com/leaderboard/benchmarks/arc-c" + ], + "additional_details": { + "raw_benchmark_id": "arc-c", + "raw_model_id": "phi-3.5-mini-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The AI2 Reasoning Challenge (ARC) Challenge Set is a multiple-choice question-answering benchmark containing grade-school level science questions that require advanced reasoning capabilities. ARC-C specifically contains questions that were answered incorrectly by both retrieval-based and word co-occurrence algorithms, making it a particularly challenging subset designed to test commonsense reasoning abilities in AI systems.", + "metric_id": "llm_stats.arc-c.score", + "metric_name": "ARC-C score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arc-c", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ARC-C", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "34" + } + }, + "score_details": { + "score": 0.846, + "details": { + "raw_score": "0.846", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-mini-instruct", + "raw_benchmark_id": "arc-c", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-mini-instruct\",\"https://llm-stats.com/benchmarks/arc-c\",\"https://api.llm-stats.com/leaderboard/benchmarks/arc-c\"]", + "raw_score_id": "arc-c::phi-3.5-mini-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "big-bench-hard::big-bench-hard-phi-3.5-mini-instruct", + "evaluation_name": "llm_stats.big-bench-hard", + "source_data": { + "dataset_name": "BIG-Bench Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-mini-instruct", + "https://llm-stats.com/benchmarks/big-bench-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard" + ], + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_model_id": "phi-3.5-mini-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIG-Bench Hard (BBH) is a subset of 23 challenging BIG-Bench tasks selected because prior language model evaluations did not outperform average human-rater performance. The benchmark contains 6,511 evaluation examples testing various forms of multi-step reasoning including arithmetic, logical reasoning (Boolean expressions, logical deduction), geometric reasoning, temporal reasoning, and language understanding. Tasks require capabilities such as causal judgment, object counting, navigation, pattern recognition, and complex problem solving.", + "metric_id": "llm_stats.big-bench-hard.score", + "metric_name": "BIG-Bench Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench Hard", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "21" + } + }, + "score_details": { + "score": 0.69, + "details": { + "raw_score": "0.69", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-mini-instruct", + "raw_benchmark_id": "big-bench-hard", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-mini-instruct\",\"https://llm-stats.com/benchmarks/big-bench-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard\"]", + "raw_score_id": "big-bench-hard::phi-3.5-mini-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "boolq::boolq-phi-3.5-mini-instruct", + "evaluation_name": "llm_stats.boolq", + "source_data": { + "dataset_name": "BoolQ", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-mini-instruct", + "https://llm-stats.com/benchmarks/boolq", + "https://api.llm-stats.com/leaderboard/benchmarks/boolq" + ], + "additional_details": { + "raw_benchmark_id": "boolq", + "raw_model_id": "phi-3.5-mini-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BoolQ is a reading comprehension dataset for yes/no questions containing 15,942 naturally occurring examples. Each example consists of a question, passage, and boolean answer, where questions are generated in unprompted and unconstrained settings. The dataset challenges models with complex, non-factoid information requiring entailment-like inference to solve.", + "metric_id": "llm_stats.boolq.score", + "metric_name": "BoolQ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "boolq", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BoolQ", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.78, + "details": { + "raw_score": "0.78", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-mini-instruct", + "raw_benchmark_id": "boolq", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-mini-instruct\",\"https://llm-stats.com/benchmarks/boolq\",\"https://api.llm-stats.com/leaderboard/benchmarks/boolq\"]", + "raw_score_id": "boolq::phi-3.5-mini-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "govreport::govreport-phi-3.5-mini-instruct", + "evaluation_name": "llm_stats.govreport", + "source_data": { + "dataset_name": "GovReport", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-mini-instruct", + "https://llm-stats.com/benchmarks/govreport", + "https://api.llm-stats.com/leaderboard/benchmarks/govreport" + ], + "additional_details": { + "raw_benchmark_id": "govreport", + "raw_model_id": "phi-3.5-mini-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A long document summarization dataset consisting of reports from government research agencies including Congressional Research Service and U.S. Government Accountability Office, with significantly longer documents and summaries than other datasets.", + "metric_id": "llm_stats.govreport.score", + "metric_name": "GovReport score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "govreport", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GovReport", + "raw_categories": "[\"long_context\",\"summarization\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.259, + "details": { + "raw_score": "0.259", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-mini-instruct", + "raw_benchmark_id": "govreport", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-mini-instruct\",\"https://llm-stats.com/benchmarks/govreport\",\"https://api.llm-stats.com/leaderboard/benchmarks/govreport\"]", + "raw_score_id": "govreport::phi-3.5-mini-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mega-mlqa::mega-mlqa-phi-3.5-mini-instruct", + "evaluation_name": "llm_stats.mega-mlqa", + "source_data": { + "dataset_name": "MEGA MLQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-mini-instruct", + "https://llm-stats.com/benchmarks/mega-mlqa", + "https://api.llm-stats.com/leaderboard/benchmarks/mega-mlqa" + ], + "additional_details": { + "raw_benchmark_id": "mega-mlqa", + "raw_model_id": "phi-3.5-mini-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MLQA as part of the MEGA (Multilingual Evaluation of Generative AI) benchmark suite. A multi-way aligned extractive QA evaluation benchmark for cross-lingual question answering across 7 languages (English, Arabic, German, Spanish, Hindi, Vietnamese, and Simplified Chinese) with over 12K QA instances in English and 5K in each other language.", + "metric_id": "llm_stats.mega-mlqa.score", + "metric_name": "MEGA MLQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mega-mlqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MEGA MLQA", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.617, + "details": { + "raw_score": "0.617", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-mini-instruct", + "raw_benchmark_id": "mega-mlqa", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-mini-instruct\",\"https://llm-stats.com/benchmarks/mega-mlqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/mega-mlqa\"]", + "raw_score_id": "mega-mlqa::phi-3.5-mini-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mega-tydi-qa::mega-tydi-qa-phi-3.5-mini-instruct", + "evaluation_name": "llm_stats.mega-tydi-qa", + "source_data": { + "dataset_name": "MEGA TyDi QA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-mini-instruct", + "https://llm-stats.com/benchmarks/mega-tydi-qa", + "https://api.llm-stats.com/leaderboard/benchmarks/mega-tydi-qa" + ], + "additional_details": { + "raw_benchmark_id": "mega-tydi-qa", + "raw_model_id": "phi-3.5-mini-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "TyDi QA as part of the MEGA benchmark suite. A question answering dataset covering 11 typologically diverse languages (Arabic, Bengali, English, Finnish, Indonesian, Japanese, Korean, Russian, Swahili, Telugu, and Thai) with 204K question-answer pairs. Features realistic information-seeking questions written by people who want to know the answer but don't know it yet.", + "metric_id": "llm_stats.mega-tydi-qa.score", + "metric_name": "MEGA TyDi QA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mega-tydi-qa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MEGA TyDi QA", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.622, + "details": { + "raw_score": "0.622", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-mini-instruct", + "raw_benchmark_id": "mega-tydi-qa", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-mini-instruct\",\"https://llm-stats.com/benchmarks/mega-tydi-qa\",\"https://api.llm-stats.com/leaderboard/benchmarks/mega-tydi-qa\"]", + "raw_score_id": "mega-tydi-qa::phi-3.5-mini-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mega-udpos::mega-udpos-phi-3.5-mini-instruct", + "evaluation_name": "llm_stats.mega-udpos", + "source_data": { + "dataset_name": "MEGA UDPOS", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-mini-instruct", + "https://llm-stats.com/benchmarks/mega-udpos", + "https://api.llm-stats.com/leaderboard/benchmarks/mega-udpos" + ], + "additional_details": { + "raw_benchmark_id": "mega-udpos", + "raw_model_id": "phi-3.5-mini-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Universal Dependencies POS tagging as part of the MEGA benchmark suite. A multilingual part-of-speech tagging dataset based on Universal Dependencies treebanks, utilizing the universal POS tag set of 17 tags across 38 diverse languages from different language families. Used for evaluating multilingual POS tagging systems.", + "metric_id": "llm_stats.mega-udpos.score", + "metric_name": "MEGA UDPOS score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mega-udpos", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MEGA UDPOS", + "raw_categories": "[\"language\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.465, + "details": { + "raw_score": "0.465", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-mini-instruct", + "raw_benchmark_id": "mega-udpos", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-mini-instruct\",\"https://llm-stats.com/benchmarks/mega-udpos\",\"https://api.llm-stats.com/leaderboard/benchmarks/mega-udpos\"]", + "raw_score_id": "mega-udpos::phi-3.5-mini-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mega-xcopa::mega-xcopa-phi-3.5-mini-instruct", + "evaluation_name": "llm_stats.mega-xcopa", + "source_data": { + "dataset_name": "MEGA XCOPA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-mini-instruct", + "https://llm-stats.com/benchmarks/mega-xcopa", + "https://api.llm-stats.com/leaderboard/benchmarks/mega-xcopa" + ], + "additional_details": { + "raw_benchmark_id": "mega-xcopa", + "raw_model_id": "phi-3.5-mini-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "XCOPA (Cross-lingual Choice of Plausible Alternatives) as part of the MEGA benchmark suite. A typologically diverse multilingual dataset for causal commonsense reasoning in 11 languages, including resource-poor languages like Eastern Apurímac Quechua and Haitian Creole. Requires models to select which choice is the effect or cause of a given premise.", + "metric_id": "llm_stats.mega-xcopa.score", + "metric_name": "MEGA XCOPA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mega-xcopa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MEGA XCOPA", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.631, + "details": { + "raw_score": "0.631", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-mini-instruct", + "raw_benchmark_id": "mega-xcopa", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-mini-instruct\",\"https://llm-stats.com/benchmarks/mega-xcopa\",\"https://api.llm-stats.com/leaderboard/benchmarks/mega-xcopa\"]", + "raw_score_id": "mega-xcopa::phi-3.5-mini-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mega-xstorycloze::mega-xstorycloze-phi-3.5-mini-instruct", + "evaluation_name": "llm_stats.mega-xstorycloze", + "source_data": { + "dataset_name": "MEGA XStoryCloze", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-mini-instruct", + "https://llm-stats.com/benchmarks/mega-xstorycloze", + "https://api.llm-stats.com/leaderboard/benchmarks/mega-xstorycloze" + ], + "additional_details": { + "raw_benchmark_id": "mega-xstorycloze", + "raw_model_id": "phi-3.5-mini-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "XStoryCloze as part of the MEGA benchmark suite. A cross-lingual story completion task that consists of professionally translated versions of the English StoryCloze dataset to 10 non-English languages. Requires models to predict the correct ending for a given four-sentence story, evaluating commonsense reasoning and narrative understanding.", + "metric_id": "llm_stats.mega-xstorycloze.score", + "metric_name": "MEGA XStoryCloze score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mega-xstorycloze", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MEGA XStoryCloze", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.735, + "details": { + "raw_score": "0.735", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-mini-instruct", + "raw_benchmark_id": "mega-xstorycloze", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-mini-instruct\",\"https://llm-stats.com/benchmarks/mega-xstorycloze\",\"https://api.llm-stats.com/leaderboard/benchmarks/mega-xstorycloze\"]", + "raw_score_id": "mega-xstorycloze::phi-3.5-mini-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "openbookqa::openbookqa-phi-3.5-mini-instruct", + "evaluation_name": "llm_stats.openbookqa", + "source_data": { + "dataset_name": "OpenBookQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-mini-instruct", + "https://llm-stats.com/benchmarks/openbookqa", + "https://api.llm-stats.com/leaderboard/benchmarks/openbookqa" + ], + "additional_details": { + "raw_benchmark_id": "openbookqa", + "raw_model_id": "phi-3.5-mini-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OpenBookQA is a question-answering dataset modeled after open book exams for assessing human understanding. It contains 5,957 multiple-choice elementary-level science questions that probe understanding of 1,326 core science facts and their application to novel situations, requiring combination of open book facts with broad common knowledge through multi-hop reasoning.", + "metric_id": "llm_stats.openbookqa.score", + "metric_name": "OpenBookQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "openbookqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OpenBookQA", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.792, + "details": { + "raw_score": "0.792", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-mini-instruct", + "raw_benchmark_id": "openbookqa", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-mini-instruct\",\"https://llm-stats.com/benchmarks/openbookqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/openbookqa\"]", + "raw_score_id": "openbookqa::phi-3.5-mini-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "piqa::piqa-phi-3.5-mini-instruct", + "evaluation_name": "llm_stats.piqa", + "source_data": { + "dataset_name": "PIQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-mini-instruct", + "https://llm-stats.com/benchmarks/piqa", + "https://api.llm-stats.com/leaderboard/benchmarks/piqa" + ], + "additional_details": { + "raw_benchmark_id": "piqa", + "raw_model_id": "phi-3.5-mini-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "PIQA (Physical Interaction: Question Answering) is a benchmark dataset for physical commonsense reasoning in natural language. It tests AI systems' ability to answer questions requiring physical world knowledge through multiple choice questions with everyday situations, focusing on atypical solutions inspired by instructables.com. The dataset contains 21,000 multiple choice questions where models must choose the most appropriate solution for physical interactions.", + "metric_id": "llm_stats.piqa.score", + "metric_name": "PIQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "piqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PIQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.81, + "details": { + "raw_score": "0.81", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-mini-instruct", + "raw_benchmark_id": "piqa", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-mini-instruct\",\"https://llm-stats.com/benchmarks/piqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/piqa\"]", + "raw_score_id": "piqa::phi-3.5-mini-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "qasper::qasper-phi-3.5-mini-instruct", + "evaluation_name": "llm_stats.qasper", + "source_data": { + "dataset_name": "Qasper", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-mini-instruct", + "https://llm-stats.com/benchmarks/qasper", + "https://api.llm-stats.com/leaderboard/benchmarks/qasper" + ], + "additional_details": { + "raw_benchmark_id": "qasper", + "raw_model_id": "phi-3.5-mini-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "QASPER is a dataset of 5,049 information-seeking questions and answers anchored in 1,585 NLP research papers. Questions are written by NLP practitioners who read only titles and abstracts, while answers require understanding the full paper text and provide supporting evidence. The dataset challenges models with complex reasoning across document sections for academic document question answering. Each question seeks information present in the full text and is answered by a separate set of NLP practitioners who also provide supporting evidence to answers.", + "metric_id": "llm_stats.qasper.score", + "metric_name": "Qasper score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "qasper", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Qasper", + "raw_categories": "[\"long_context\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.419, + "details": { + "raw_score": "0.419", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-mini-instruct", + "raw_benchmark_id": "qasper", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-mini-instruct\",\"https://llm-stats.com/benchmarks/qasper\",\"https://api.llm-stats.com/leaderboard/benchmarks/qasper\"]", + "raw_score_id": "qasper::phi-3.5-mini-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "qmsum::qmsum-phi-3.5-mini-instruct", + "evaluation_name": "llm_stats.qmsum", + "source_data": { + "dataset_name": "QMSum", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-mini-instruct", + "https://llm-stats.com/benchmarks/qmsum", + "https://api.llm-stats.com/leaderboard/benchmarks/qmsum" + ], + "additional_details": { + "raw_benchmark_id": "qmsum", + "raw_model_id": "phi-3.5-mini-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "QMSum is a benchmark for query-based multi-domain meeting summarization consisting of 1,808 query-summary pairs over 232 meetings across academic, product, and committee domains. The dataset enables models to select and summarize relevant spans of meetings in response to specific queries. Published at NAACL 2021, QMSum presents significant challenges in long meeting summarization where models must identify and summarize relevant content based on user queries.", + "metric_id": "llm_stats.qmsum.score", + "metric_name": "QMSum score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "qmsum", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "QMSum", + "raw_categories": "[\"long_context\",\"summarization\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.213, + "details": { + "raw_score": "0.213", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-mini-instruct", + "raw_benchmark_id": "qmsum", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-mini-instruct\",\"https://llm-stats.com/benchmarks/qmsum\",\"https://api.llm-stats.com/leaderboard/benchmarks/qmsum\"]", + "raw_score_id": "qmsum::phi-3.5-mini-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "repoqa::repoqa-phi-3.5-mini-instruct", + "evaluation_name": "llm_stats.repoqa", + "source_data": { + "dataset_name": "RepoQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-mini-instruct", + "https://llm-stats.com/benchmarks/repoqa", + "https://api.llm-stats.com/leaderboard/benchmarks/repoqa" + ], + "additional_details": { + "raw_benchmark_id": "repoqa", + "raw_model_id": "phi-3.5-mini-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "RepoQA is a benchmark for evaluating long-context code understanding capabilities of Large Language Models through the Searching Needle Function (SNF) task, where LLMs must locate specific functions in code repositories using natural language descriptions. The benchmark contains 500 code search tasks spanning 50 repositories across 5 modern programming languages (Python, Java, TypeScript, C++, and Rust), tested on 26 general and code-specific LLMs to assess their ability to comprehend and navigate code repositories.", + "metric_id": "llm_stats.repoqa.score", + "metric_name": "RepoQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "repoqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "RepoQA", + "raw_categories": "[\"long_context\",\"reasoning\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.77, + "details": { + "raw_score": "0.77", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-mini-instruct", + "raw_benchmark_id": "repoqa", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-mini-instruct\",\"https://llm-stats.com/benchmarks/repoqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/repoqa\"]", + "raw_score_id": "repoqa::phi-3.5-mini-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ruler::ruler-phi-3.5-mini-instruct", + "evaluation_name": "llm_stats.ruler", + "source_data": { + "dataset_name": "RULER", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-mini-instruct", + "https://llm-stats.com/benchmarks/ruler", + "https://api.llm-stats.com/leaderboard/benchmarks/ruler" + ], + "additional_details": { + "raw_benchmark_id": "ruler", + "raw_model_id": "phi-3.5-mini-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "RULER v1 is a synthetic long-context benchmark for measuring how model quality degrades as input length increases. This packaging follows the public standalone NVIDIA RULER implementation with 13 official tasks spanning retrieval, multi-hop tracing, aggregation, and QA.", + "metric_id": "llm_stats.ruler.score", + "metric_name": "RULER score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ruler", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "RULER", + "raw_categories": "[\"long_context\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.841, + "details": { + "raw_score": "0.841", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-mini-instruct", + "raw_benchmark_id": "ruler", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-mini-instruct\",\"https://llm-stats.com/benchmarks/ruler\",\"https://api.llm-stats.com/leaderboard/benchmarks/ruler\"]", + "raw_score_id": "ruler::phi-3.5-mini-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "social-iqa::social-iqa-phi-3.5-mini-instruct", + "evaluation_name": "llm_stats.social-iqa", + "source_data": { + "dataset_name": "Social IQa", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-mini-instruct", + "https://llm-stats.com/benchmarks/social-iqa", + "https://api.llm-stats.com/leaderboard/benchmarks/social-iqa" + ], + "additional_details": { + "raw_benchmark_id": "social-iqa", + "raw_model_id": "phi-3.5-mini-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The first large-scale benchmark for commonsense reasoning about social situations. Contains 38,000 multiple choice questions probing emotional and social intelligence in everyday situations, testing commonsense understanding of social interactions and theory of mind reasoning about the implied emotions and behavior of others.", + "metric_id": "llm_stats.social-iqa.score", + "metric_name": "Social IQa score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "social-iqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Social IQa", + "raw_categories": "[\"psychology\",\"reasoning\",\"creativity\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.747, + "details": { + "raw_score": "0.747", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-mini-instruct", + "raw_benchmark_id": "social-iqa", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-mini-instruct\",\"https://llm-stats.com/benchmarks/social-iqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/social-iqa\"]", + "raw_score_id": "social-iqa::phi-3.5-mini-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "squality::squality-phi-3.5-mini-instruct", + "evaluation_name": "llm_stats.squality", + "source_data": { + "dataset_name": "SQuALITY", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-mini-instruct", + "https://llm-stats.com/benchmarks/squality", + "https://api.llm-stats.com/leaderboard/benchmarks/squality" + ], + "additional_details": { + "raw_benchmark_id": "squality", + "raw_model_id": "phi-3.5-mini-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SQuALITY (Summarization-format QUestion Answering with Long Input Texts, Yes!) is a long-document summarization dataset built by hiring highly-qualified contractors to read public-domain short stories (3000-6000 words) and write original summaries from scratch. Each document has five summaries: one overview and four question-focused summaries. Designed to address limitations in existing summarization datasets by providing high-quality, faithful summaries.", + "metric_id": "llm_stats.squality.score", + "metric_name": "SQuALITY score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "squality", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SQuALITY", + "raw_categories": "[\"language\",\"long_context\",\"summarization\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.243, + "details": { + "raw_score": "0.243", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-mini-instruct", + "raw_benchmark_id": "squality", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-mini-instruct\",\"https://llm-stats.com/benchmarks/squality\",\"https://api.llm-stats.com/leaderboard/benchmarks/squality\"]", + "raw_score_id": "squality::phi-3.5-mini-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "summscreenfd::summscreenfd-phi-3.5-mini-instruct", + "evaluation_name": "llm_stats.summscreenfd", + "source_data": { + "dataset_name": "SummScreenFD", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-mini-instruct", + "https://llm-stats.com/benchmarks/summscreenfd", + "https://api.llm-stats.com/leaderboard/benchmarks/summscreenfd" + ], + "additional_details": { + "raw_benchmark_id": "summscreenfd", + "raw_model_id": "phi-3.5-mini-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SummScreenFD is the ForeverDreaming subset of the SummScreen dataset for abstractive screenplay summarization, comprising pairs of TV series transcripts and human-written recaps from 88 different shows. The dataset provides a challenging testbed for abstractive summarization where plot details are often expressed indirectly in character dialogues and scattered across the entirety of the transcript, requiring models to find and integrate these details to form succinct plot descriptions.", + "metric_id": "llm_stats.summscreenfd.score", + "metric_name": "SummScreenFD score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "summscreenfd", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SummScreenFD", + "raw_categories": "[\"long_context\",\"summarization\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.16, + "details": { + "raw_score": "0.16", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-mini-instruct", + "raw_benchmark_id": "summscreenfd", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-mini-instruct\",\"https://llm-stats.com/benchmarks/summscreenfd\",\"https://api.llm-stats.com/leaderboard/benchmarks/summscreenfd\"]", + "raw_score_id": "summscreenfd::phi-3.5-mini-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "truthfulqa::truthfulqa-phi-3.5-mini-instruct", + "evaluation_name": "llm_stats.truthfulqa", + "source_data": { + "dataset_name": "TruthfulQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-mini-instruct", + "https://llm-stats.com/benchmarks/truthfulqa", + "https://api.llm-stats.com/leaderboard/benchmarks/truthfulqa" + ], + "additional_details": { + "raw_benchmark_id": "truthfulqa", + "raw_model_id": "phi-3.5-mini-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "TruthfulQA is a benchmark to measure whether language models are truthful in generating answers to questions. It comprises 817 questions that span 38 categories, including health, law, finance and politics. The questions are crafted such that some humans would answer falsely due to a false belief or misconception, testing models' ability to avoid generating false answers learned from human texts.", + "metric_id": "llm_stats.truthfulqa.score", + "metric_name": "TruthfulQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "truthfulqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TruthfulQA", + "raw_categories": "[\"legal\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.64, + "details": { + "raw_score": "0.64", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-mini-instruct", + "raw_benchmark_id": "truthfulqa", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-mini-instruct\",\"https://llm-stats.com/benchmarks/truthfulqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/truthfulqa\"]", + "raw_score_id": "truthfulqa::phi-3.5-mini-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "winogrande::winogrande-phi-3.5-mini-instruct", + "evaluation_name": "llm_stats.winogrande", + "source_data": { + "dataset_name": "Winogrande", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-mini-instruct", + "https://llm-stats.com/benchmarks/winogrande", + "https://api.llm-stats.com/leaderboard/benchmarks/winogrande" + ], + "additional_details": { + "raw_benchmark_id": "winogrande", + "raw_model_id": "phi-3.5-mini-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WinoGrande: An Adversarial Winograd Schema Challenge at Scale. A large-scale dataset of 44,000 pronoun resolution problems designed to test machine commonsense reasoning. Uses adversarial filtering to reduce spurious biases and provides a more robust evaluation of whether AI systems truly understand commonsense or exploit statistical shortcuts. Current best AI methods achieve 59.4-79.1% accuracy, significantly below human performance of 94.0%.", + "metric_id": "llm_stats.winogrande.score", + "metric_name": "Winogrande score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "winogrande", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Winogrande", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "22" + } + }, + "score_details": { + "score": 0.685, + "details": { + "raw_score": "0.685", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-mini-instruct", + "raw_benchmark_id": "winogrande", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-mini-instruct\",\"https://llm-stats.com/benchmarks/winogrande\",\"https://api.llm-stats.com/leaderboard/benchmarks/winogrande\"]", + "raw_score_id": "winogrande::phi-3.5-mini-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/microsoft/phi-3.5-moe-instruct/cf75d602-ceed-4dde-b67c-4c1aa62e905d.json b/data/llm_stats/microsoft/phi-3.5-moe-instruct/cf75d602-ceed-4dde-b67c-4c1aa62e905d.json new file mode 100644 index 0000000000000000000000000000000000000000..7d621355950c4ad98e2a55b2fe1326bab0694bc0 --- /dev/null +++ b/data/llm_stats/microsoft/phi-3.5-moe-instruct/cf75d602-ceed-4dde-b67c-4c1aa62e905d.json @@ -0,0 +1,1191 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/microsoft_phi-3.5-moe-instruct/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Phi-3.5-MoE-instruct", + "id": "microsoft/phi-3.5-moe-instruct", + "developer": "microsoft", + "additional_details": { + "raw_id": "phi-3.5-moe-instruct", + "raw_name": "Phi-3.5-MoE-instruct", + "raw_model_id": "phi-3.5-moe-instruct", + "raw_model_name": "Phi-3.5-MoE-instruct", + "raw_organization_id": "microsoft", + "raw_organization_name": "Microsoft", + "raw_release_date": "2024-08-23", + "raw_announcement_date": "2024-08-23", + "raw_multimodal": "false", + "raw_param_count": "60000000000", + "raw_provider_slug": "microsoft", + "raw_provider_name": "Microsoft" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "arc-c::arc-c-phi-3.5-moe-instruct", + "evaluation_name": "llm_stats.arc-c", + "source_data": { + "dataset_name": "ARC-C", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-moe-instruct", + "https://llm-stats.com/benchmarks/arc-c", + "https://api.llm-stats.com/leaderboard/benchmarks/arc-c" + ], + "additional_details": { + "raw_benchmark_id": "arc-c", + "raw_model_id": "phi-3.5-moe-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The AI2 Reasoning Challenge (ARC) Challenge Set is a multiple-choice question-answering benchmark containing grade-school level science questions that require advanced reasoning capabilities. ARC-C specifically contains questions that were answered incorrectly by both retrieval-based and word co-occurrence algorithms, making it a particularly challenging subset designed to test commonsense reasoning abilities in AI systems.", + "metric_id": "llm_stats.arc-c.score", + "metric_name": "ARC-C score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arc-c", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ARC-C", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "34" + } + }, + "score_details": { + "score": 0.91, + "details": { + "raw_score": "0.91", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-moe-instruct", + "raw_benchmark_id": "arc-c", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-moe-instruct\",\"https://llm-stats.com/benchmarks/arc-c\",\"https://api.llm-stats.com/leaderboard/benchmarks/arc-c\"]", + "raw_score_id": "arc-c::phi-3.5-moe-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "big-bench-hard::big-bench-hard-phi-3.5-moe-instruct", + "evaluation_name": "llm_stats.big-bench-hard", + "source_data": { + "dataset_name": "BIG-Bench Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-moe-instruct", + "https://llm-stats.com/benchmarks/big-bench-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard" + ], + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_model_id": "phi-3.5-moe-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIG-Bench Hard (BBH) is a subset of 23 challenging BIG-Bench tasks selected because prior language model evaluations did not outperform average human-rater performance. The benchmark contains 6,511 evaluation examples testing various forms of multi-step reasoning including arithmetic, logical reasoning (Boolean expressions, logical deduction), geometric reasoning, temporal reasoning, and language understanding. Tasks require capabilities such as causal judgment, object counting, navigation, pattern recognition, and complex problem solving.", + "metric_id": "llm_stats.big-bench-hard.score", + "metric_name": "BIG-Bench Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench Hard", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "21" + } + }, + "score_details": { + "score": 0.791, + "details": { + "raw_score": "0.791", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-moe-instruct", + "raw_benchmark_id": "big-bench-hard", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-moe-instruct\",\"https://llm-stats.com/benchmarks/big-bench-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard\"]", + "raw_score_id": "big-bench-hard::phi-3.5-moe-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "boolq::boolq-phi-3.5-moe-instruct", + "evaluation_name": "llm_stats.boolq", + "source_data": { + "dataset_name": "BoolQ", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-moe-instruct", + "https://llm-stats.com/benchmarks/boolq", + "https://api.llm-stats.com/leaderboard/benchmarks/boolq" + ], + "additional_details": { + "raw_benchmark_id": "boolq", + "raw_model_id": "phi-3.5-moe-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BoolQ is a reading comprehension dataset for yes/no questions containing 15,942 naturally occurring examples. Each example consists of a question, passage, and boolean answer, where questions are generated in unprompted and unconstrained settings. The dataset challenges models with complex, non-factoid information requiring entailment-like inference to solve.", + "metric_id": "llm_stats.boolq.score", + "metric_name": "BoolQ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "boolq", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BoolQ", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.846, + "details": { + "raw_score": "0.846", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-moe-instruct", + "raw_benchmark_id": "boolq", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-moe-instruct\",\"https://llm-stats.com/benchmarks/boolq\",\"https://api.llm-stats.com/leaderboard/benchmarks/boolq\"]", + "raw_score_id": "boolq::phi-3.5-moe-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "govreport::govreport-phi-3.5-moe-instruct", + "evaluation_name": "llm_stats.govreport", + "source_data": { + "dataset_name": "GovReport", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-moe-instruct", + "https://llm-stats.com/benchmarks/govreport", + "https://api.llm-stats.com/leaderboard/benchmarks/govreport" + ], + "additional_details": { + "raw_benchmark_id": "govreport", + "raw_model_id": "phi-3.5-moe-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A long document summarization dataset consisting of reports from government research agencies including Congressional Research Service and U.S. Government Accountability Office, with significantly longer documents and summaries than other datasets.", + "metric_id": "llm_stats.govreport.score", + "metric_name": "GovReport score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "govreport", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GovReport", + "raw_categories": "[\"long_context\",\"summarization\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.264, + "details": { + "raw_score": "0.264", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-moe-instruct", + "raw_benchmark_id": "govreport", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-moe-instruct\",\"https://llm-stats.com/benchmarks/govreport\",\"https://api.llm-stats.com/leaderboard/benchmarks/govreport\"]", + "raw_score_id": "govreport::phi-3.5-moe-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hellaswag::hellaswag-phi-3.5-moe-instruct", + "evaluation_name": "llm_stats.hellaswag", + "source_data": { + "dataset_name": "HellaSwag", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-moe-instruct", + "https://llm-stats.com/benchmarks/hellaswag", + "https://api.llm-stats.com/leaderboard/benchmarks/hellaswag" + ], + "additional_details": { + "raw_benchmark_id": "hellaswag", + "raw_model_id": "phi-3.5-moe-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A challenging commonsense natural language inference dataset that uses Adversarial Filtering to create questions trivial for humans (>95% accuracy) but difficult for state-of-the-art models, requiring completion of sentence endings based on physical situations and everyday activities", + "metric_id": "llm_stats.hellaswag.score", + "metric_name": "HellaSwag score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hellaswag", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HellaSwag", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.838, + "details": { + "raw_score": "0.838", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-moe-instruct", + "raw_benchmark_id": "hellaswag", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-moe-instruct\",\"https://llm-stats.com/benchmarks/hellaswag\",\"https://api.llm-stats.com/leaderboard/benchmarks/hellaswag\"]", + "raw_score_id": "hellaswag::phi-3.5-moe-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mbpp::mbpp-phi-3.5-moe-instruct", + "evaluation_name": "llm_stats.mbpp", + "source_data": { + "dataset_name": "MBPP", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-moe-instruct", + "https://llm-stats.com/benchmarks/mbpp", + "https://api.llm-stats.com/leaderboard/benchmarks/mbpp" + ], + "additional_details": { + "raw_benchmark_id": "mbpp", + "raw_model_id": "phi-3.5-moe-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MBPP (Mostly Basic Python Problems) is a benchmark of 974 crowd-sourced Python programming problems designed to be solvable by entry-level programmers. Each problem consists of a task description, code solution, and 3 automated test cases covering programming fundamentals and standard library functionality.", + "metric_id": "llm_stats.mbpp.score", + "metric_name": "MBPP score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mbpp", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MBPP", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "33" + } + }, + "score_details": { + "score": 0.808, + "details": { + "raw_score": "0.808", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-moe-instruct", + "raw_benchmark_id": "mbpp", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-moe-instruct\",\"https://llm-stats.com/benchmarks/mbpp\",\"https://api.llm-stats.com/leaderboard/benchmarks/mbpp\"]", + "raw_score_id": "mbpp::phi-3.5-moe-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mega-mlqa::mega-mlqa-phi-3.5-moe-instruct", + "evaluation_name": "llm_stats.mega-mlqa", + "source_data": { + "dataset_name": "MEGA MLQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-moe-instruct", + "https://llm-stats.com/benchmarks/mega-mlqa", + "https://api.llm-stats.com/leaderboard/benchmarks/mega-mlqa" + ], + "additional_details": { + "raw_benchmark_id": "mega-mlqa", + "raw_model_id": "phi-3.5-moe-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MLQA as part of the MEGA (Multilingual Evaluation of Generative AI) benchmark suite. A multi-way aligned extractive QA evaluation benchmark for cross-lingual question answering across 7 languages (English, Arabic, German, Spanish, Hindi, Vietnamese, and Simplified Chinese) with over 12K QA instances in English and 5K in each other language.", + "metric_id": "llm_stats.mega-mlqa.score", + "metric_name": "MEGA MLQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mega-mlqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MEGA MLQA", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.653, + "details": { + "raw_score": "0.653", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-moe-instruct", + "raw_benchmark_id": "mega-mlqa", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-moe-instruct\",\"https://llm-stats.com/benchmarks/mega-mlqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/mega-mlqa\"]", + "raw_score_id": "mega-mlqa::phi-3.5-moe-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mega-tydi-qa::mega-tydi-qa-phi-3.5-moe-instruct", + "evaluation_name": "llm_stats.mega-tydi-qa", + "source_data": { + "dataset_name": "MEGA TyDi QA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-moe-instruct", + "https://llm-stats.com/benchmarks/mega-tydi-qa", + "https://api.llm-stats.com/leaderboard/benchmarks/mega-tydi-qa" + ], + "additional_details": { + "raw_benchmark_id": "mega-tydi-qa", + "raw_model_id": "phi-3.5-moe-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "TyDi QA as part of the MEGA benchmark suite. A question answering dataset covering 11 typologically diverse languages (Arabic, Bengali, English, Finnish, Indonesian, Japanese, Korean, Russian, Swahili, Telugu, and Thai) with 204K question-answer pairs. Features realistic information-seeking questions written by people who want to know the answer but don't know it yet.", + "metric_id": "llm_stats.mega-tydi-qa.score", + "metric_name": "MEGA TyDi QA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mega-tydi-qa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MEGA TyDi QA", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.671, + "details": { + "raw_score": "0.671", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-moe-instruct", + "raw_benchmark_id": "mega-tydi-qa", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-moe-instruct\",\"https://llm-stats.com/benchmarks/mega-tydi-qa\",\"https://api.llm-stats.com/leaderboard/benchmarks/mega-tydi-qa\"]", + "raw_score_id": "mega-tydi-qa::phi-3.5-moe-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mega-udpos::mega-udpos-phi-3.5-moe-instruct", + "evaluation_name": "llm_stats.mega-udpos", + "source_data": { + "dataset_name": "MEGA UDPOS", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-moe-instruct", + "https://llm-stats.com/benchmarks/mega-udpos", + "https://api.llm-stats.com/leaderboard/benchmarks/mega-udpos" + ], + "additional_details": { + "raw_benchmark_id": "mega-udpos", + "raw_model_id": "phi-3.5-moe-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Universal Dependencies POS tagging as part of the MEGA benchmark suite. A multilingual part-of-speech tagging dataset based on Universal Dependencies treebanks, utilizing the universal POS tag set of 17 tags across 38 diverse languages from different language families. Used for evaluating multilingual POS tagging systems.", + "metric_id": "llm_stats.mega-udpos.score", + "metric_name": "MEGA UDPOS score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mega-udpos", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MEGA UDPOS", + "raw_categories": "[\"language\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.604, + "details": { + "raw_score": "0.604", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-moe-instruct", + "raw_benchmark_id": "mega-udpos", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-moe-instruct\",\"https://llm-stats.com/benchmarks/mega-udpos\",\"https://api.llm-stats.com/leaderboard/benchmarks/mega-udpos\"]", + "raw_score_id": "mega-udpos::phi-3.5-moe-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mega-xcopa::mega-xcopa-phi-3.5-moe-instruct", + "evaluation_name": "llm_stats.mega-xcopa", + "source_data": { + "dataset_name": "MEGA XCOPA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-moe-instruct", + "https://llm-stats.com/benchmarks/mega-xcopa", + "https://api.llm-stats.com/leaderboard/benchmarks/mega-xcopa" + ], + "additional_details": { + "raw_benchmark_id": "mega-xcopa", + "raw_model_id": "phi-3.5-moe-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "XCOPA (Cross-lingual Choice of Plausible Alternatives) as part of the MEGA benchmark suite. A typologically diverse multilingual dataset for causal commonsense reasoning in 11 languages, including resource-poor languages like Eastern Apurímac Quechua and Haitian Creole. Requires models to select which choice is the effect or cause of a given premise.", + "metric_id": "llm_stats.mega-xcopa.score", + "metric_name": "MEGA XCOPA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mega-xcopa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MEGA XCOPA", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.766, + "details": { + "raw_score": "0.766", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-moe-instruct", + "raw_benchmark_id": "mega-xcopa", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-moe-instruct\",\"https://llm-stats.com/benchmarks/mega-xcopa\",\"https://api.llm-stats.com/leaderboard/benchmarks/mega-xcopa\"]", + "raw_score_id": "mega-xcopa::phi-3.5-moe-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mega-xstorycloze::mega-xstorycloze-phi-3.5-moe-instruct", + "evaluation_name": "llm_stats.mega-xstorycloze", + "source_data": { + "dataset_name": "MEGA XStoryCloze", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-moe-instruct", + "https://llm-stats.com/benchmarks/mega-xstorycloze", + "https://api.llm-stats.com/leaderboard/benchmarks/mega-xstorycloze" + ], + "additional_details": { + "raw_benchmark_id": "mega-xstorycloze", + "raw_model_id": "phi-3.5-moe-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "XStoryCloze as part of the MEGA benchmark suite. A cross-lingual story completion task that consists of professionally translated versions of the English StoryCloze dataset to 10 non-English languages. Requires models to predict the correct ending for a given four-sentence story, evaluating commonsense reasoning and narrative understanding.", + "metric_id": "llm_stats.mega-xstorycloze.score", + "metric_name": "MEGA XStoryCloze score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mega-xstorycloze", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MEGA XStoryCloze", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.828, + "details": { + "raw_score": "0.828", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-moe-instruct", + "raw_benchmark_id": "mega-xstorycloze", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-moe-instruct\",\"https://llm-stats.com/benchmarks/mega-xstorycloze\",\"https://api.llm-stats.com/leaderboard/benchmarks/mega-xstorycloze\"]", + "raw_score_id": "mega-xstorycloze::phi-3.5-moe-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "openbookqa::openbookqa-phi-3.5-moe-instruct", + "evaluation_name": "llm_stats.openbookqa", + "source_data": { + "dataset_name": "OpenBookQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-moe-instruct", + "https://llm-stats.com/benchmarks/openbookqa", + "https://api.llm-stats.com/leaderboard/benchmarks/openbookqa" + ], + "additional_details": { + "raw_benchmark_id": "openbookqa", + "raw_model_id": "phi-3.5-moe-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OpenBookQA is a question-answering dataset modeled after open book exams for assessing human understanding. It contains 5,957 multiple-choice elementary-level science questions that probe understanding of 1,326 core science facts and their application to novel situations, requiring combination of open book facts with broad common knowledge through multi-hop reasoning.", + "metric_id": "llm_stats.openbookqa.score", + "metric_name": "OpenBookQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "openbookqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OpenBookQA", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.896, + "details": { + "raw_score": "0.896", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-moe-instruct", + "raw_benchmark_id": "openbookqa", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-moe-instruct\",\"https://llm-stats.com/benchmarks/openbookqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/openbookqa\"]", + "raw_score_id": "openbookqa::phi-3.5-moe-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "piqa::piqa-phi-3.5-moe-instruct", + "evaluation_name": "llm_stats.piqa", + "source_data": { + "dataset_name": "PIQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-moe-instruct", + "https://llm-stats.com/benchmarks/piqa", + "https://api.llm-stats.com/leaderboard/benchmarks/piqa" + ], + "additional_details": { + "raw_benchmark_id": "piqa", + "raw_model_id": "phi-3.5-moe-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "PIQA (Physical Interaction: Question Answering) is a benchmark dataset for physical commonsense reasoning in natural language. It tests AI systems' ability to answer questions requiring physical world knowledge through multiple choice questions with everyday situations, focusing on atypical solutions inspired by instructables.com. The dataset contains 21,000 multiple choice questions where models must choose the most appropriate solution for physical interactions.", + "metric_id": "llm_stats.piqa.score", + "metric_name": "PIQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "piqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PIQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.886, + "details": { + "raw_score": "0.886", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-moe-instruct", + "raw_benchmark_id": "piqa", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-moe-instruct\",\"https://llm-stats.com/benchmarks/piqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/piqa\"]", + "raw_score_id": "piqa::phi-3.5-moe-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "qasper::qasper-phi-3.5-moe-instruct", + "evaluation_name": "llm_stats.qasper", + "source_data": { + "dataset_name": "Qasper", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-moe-instruct", + "https://llm-stats.com/benchmarks/qasper", + "https://api.llm-stats.com/leaderboard/benchmarks/qasper" + ], + "additional_details": { + "raw_benchmark_id": "qasper", + "raw_model_id": "phi-3.5-moe-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "QASPER is a dataset of 5,049 information-seeking questions and answers anchored in 1,585 NLP research papers. Questions are written by NLP practitioners who read only titles and abstracts, while answers require understanding the full paper text and provide supporting evidence. The dataset challenges models with complex reasoning across document sections for academic document question answering. Each question seeks information present in the full text and is answered by a separate set of NLP practitioners who also provide supporting evidence to answers.", + "metric_id": "llm_stats.qasper.score", + "metric_name": "Qasper score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "qasper", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Qasper", + "raw_categories": "[\"long_context\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.4, + "details": { + "raw_score": "0.4", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-moe-instruct", + "raw_benchmark_id": "qasper", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-moe-instruct\",\"https://llm-stats.com/benchmarks/qasper\",\"https://api.llm-stats.com/leaderboard/benchmarks/qasper\"]", + "raw_score_id": "qasper::phi-3.5-moe-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "qmsum::qmsum-phi-3.5-moe-instruct", + "evaluation_name": "llm_stats.qmsum", + "source_data": { + "dataset_name": "QMSum", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-moe-instruct", + "https://llm-stats.com/benchmarks/qmsum", + "https://api.llm-stats.com/leaderboard/benchmarks/qmsum" + ], + "additional_details": { + "raw_benchmark_id": "qmsum", + "raw_model_id": "phi-3.5-moe-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "QMSum is a benchmark for query-based multi-domain meeting summarization consisting of 1,808 query-summary pairs over 232 meetings across academic, product, and committee domains. The dataset enables models to select and summarize relevant spans of meetings in response to specific queries. Published at NAACL 2021, QMSum presents significant challenges in long meeting summarization where models must identify and summarize relevant content based on user queries.", + "metric_id": "llm_stats.qmsum.score", + "metric_name": "QMSum score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "qmsum", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "QMSum", + "raw_categories": "[\"long_context\",\"summarization\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.199, + "details": { + "raw_score": "0.199", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-moe-instruct", + "raw_benchmark_id": "qmsum", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-moe-instruct\",\"https://llm-stats.com/benchmarks/qmsum\",\"https://api.llm-stats.com/leaderboard/benchmarks/qmsum\"]", + "raw_score_id": "qmsum::phi-3.5-moe-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "repoqa::repoqa-phi-3.5-moe-instruct", + "evaluation_name": "llm_stats.repoqa", + "source_data": { + "dataset_name": "RepoQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-moe-instruct", + "https://llm-stats.com/benchmarks/repoqa", + "https://api.llm-stats.com/leaderboard/benchmarks/repoqa" + ], + "additional_details": { + "raw_benchmark_id": "repoqa", + "raw_model_id": "phi-3.5-moe-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "RepoQA is a benchmark for evaluating long-context code understanding capabilities of Large Language Models through the Searching Needle Function (SNF) task, where LLMs must locate specific functions in code repositories using natural language descriptions. The benchmark contains 500 code search tasks spanning 50 repositories across 5 modern programming languages (Python, Java, TypeScript, C++, and Rust), tested on 26 general and code-specific LLMs to assess their ability to comprehend and navigate code repositories.", + "metric_id": "llm_stats.repoqa.score", + "metric_name": "RepoQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "repoqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "RepoQA", + "raw_categories": "[\"long_context\",\"reasoning\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.85, + "details": { + "raw_score": "0.85", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-moe-instruct", + "raw_benchmark_id": "repoqa", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-moe-instruct\",\"https://llm-stats.com/benchmarks/repoqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/repoqa\"]", + "raw_score_id": "repoqa::phi-3.5-moe-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ruler::ruler-phi-3.5-moe-instruct", + "evaluation_name": "llm_stats.ruler", + "source_data": { + "dataset_name": "RULER", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-moe-instruct", + "https://llm-stats.com/benchmarks/ruler", + "https://api.llm-stats.com/leaderboard/benchmarks/ruler" + ], + "additional_details": { + "raw_benchmark_id": "ruler", + "raw_model_id": "phi-3.5-moe-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "RULER v1 is a synthetic long-context benchmark for measuring how model quality degrades as input length increases. This packaging follows the public standalone NVIDIA RULER implementation with 13 official tasks spanning retrieval, multi-hop tracing, aggregation, and QA.", + "metric_id": "llm_stats.ruler.score", + "metric_name": "RULER score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ruler", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "RULER", + "raw_categories": "[\"long_context\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.871, + "details": { + "raw_score": "0.871", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-moe-instruct", + "raw_benchmark_id": "ruler", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-moe-instruct\",\"https://llm-stats.com/benchmarks/ruler\",\"https://api.llm-stats.com/leaderboard/benchmarks/ruler\"]", + "raw_score_id": "ruler::phi-3.5-moe-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "social-iqa::social-iqa-phi-3.5-moe-instruct", + "evaluation_name": "llm_stats.social-iqa", + "source_data": { + "dataset_name": "Social IQa", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-moe-instruct", + "https://llm-stats.com/benchmarks/social-iqa", + "https://api.llm-stats.com/leaderboard/benchmarks/social-iqa" + ], + "additional_details": { + "raw_benchmark_id": "social-iqa", + "raw_model_id": "phi-3.5-moe-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The first large-scale benchmark for commonsense reasoning about social situations. Contains 38,000 multiple choice questions probing emotional and social intelligence in everyday situations, testing commonsense understanding of social interactions and theory of mind reasoning about the implied emotions and behavior of others.", + "metric_id": "llm_stats.social-iqa.score", + "metric_name": "Social IQa score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "social-iqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Social IQa", + "raw_categories": "[\"psychology\",\"reasoning\",\"creativity\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.78, + "details": { + "raw_score": "0.78", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-moe-instruct", + "raw_benchmark_id": "social-iqa", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-moe-instruct\",\"https://llm-stats.com/benchmarks/social-iqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/social-iqa\"]", + "raw_score_id": "social-iqa::phi-3.5-moe-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "squality::squality-phi-3.5-moe-instruct", + "evaluation_name": "llm_stats.squality", + "source_data": { + "dataset_name": "SQuALITY", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-moe-instruct", + "https://llm-stats.com/benchmarks/squality", + "https://api.llm-stats.com/leaderboard/benchmarks/squality" + ], + "additional_details": { + "raw_benchmark_id": "squality", + "raw_model_id": "phi-3.5-moe-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SQuALITY (Summarization-format QUestion Answering with Long Input Texts, Yes!) is a long-document summarization dataset built by hiring highly-qualified contractors to read public-domain short stories (3000-6000 words) and write original summaries from scratch. Each document has five summaries: one overview and four question-focused summaries. Designed to address limitations in existing summarization datasets by providing high-quality, faithful summaries.", + "metric_id": "llm_stats.squality.score", + "metric_name": "SQuALITY score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "squality", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SQuALITY", + "raw_categories": "[\"language\",\"long_context\",\"summarization\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.241, + "details": { + "raw_score": "0.241", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-moe-instruct", + "raw_benchmark_id": "squality", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-moe-instruct\",\"https://llm-stats.com/benchmarks/squality\",\"https://api.llm-stats.com/leaderboard/benchmarks/squality\"]", + "raw_score_id": "squality::phi-3.5-moe-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "summscreenfd::summscreenfd-phi-3.5-moe-instruct", + "evaluation_name": "llm_stats.summscreenfd", + "source_data": { + "dataset_name": "SummScreenFD", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-moe-instruct", + "https://llm-stats.com/benchmarks/summscreenfd", + "https://api.llm-stats.com/leaderboard/benchmarks/summscreenfd" + ], + "additional_details": { + "raw_benchmark_id": "summscreenfd", + "raw_model_id": "phi-3.5-moe-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SummScreenFD is the ForeverDreaming subset of the SummScreen dataset for abstractive screenplay summarization, comprising pairs of TV series transcripts and human-written recaps from 88 different shows. The dataset provides a challenging testbed for abstractive summarization where plot details are often expressed indirectly in character dialogues and scattered across the entirety of the transcript, requiring models to find and integrate these details to form succinct plot descriptions.", + "metric_id": "llm_stats.summscreenfd.score", + "metric_name": "SummScreenFD score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "summscreenfd", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SummScreenFD", + "raw_categories": "[\"long_context\",\"summarization\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.169, + "details": { + "raw_score": "0.169", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-moe-instruct", + "raw_benchmark_id": "summscreenfd", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-moe-instruct\",\"https://llm-stats.com/benchmarks/summscreenfd\",\"https://api.llm-stats.com/leaderboard/benchmarks/summscreenfd\"]", + "raw_score_id": "summscreenfd::phi-3.5-moe-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "truthfulqa::truthfulqa-phi-3.5-moe-instruct", + "evaluation_name": "llm_stats.truthfulqa", + "source_data": { + "dataset_name": "TruthfulQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-moe-instruct", + "https://llm-stats.com/benchmarks/truthfulqa", + "https://api.llm-stats.com/leaderboard/benchmarks/truthfulqa" + ], + "additional_details": { + "raw_benchmark_id": "truthfulqa", + "raw_model_id": "phi-3.5-moe-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "TruthfulQA is a benchmark to measure whether language models are truthful in generating answers to questions. It comprises 817 questions that span 38 categories, including health, law, finance and politics. The questions are crafted such that some humans would answer falsely due to a false belief or misconception, testing models' ability to avoid generating false answers learned from human texts.", + "metric_id": "llm_stats.truthfulqa.score", + "metric_name": "TruthfulQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "truthfulqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TruthfulQA", + "raw_categories": "[\"legal\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.775, + "details": { + "raw_score": "0.775", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-moe-instruct", + "raw_benchmark_id": "truthfulqa", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-moe-instruct\",\"https://llm-stats.com/benchmarks/truthfulqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/truthfulqa\"]", + "raw_score_id": "truthfulqa::phi-3.5-moe-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "winogrande::winogrande-phi-3.5-moe-instruct", + "evaluation_name": "llm_stats.winogrande", + "source_data": { + "dataset_name": "Winogrande", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-3.5-moe-instruct", + "https://llm-stats.com/benchmarks/winogrande", + "https://api.llm-stats.com/leaderboard/benchmarks/winogrande" + ], + "additional_details": { + "raw_benchmark_id": "winogrande", + "raw_model_id": "phi-3.5-moe-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WinoGrande: An Adversarial Winograd Schema Challenge at Scale. A large-scale dataset of 44,000 pronoun resolution problems designed to test machine commonsense reasoning. Uses adversarial filtering to reduce spurious biases and provides a more robust evaluation of whether AI systems truly understand commonsense or exploit statistical shortcuts. Current best AI methods achieve 59.4-79.1% accuracy, significantly below human performance of 94.0%.", + "metric_id": "llm_stats.winogrande.score", + "metric_name": "Winogrande score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "winogrande", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Winogrande", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "22" + } + }, + "score_details": { + "score": 0.813, + "details": { + "raw_score": "0.813", + "raw_score_field": "score", + "raw_model_id": "phi-3.5-moe-instruct", + "raw_benchmark_id": "winogrande", + "source_urls_json": "[\"https://llm-stats.com/models/phi-3.5-moe-instruct\",\"https://llm-stats.com/benchmarks/winogrande\",\"https://api.llm-stats.com/leaderboard/benchmarks/winogrande\"]", + "raw_score_id": "winogrande::phi-3.5-moe-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/microsoft/phi-4-mini/4b2744e5-3793-4713-9940-147b9e096119.json b/data/llm_stats/microsoft/phi-4-mini/4b2744e5-3793-4713-9940-147b9e096119.json new file mode 100644 index 0000000000000000000000000000000000000000..a09a1d154dc7049392d39fd97d2d04299ef1fcf2 --- /dev/null +++ b/data/llm_stats/microsoft/phi-4-mini/4b2744e5-3793-4713-9940-147b9e096119.json @@ -0,0 +1,515 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/microsoft_phi-4-mini/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Phi 4 Mini", + "id": "microsoft/phi-4-mini", + "developer": "microsoft", + "additional_details": { + "raw_id": "phi-4-mini", + "raw_name": "Phi 4 Mini", + "raw_model_id": "phi-4-mini", + "raw_model_name": "Phi 4 Mini", + "raw_organization_id": "microsoft", + "raw_organization_name": "Microsoft", + "raw_release_date": "2025-02-01", + "raw_announcement_date": "2025-02-01", + "raw_multimodal": "false", + "raw_param_count": "3840000000", + "raw_provider_slug": "microsoft", + "raw_provider_name": "Microsoft" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "arc-c::arc-c-phi-4-mini", + "evaluation_name": "llm_stats.arc-c", + "source_data": { + "dataset_name": "ARC-C", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-4-mini", + "https://llm-stats.com/benchmarks/arc-c", + "https://api.llm-stats.com/leaderboard/benchmarks/arc-c" + ], + "additional_details": { + "raw_benchmark_id": "arc-c", + "raw_model_id": "phi-4-mini", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The AI2 Reasoning Challenge (ARC) Challenge Set is a multiple-choice question-answering benchmark containing grade-school level science questions that require advanced reasoning capabilities. ARC-C specifically contains questions that were answered incorrectly by both retrieval-based and word co-occurrence algorithms, making it a particularly challenging subset designed to test commonsense reasoning abilities in AI systems.", + "metric_id": "llm_stats.arc-c.score", + "metric_name": "ARC-C score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arc-c", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ARC-C", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "34" + } + }, + "score_details": { + "score": 0.837, + "details": { + "raw_score": "0.837", + "raw_score_field": "score", + "raw_model_id": "phi-4-mini", + "raw_benchmark_id": "arc-c", + "source_urls_json": "[\"https://llm-stats.com/models/phi-4-mini\",\"https://llm-stats.com/benchmarks/arc-c\",\"https://api.llm-stats.com/leaderboard/benchmarks/arc-c\"]", + "raw_score_id": "arc-c::phi-4-mini", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "big-bench-hard::big-bench-hard-phi-4-mini", + "evaluation_name": "llm_stats.big-bench-hard", + "source_data": { + "dataset_name": "BIG-Bench Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-4-mini", + "https://llm-stats.com/benchmarks/big-bench-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard" + ], + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_model_id": "phi-4-mini", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIG-Bench Hard (BBH) is a subset of 23 challenging BIG-Bench tasks selected because prior language model evaluations did not outperform average human-rater performance. The benchmark contains 6,511 evaluation examples testing various forms of multi-step reasoning including arithmetic, logical reasoning (Boolean expressions, logical deduction), geometric reasoning, temporal reasoning, and language understanding. Tasks require capabilities such as causal judgment, object counting, navigation, pattern recognition, and complex problem solving.", + "metric_id": "llm_stats.big-bench-hard.score", + "metric_name": "BIG-Bench Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "big-bench-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BIG-Bench Hard", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "21" + } + }, + "score_details": { + "score": 0.704, + "details": { + "raw_score": "0.704", + "raw_score_field": "score", + "raw_model_id": "phi-4-mini", + "raw_benchmark_id": "big-bench-hard", + "source_urls_json": "[\"https://llm-stats.com/models/phi-4-mini\",\"https://llm-stats.com/benchmarks/big-bench-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/big-bench-hard\"]", + "raw_score_id": "big-bench-hard::phi-4-mini", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "boolq::boolq-phi-4-mini", + "evaluation_name": "llm_stats.boolq", + "source_data": { + "dataset_name": "BoolQ", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-4-mini", + "https://llm-stats.com/benchmarks/boolq", + "https://api.llm-stats.com/leaderboard/benchmarks/boolq" + ], + "additional_details": { + "raw_benchmark_id": "boolq", + "raw_model_id": "phi-4-mini", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BoolQ is a reading comprehension dataset for yes/no questions containing 15,942 naturally occurring examples. Each example consists of a question, passage, and boolean answer, where questions are generated in unprompted and unconstrained settings. The dataset challenges models with complex, non-factoid information requiring entailment-like inference to solve.", + "metric_id": "llm_stats.boolq.score", + "metric_name": "BoolQ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "boolq", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BoolQ", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.812, + "details": { + "raw_score": "0.812", + "raw_score_field": "score", + "raw_model_id": "phi-4-mini", + "raw_benchmark_id": "boolq", + "source_urls_json": "[\"https://llm-stats.com/models/phi-4-mini\",\"https://llm-stats.com/benchmarks/boolq\",\"https://api.llm-stats.com/leaderboard/benchmarks/boolq\"]", + "raw_score_id": "boolq::phi-4-mini", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "multilingual-mmlu::multilingual-mmlu-phi-4-mini", + "evaluation_name": "llm_stats.multilingual-mmlu", + "source_data": { + "dataset_name": "Multilingual MMLU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-4-mini", + "https://llm-stats.com/benchmarks/multilingual-mmlu", + "https://api.llm-stats.com/leaderboard/benchmarks/multilingual-mmlu" + ], + "additional_details": { + "raw_benchmark_id": "multilingual-mmlu", + "raw_model_id": "phi-4-mini", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MMLU-ProX is a comprehensive multilingual benchmark covering 29 typologically diverse languages, building upon MMLU-Pro. Each language version consists of 11,829 identical questions enabling direct cross-linguistic comparisons. The benchmark evaluates large language models' reasoning capabilities across linguistic and cultural boundaries through challenging, reasoning-focused questions with 10 answer choices.", + "metric_id": "llm_stats.multilingual-mmlu.score", + "metric_name": "Multilingual MMLU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "multilingual-mmlu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Multilingual MMLU", + "raw_categories": "[\"language\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.493, + "details": { + "raw_score": "0.493", + "raw_score_field": "score", + "raw_model_id": "phi-4-mini", + "raw_benchmark_id": "multilingual-mmlu", + "source_urls_json": "[\"https://llm-stats.com/models/phi-4-mini\",\"https://llm-stats.com/benchmarks/multilingual-mmlu\",\"https://api.llm-stats.com/leaderboard/benchmarks/multilingual-mmlu\"]", + "raw_score_id": "multilingual-mmlu::phi-4-mini", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "openbookqa::openbookqa-phi-4-mini", + "evaluation_name": "llm_stats.openbookqa", + "source_data": { + "dataset_name": "OpenBookQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-4-mini", + "https://llm-stats.com/benchmarks/openbookqa", + "https://api.llm-stats.com/leaderboard/benchmarks/openbookqa" + ], + "additional_details": { + "raw_benchmark_id": "openbookqa", + "raw_model_id": "phi-4-mini", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OpenBookQA is a question-answering dataset modeled after open book exams for assessing human understanding. It contains 5,957 multiple-choice elementary-level science questions that probe understanding of 1,326 core science facts and their application to novel situations, requiring combination of open book facts with broad common knowledge through multi-hop reasoning.", + "metric_id": "llm_stats.openbookqa.score", + "metric_name": "OpenBookQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "openbookqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OpenBookQA", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.792, + "details": { + "raw_score": "0.792", + "raw_score_field": "score", + "raw_model_id": "phi-4-mini", + "raw_benchmark_id": "openbookqa", + "source_urls_json": "[\"https://llm-stats.com/models/phi-4-mini\",\"https://llm-stats.com/benchmarks/openbookqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/openbookqa\"]", + "raw_score_id": "openbookqa::phi-4-mini", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "piqa::piqa-phi-4-mini", + "evaluation_name": "llm_stats.piqa", + "source_data": { + "dataset_name": "PIQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-4-mini", + "https://llm-stats.com/benchmarks/piqa", + "https://api.llm-stats.com/leaderboard/benchmarks/piqa" + ], + "additional_details": { + "raw_benchmark_id": "piqa", + "raw_model_id": "phi-4-mini", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "PIQA (Physical Interaction: Question Answering) is a benchmark dataset for physical commonsense reasoning in natural language. It tests AI systems' ability to answer questions requiring physical world knowledge through multiple choice questions with everyday situations, focusing on atypical solutions inspired by instructables.com. The dataset contains 21,000 multiple choice questions where models must choose the most appropriate solution for physical interactions.", + "metric_id": "llm_stats.piqa.score", + "metric_name": "PIQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "piqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PIQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.776, + "details": { + "raw_score": "0.776", + "raw_score_field": "score", + "raw_model_id": "phi-4-mini", + "raw_benchmark_id": "piqa", + "source_urls_json": "[\"https://llm-stats.com/models/phi-4-mini\",\"https://llm-stats.com/benchmarks/piqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/piqa\"]", + "raw_score_id": "piqa::phi-4-mini", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "social-iqa::social-iqa-phi-4-mini", + "evaluation_name": "llm_stats.social-iqa", + "source_data": { + "dataset_name": "Social IQa", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-4-mini", + "https://llm-stats.com/benchmarks/social-iqa", + "https://api.llm-stats.com/leaderboard/benchmarks/social-iqa" + ], + "additional_details": { + "raw_benchmark_id": "social-iqa", + "raw_model_id": "phi-4-mini", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The first large-scale benchmark for commonsense reasoning about social situations. Contains 38,000 multiple choice questions probing emotional and social intelligence in everyday situations, testing commonsense understanding of social interactions and theory of mind reasoning about the implied emotions and behavior of others.", + "metric_id": "llm_stats.social-iqa.score", + "metric_name": "Social IQa score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "social-iqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Social IQa", + "raw_categories": "[\"psychology\",\"reasoning\",\"creativity\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.725, + "details": { + "raw_score": "0.725", + "raw_score_field": "score", + "raw_model_id": "phi-4-mini", + "raw_benchmark_id": "social-iqa", + "source_urls_json": "[\"https://llm-stats.com/models/phi-4-mini\",\"https://llm-stats.com/benchmarks/social-iqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/social-iqa\"]", + "raw_score_id": "social-iqa::phi-4-mini", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "truthfulqa::truthfulqa-phi-4-mini", + "evaluation_name": "llm_stats.truthfulqa", + "source_data": { + "dataset_name": "TruthfulQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-4-mini", + "https://llm-stats.com/benchmarks/truthfulqa", + "https://api.llm-stats.com/leaderboard/benchmarks/truthfulqa" + ], + "additional_details": { + "raw_benchmark_id": "truthfulqa", + "raw_model_id": "phi-4-mini", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "TruthfulQA is a benchmark to measure whether language models are truthful in generating answers to questions. It comprises 817 questions that span 38 categories, including health, law, finance and politics. The questions are crafted such that some humans would answer falsely due to a false belief or misconception, testing models' ability to avoid generating false answers learned from human texts.", + "metric_id": "llm_stats.truthfulqa.score", + "metric_name": "TruthfulQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "truthfulqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TruthfulQA", + "raw_categories": "[\"legal\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.664, + "details": { + "raw_score": "0.664", + "raw_score_field": "score", + "raw_model_id": "phi-4-mini", + "raw_benchmark_id": "truthfulqa", + "source_urls_json": "[\"https://llm-stats.com/models/phi-4-mini\",\"https://llm-stats.com/benchmarks/truthfulqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/truthfulqa\"]", + "raw_score_id": "truthfulqa::phi-4-mini", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "winogrande::winogrande-phi-4-mini", + "evaluation_name": "llm_stats.winogrande", + "source_data": { + "dataset_name": "Winogrande", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/phi-4-mini", + "https://llm-stats.com/benchmarks/winogrande", + "https://api.llm-stats.com/leaderboard/benchmarks/winogrande" + ], + "additional_details": { + "raw_benchmark_id": "winogrande", + "raw_model_id": "phi-4-mini", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WinoGrande: An Adversarial Winograd Schema Challenge at Scale. A large-scale dataset of 44,000 pronoun resolution problems designed to test machine commonsense reasoning. Uses adversarial filtering to reduce spurious biases and provides a more robust evaluation of whether AI systems truly understand commonsense or exploit statistical shortcuts. Current best AI methods achieve 59.4-79.1% accuracy, significantly below human performance of 94.0%.", + "metric_id": "llm_stats.winogrande.score", + "metric_name": "Winogrande score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "winogrande", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Winogrande", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "22" + } + }, + "score_details": { + "score": 0.67, + "details": { + "raw_score": "0.67", + "raw_score_field": "score", + "raw_model_id": "phi-4-mini", + "raw_benchmark_id": "winogrande", + "source_urls_json": "[\"https://llm-stats.com/models/phi-4-mini\",\"https://llm-stats.com/benchmarks/winogrande\",\"https://api.llm-stats.com/leaderboard/benchmarks/winogrande\"]", + "raw_score_id": "winogrande::phi-4-mini", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/minimax/minimax-m2.1/df7327e3-1be4-4aaa-8347-4d47391be13a.json b/data/llm_stats/minimax/minimax-m2.1/df7327e3-1be4-4aaa-8347-4d47391be13a.json new file mode 100644 index 0000000000000000000000000000000000000000..2c50dc12a6f7f304922e6178521461f494889520 --- /dev/null +++ b/data/llm_stats/minimax/minimax-m2.1/df7327e3-1be4-4aaa-8347-4d47391be13a.json @@ -0,0 +1,1088 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/minimax_minimax-m2.1/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "minimax-m2.1", + "id": "minimax/minimax-m2.1", + "developer": "minimax", + "additional_details": { + "raw_id": "minimax-m2.1", + "raw_model_name": "minimax-m2.1", + "raw_display_name": "MiniMax M2.1", + "raw_organization_id": "minimax", + "raw_organization_name": "MiniMax", + "raw_context_length": "1000000", + "raw_input_modalities": "[\"text\"]", + "raw_output_modalities": "[\"text\"]", + "raw_input_price": "0.3", + "raw_output_price": "1.2", + "raw_model_id": "minimax-m2.1", + "raw_provider_slug": "minimax", + "raw_provider_name": "ZeroEval" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aa-lcr::aa-lcr-minimax-m2.1", + "evaluation_name": "llm_stats.aa-lcr", + "source_data": { + "dataset_name": "AA-LCR", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m2.1", + "https://llm-stats.com/benchmarks/aa-lcr", + "https://api.llm-stats.com/leaderboard/benchmarks/aa-lcr" + ], + "additional_details": { + "raw_benchmark_id": "aa-lcr", + "raw_model_id": "minimax-m2.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Agent Arena Long Context Reasoning benchmark", + "metric_id": "llm_stats.aa-lcr.score", + "metric_name": "AA-LCR score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aa-lcr", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AA-LCR", + "raw_categories": "[\"long_context\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "14" + } + }, + "score_details": { + "score": 0.62, + "details": { + "raw_score": "0.62", + "raw_score_field": "score", + "raw_model_id": "minimax-m2.1", + "raw_benchmark_id": "aa-lcr", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m2.1\",\"https://llm-stats.com/benchmarks/aa-lcr\",\"https://api.llm-stats.com/leaderboard/benchmarks/aa-lcr\"]", + "raw_score_id": "aa-lcr::minimax-m2.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ifbench::ifbench-minimax-m2.1", + "evaluation_name": "llm_stats.ifbench", + "source_data": { + "dataset_name": "IFBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m2.1", + "https://llm-stats.com/benchmarks/ifbench", + "https://api.llm-stats.com/leaderboard/benchmarks/ifbench" + ], + "additional_details": { + "raw_benchmark_id": "ifbench", + "raw_model_id": "minimax-m2.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Instruction Following Benchmark evaluating model's ability to follow complex instructions", + "metric_id": "llm_stats.ifbench.score", + "metric_name": "IFBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ifbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IFBench", + "raw_categories": "[\"instruction_following\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.7, + "details": { + "raw_score": "0.7", + "raw_score_field": "score", + "raw_model_id": "minimax-m2.1", + "raw_benchmark_id": "ifbench", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m2.1\",\"https://llm-stats.com/benchmarks/ifbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/ifbench\"]", + "raw_score_id": "ifbench::minimax-m2.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "livecodebench::livecodebench-minimax-m2.1", + "evaluation_name": "llm_stats.livecodebench", + "source_data": { + "dataset_name": "LiveCodeBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m2.1", + "https://llm-stats.com/benchmarks/livecodebench", + "https://api.llm-stats.com/leaderboard/benchmarks/livecodebench" + ], + "additional_details": { + "raw_benchmark_id": "livecodebench", + "raw_model_id": "minimax-m2.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LiveCodeBench is a holistic and contamination-free evaluation benchmark for large language models for code. It continuously collects new problems from programming contests (LeetCode, AtCoder, CodeForces) and evaluates four different scenarios: code generation, self-repair, code execution, and test output prediction. Problems are annotated with release dates to enable evaluation on unseen problems released after a model's training cutoff.", + "metric_id": "llm_stats.livecodebench.score", + "metric_name": "LiveCodeBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "livecodebench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LiveCodeBench", + "raw_categories": "[\"reasoning\",\"general\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "73" + } + }, + "score_details": { + "score": 0.78, + "details": { + "raw_score": "0.78", + "raw_score_field": "score", + "raw_model_id": "minimax-m2.1", + "raw_benchmark_id": "livecodebench", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m2.1\",\"https://llm-stats.com/benchmarks/livecodebench\",\"https://api.llm-stats.com/leaderboard/benchmarks/livecodebench\"]", + "raw_score_id": "livecodebench::minimax-m2.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu-pro::mmlu-pro-minimax-m2.1", + "evaluation_name": "llm_stats.mmlu-pro", + "source_data": { + "dataset_name": "MMLU-Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m2.1", + "https://llm-stats.com/benchmarks/mmlu-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu-pro" + ], + "additional_details": { + "raw_benchmark_id": "mmlu-pro", + "raw_model_id": "minimax-m2.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A more robust and challenging multi-task language understanding benchmark that extends MMLU by expanding multiple-choice options from 4 to 10, eliminating trivial questions, and focusing on reasoning-intensive tasks. Features over 12,000 curated questions across 14 domains and causes a 16-33% accuracy drop compared to original MMLU.", + "metric_id": "llm_stats.mmlu-pro.score", + "metric_name": "MMLU-Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU-Pro", + "raw_categories": "[\"language\",\"legal\",\"math\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "129" + } + }, + "score_details": { + "score": 0.88, + "details": { + "raw_score": "0.88", + "raw_score_field": "score", + "raw_model_id": "minimax-m2.1", + "raw_benchmark_id": "mmlu-pro", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m2.1\",\"https://llm-stats.com/benchmarks/mmlu-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu-pro\"]", + "raw_score_id": "mmlu-pro::minimax-m2.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "multi-swe-bench::multi-swe-bench-minimax-m2.1", + "evaluation_name": "llm_stats.multi-swe-bench", + "source_data": { + "dataset_name": "Multi-SWE-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m2.1", + "https://llm-stats.com/benchmarks/multi-swe-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/multi-swe-bench" + ], + "additional_details": { + "raw_benchmark_id": "multi-swe-bench", + "raw_model_id": "minimax-m2.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multilingual benchmark for issue resolving that evaluates Large Language Models' ability to resolve software issues across diverse programming ecosystems. Covers 7 programming languages (Java, TypeScript, JavaScript, Go, Rust, C, and C++) with 1,632 high-quality instances carefully annotated by 68 expert annotators. Addresses limitations of existing benchmarks that focus almost exclusively on Python.", + "metric_id": "llm_stats.multi-swe-bench.score", + "metric_name": "Multi-SWE-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "multi-swe-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Multi-SWE-Bench", + "raw_categories": "[\"reasoning\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.494, + "details": { + "raw_score": "0.494", + "raw_score_field": "score", + "raw_model_id": "minimax-m2.1", + "raw_benchmark_id": "multi-swe-bench", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m2.1\",\"https://llm-stats.com/benchmarks/multi-swe-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/multi-swe-bench\"]", + "raw_score_id": "multi-swe-bench::minimax-m2.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "octocodingbench::octocodingbench-minimax-m2.1", + "evaluation_name": "llm_stats.octocodingbench", + "source_data": { + "dataset_name": "OctoCodingBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m2.1", + "https://llm-stats.com/benchmarks/octocodingbench", + "https://api.llm-stats.com/leaderboard/benchmarks/octocodingbench" + ], + "additional_details": { + "raw_benchmark_id": "octocodingbench", + "raw_model_id": "minimax-m2.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Octopus coding benchmark for evaluating multi-language programming capabilities", + "metric_id": "llm_stats.octocodingbench.score", + "metric_name": "OctoCodingBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "octocodingbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OctoCodingBench", + "raw_categories": "[\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.261, + "details": { + "raw_score": "0.261", + "raw_score_field": "score", + "raw_model_id": "minimax-m2.1", + "raw_benchmark_id": "octocodingbench", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m2.1\",\"https://llm-stats.com/benchmarks/octocodingbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/octocodingbench\"]", + "raw_score_id": "octocodingbench::minimax-m2.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "scicode::scicode-minimax-m2.1", + "evaluation_name": "llm_stats.scicode", + "source_data": { + "dataset_name": "SciCode", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m2.1", + "https://llm-stats.com/benchmarks/scicode", + "https://api.llm-stats.com/leaderboard/benchmarks/scicode" + ], + "additional_details": { + "raw_benchmark_id": "scicode", + "raw_model_id": "minimax-m2.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SciCode is a research coding benchmark curated by scientists that challenges language models to code solutions for scientific problems. It contains 338 subproblems decomposed from 80 challenging main problems across 16 natural science sub-fields including mathematics, physics, chemistry, biology, and materials science. Problems require knowledge recall, reasoning, and code synthesis skills.", + "metric_id": "llm_stats.scicode.score", + "metric_name": "SciCode score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "scicode", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SciCode", + "raw_categories": "[\"math\",\"physics\",\"reasoning\",\"biology\",\"chemistry\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.39, + "details": { + "raw_score": "0.39", + "raw_score_field": "score", + "raw_model_id": "minimax-m2.1", + "raw_benchmark_id": "scicode", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m2.1\",\"https://llm-stats.com/benchmarks/scicode\",\"https://api.llm-stats.com/leaderboard/benchmarks/scicode\"]", + "raw_score_id": "scicode::minimax-m2.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-multilingual::swe-bench-multilingual-minimax-m2.1", + "evaluation_name": "llm_stats.swe-bench-multilingual", + "source_data": { + "dataset_name": "SWE-bench Multilingual", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m2.1", + "https://llm-stats.com/benchmarks/swe-bench-multilingual", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-multilingual" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-multilingual", + "raw_model_id": "minimax-m2.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multilingual benchmark for issue resolving in software engineering that covers Java, TypeScript, JavaScript, Go, Rust, C, and C++. Contains 1,632 high-quality instances carefully annotated from 2,456 candidates by 68 expert annotators, designed to evaluate Large Language Models across diverse software ecosystems beyond Python.", + "metric_id": "llm_stats.swe-bench-multilingual.score", + "metric_name": "SWE-bench Multilingual score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-multilingual", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-bench Multilingual", + "raw_categories": "[\"reasoning\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "33" + } + }, + "score_details": { + "score": 0.725, + "details": { + "raw_score": "0.725", + "raw_score_field": "score", + "raw_model_id": "minimax-m2.1", + "raw_benchmark_id": "swe-bench-multilingual", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m2.1\",\"https://llm-stats.com/benchmarks/swe-bench-multilingual\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-multilingual\"]", + "raw_score_id": "swe-bench-multilingual::minimax-m2.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-perf::swe-perf-minimax-m2.1", + "evaluation_name": "llm_stats.swe-perf", + "source_data": { + "dataset_name": "SWE-Perf", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m2.1", + "https://llm-stats.com/benchmarks/swe-perf", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-perf" + ], + "additional_details": { + "raw_benchmark_id": "swe-perf", + "raw_model_id": "minimax-m2.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Software Engineering Performance benchmark measuring code optimization capabilities", + "metric_id": "llm_stats.swe-perf.score", + "metric_name": "SWE-Perf score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-perf", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Perf", + "raw_categories": "[\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.031, + "details": { + "raw_score": "0.031", + "raw_score_field": "score", + "raw_model_id": "minimax-m2.1", + "raw_benchmark_id": "swe-perf", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m2.1\",\"https://llm-stats.com/benchmarks/swe-perf\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-perf\"]", + "raw_score_id": "swe-perf::minimax-m2.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-review::swe-review-minimax-m2.1", + "evaluation_name": "llm_stats.swe-review", + "source_data": { + "dataset_name": "SWE-Review", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m2.1", + "https://llm-stats.com/benchmarks/swe-review", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-review" + ], + "additional_details": { + "raw_benchmark_id": "swe-review", + "raw_model_id": "minimax-m2.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Software Engineering Review benchmark evaluating code review capabilities", + "metric_id": "llm_stats.swe-review.score", + "metric_name": "SWE-Review score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-review", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Review", + "raw_categories": "[\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.089, + "details": { + "raw_score": "0.089", + "raw_score_field": "score", + "raw_model_id": "minimax-m2.1", + "raw_benchmark_id": "swe-review", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m2.1\",\"https://llm-stats.com/benchmarks/swe-review\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-review\"]", + "raw_score_id": "swe-review::minimax-m2.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swt-bench::swt-bench-minimax-m2.1", + "evaluation_name": "llm_stats.swt-bench", + "source_data": { + "dataset_name": "SWT-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m2.1", + "https://llm-stats.com/benchmarks/swt-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/swt-bench" + ], + "additional_details": { + "raw_benchmark_id": "swt-bench", + "raw_model_id": "minimax-m2.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Software Test Benchmark evaluating LLM ability to write tests for software repositories", + "metric_id": "llm_stats.swt-bench.score", + "metric_name": "SWT-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swt-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWT-Bench", + "raw_categories": "[\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.693, + "details": { + "raw_score": "0.693", + "raw_score_field": "score", + "raw_model_id": "minimax-m2.1", + "raw_benchmark_id": "swt-bench", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m2.1\",\"https://llm-stats.com/benchmarks/swt-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/swt-bench\"]", + "raw_score_id": "swt-bench::minimax-m2.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "tau2-telecom::tau2-telecom-minimax-m2.1", + "evaluation_name": "llm_stats.tau2-telecom", + "source_data": { + "dataset_name": "Tau2 Telecom", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m2.1", + "https://llm-stats.com/benchmarks/tau2-telecom", + "https://api.llm-stats.com/leaderboard/benchmarks/tau2-telecom" + ], + "additional_details": { + "raw_benchmark_id": "tau2-telecom", + "raw_model_id": "minimax-m2.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "τ²-Bench telecom domain evaluates conversational agents in a dual-control environment modeled as a Dec-POMDP, where both agent and user use tools in shared telecommunications troubleshooting scenarios that test coordination and communication capabilities.", + "metric_id": "llm_stats.tau2-telecom.score", + "metric_name": "Tau2 Telecom score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "tau2-telecom", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Tau2 Telecom", + "raw_categories": "[\"reasoning\",\"communication\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "35" + } + }, + "score_details": { + "score": 0.87, + "details": { + "raw_score": "0.87", + "raw_score_field": "score", + "raw_model_id": "minimax-m2.1", + "raw_benchmark_id": "tau2-telecom", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m2.1\",\"https://llm-stats.com/benchmarks/tau2-telecom\",\"https://api.llm-stats.com/leaderboard/benchmarks/tau2-telecom\"]", + "raw_score_id": "tau2-telecom::minimax-m2.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "terminal-bench::terminal-bench-minimax-m2.1", + "evaluation_name": "llm_stats.terminal-bench", + "source_data": { + "dataset_name": "Terminal-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m2.1", + "https://llm-stats.com/benchmarks/terminal-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench" + ], + "additional_details": { + "raw_benchmark_id": "terminal-bench", + "raw_model_id": "minimax-m2.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Terminal-Bench is a benchmark for testing AI agents in real terminal environments. It evaluates how well agents can handle real-world, end-to-end tasks autonomously, including compiling code, training models, setting up servers, system administration, security tasks, data science workflows, and cybersecurity vulnerabilities. The benchmark consists of a dataset of ~100 hand-crafted, human-verified tasks and an execution harness that connects language models to a terminal sandbox.", + "metric_id": "llm_stats.terminal-bench.score", + "metric_name": "Terminal-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "terminal-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Terminal-Bench", + "raw_categories": "[\"reasoning\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "25" + } + }, + "score_details": { + "score": 0.479, + "details": { + "raw_score": "0.479", + "raw_score_field": "score", + "raw_model_id": "minimax-m2.1", + "raw_benchmark_id": "terminal-bench", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m2.1\",\"https://llm-stats.com/benchmarks/terminal-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench\"]", + "raw_score_id": "terminal-bench::minimax-m2.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "toolathlon::toolathlon-minimax-m2.1", + "evaluation_name": "llm_stats.toolathlon", + "source_data": { + "dataset_name": "Toolathlon", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m2.1", + "https://llm-stats.com/benchmarks/toolathlon", + "https://api.llm-stats.com/leaderboard/benchmarks/toolathlon" + ], + "additional_details": { + "raw_benchmark_id": "toolathlon", + "raw_model_id": "minimax-m2.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Tool Decathlon is a comprehensive benchmark for evaluating AI agents' ability to use multiple tools across diverse task categories. It measures proficiency in tool selection, sequencing, and execution across ten different tool-use scenarios.", + "metric_id": "llm_stats.toolathlon.score", + "metric_name": "Toolathlon score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "toolathlon", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Toolathlon", + "raw_categories": "[\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "24" + } + }, + "score_details": { + "score": 0.435, + "details": { + "raw_score": "0.435", + "raw_score_field": "score", + "raw_model_id": "minimax-m2.1", + "raw_benchmark_id": "toolathlon", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m2.1\",\"https://llm-stats.com/benchmarks/toolathlon\",\"https://api.llm-stats.com/leaderboard/benchmarks/toolathlon\"]", + "raw_score_id": "toolathlon::minimax-m2.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "vibe-android::vibe-android-minimax-m2.1", + "evaluation_name": "llm_stats.vibe-android", + "source_data": { + "dataset_name": "VIBE Android", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m2.1", + "https://llm-stats.com/benchmarks/vibe-android", + "https://api.llm-stats.com/leaderboard/benchmarks/vibe-android" + ], + "additional_details": { + "raw_benchmark_id": "vibe-android", + "raw_model_id": "minimax-m2.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VIBE benchmark subset for Android application generation", + "metric_id": "llm_stats.vibe-android.score", + "metric_name": "VIBE Android score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "vibe-android", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VIBE Android", + "raw_categories": "[\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.897, + "details": { + "raw_score": "0.897", + "raw_score_field": "score", + "raw_model_id": "minimax-m2.1", + "raw_benchmark_id": "vibe-android", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m2.1\",\"https://llm-stats.com/benchmarks/vibe-android\",\"https://api.llm-stats.com/leaderboard/benchmarks/vibe-android\"]", + "raw_score_id": "vibe-android::minimax-m2.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "vibe-backend::vibe-backend-minimax-m2.1", + "evaluation_name": "llm_stats.vibe-backend", + "source_data": { + "dataset_name": "VIBE Backend", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m2.1", + "https://llm-stats.com/benchmarks/vibe-backend", + "https://api.llm-stats.com/leaderboard/benchmarks/vibe-backend" + ], + "additional_details": { + "raw_benchmark_id": "vibe-backend", + "raw_model_id": "minimax-m2.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VIBE benchmark subset for backend service generation", + "metric_id": "llm_stats.vibe-backend.score", + "metric_name": "VIBE Backend score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "vibe-backend", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VIBE Backend", + "raw_categories": "[\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.867, + "details": { + "raw_score": "0.867", + "raw_score_field": "score", + "raw_model_id": "minimax-m2.1", + "raw_benchmark_id": "vibe-backend", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m2.1\",\"https://llm-stats.com/benchmarks/vibe-backend\",\"https://api.llm-stats.com/leaderboard/benchmarks/vibe-backend\"]", + "raw_score_id": "vibe-backend::minimax-m2.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "vibe-ios::vibe-ios-minimax-m2.1", + "evaluation_name": "llm_stats.vibe-ios", + "source_data": { + "dataset_name": "VIBE iOS", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m2.1", + "https://llm-stats.com/benchmarks/vibe-ios", + "https://api.llm-stats.com/leaderboard/benchmarks/vibe-ios" + ], + "additional_details": { + "raw_benchmark_id": "vibe-ios", + "raw_model_id": "minimax-m2.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VIBE benchmark subset for iOS application generation", + "metric_id": "llm_stats.vibe-ios.score", + "metric_name": "VIBE iOS score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "vibe-ios", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VIBE iOS", + "raw_categories": "[\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.88, + "details": { + "raw_score": "0.88", + "raw_score_field": "score", + "raw_model_id": "minimax-m2.1", + "raw_benchmark_id": "vibe-ios", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m2.1\",\"https://llm-stats.com/benchmarks/vibe-ios\",\"https://api.llm-stats.com/leaderboard/benchmarks/vibe-ios\"]", + "raw_score_id": "vibe-ios::minimax-m2.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "vibe-simulation::vibe-simulation-minimax-m2.1", + "evaluation_name": "llm_stats.vibe-simulation", + "source_data": { + "dataset_name": "VIBE Simulation", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m2.1", + "https://llm-stats.com/benchmarks/vibe-simulation", + "https://api.llm-stats.com/leaderboard/benchmarks/vibe-simulation" + ], + "additional_details": { + "raw_benchmark_id": "vibe-simulation", + "raw_model_id": "minimax-m2.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VIBE benchmark subset for simulation code generation", + "metric_id": "llm_stats.vibe-simulation.score", + "metric_name": "VIBE Simulation score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "vibe-simulation", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VIBE Simulation", + "raw_categories": "[\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.871, + "details": { + "raw_score": "0.871", + "raw_score_field": "score", + "raw_model_id": "minimax-m2.1", + "raw_benchmark_id": "vibe-simulation", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m2.1\",\"https://llm-stats.com/benchmarks/vibe-simulation\",\"https://api.llm-stats.com/leaderboard/benchmarks/vibe-simulation\"]", + "raw_score_id": "vibe-simulation::minimax-m2.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "vibe-web::vibe-web-minimax-m2.1", + "evaluation_name": "llm_stats.vibe-web", + "source_data": { + "dataset_name": "VIBE Web", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m2.1", + "https://llm-stats.com/benchmarks/vibe-web", + "https://api.llm-stats.com/leaderboard/benchmarks/vibe-web" + ], + "additional_details": { + "raw_benchmark_id": "vibe-web", + "raw_model_id": "minimax-m2.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VIBE benchmark subset for web application generation", + "metric_id": "llm_stats.vibe-web.score", + "metric_name": "VIBE Web score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "vibe-web", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VIBE Web", + "raw_categories": "[\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.915, + "details": { + "raw_score": "0.915", + "raw_score_field": "score", + "raw_model_id": "minimax-m2.1", + "raw_benchmark_id": "vibe-web", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m2.1\",\"https://llm-stats.com/benchmarks/vibe-web\",\"https://api.llm-stats.com/leaderboard/benchmarks/vibe-web\"]", + "raw_score_id": "vibe-web::minimax-m2.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "vibe::vibe-minimax-m2.1", + "evaluation_name": "llm_stats.vibe", + "source_data": { + "dataset_name": "VIBE", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m2.1", + "https://llm-stats.com/benchmarks/vibe", + "https://api.llm-stats.com/leaderboard/benchmarks/vibe" + ], + "additional_details": { + "raw_benchmark_id": "vibe", + "raw_model_id": "minimax-m2.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Visual Interface Building Evaluation benchmark for UI/app generation", + "metric_id": "llm_stats.vibe.score", + "metric_name": "VIBE score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "vibe", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VIBE", + "raw_categories": "[\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.886, + "details": { + "raw_score": "0.886", + "raw_score_field": "score", + "raw_model_id": "minimax-m2.1", + "raw_benchmark_id": "vibe", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m2.1\",\"https://llm-stats.com/benchmarks/vibe\",\"https://api.llm-stats.com/leaderboard/benchmarks/vibe\"]", + "raw_score_id": "vibe::minimax-m2.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/minimax/minimax-m2.5/44e599be-8f6b-45c3-9588-c92c01cd17e0.json b/data/llm_stats/minimax/minimax-m2.5/44e599be-8f6b-45c3-9588-c92c01cd17e0.json new file mode 100644 index 0000000000000000000000000000000000000000..5a321acc6aa1f2dae76f1209ee8b6144a4bd2c51 --- /dev/null +++ b/data/llm_stats/minimax/minimax-m2.5/44e599be-8f6b-45c3-9588-c92c01cd17e0.json @@ -0,0 +1,412 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/minimax_minimax-m2.5/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "minimax-m2.5", + "id": "minimax/minimax-m2.5", + "developer": "minimax", + "additional_details": { + "raw_id": "minimax-m2.5", + "raw_model_name": "minimax-m2.5", + "raw_display_name": "MiniMax M2.5", + "raw_organization_id": "minimax", + "raw_organization_name": "MiniMax", + "raw_context_length": "1000000", + "raw_input_modalities": "[\"image\",\"text\"]", + "raw_output_modalities": "[\"text\"]", + "raw_input_price": "0.3", + "raw_output_price": "1.2", + "raw_model_id": "minimax-m2.5", + "raw_provider_slug": "minimax", + "raw_provider_name": "ZeroEval" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "bfcl-v3-multiturn::bfcl-v3-multiturn-minimax-m2.5", + "evaluation_name": "llm_stats.bfcl-v3-multiturn", + "source_data": { + "dataset_name": "BFCL_v3_MultiTurn", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m2.5", + "https://llm-stats.com/benchmarks/bfcl-v3-multiturn", + "https://api.llm-stats.com/leaderboard/benchmarks/bfcl-v3-multiturn" + ], + "additional_details": { + "raw_benchmark_id": "bfcl-v3-multiturn", + "raw_model_id": "minimax-m2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Berkeley Function Calling Leaderboard (BFCL) V3 MultiTurn benchmark that evaluates large language models' ability to handle multi-turn and multi-step function calling scenarios. The benchmark introduces complex interactions requiring models to manage sequential function calls, handle conversational context across multiple turns, and make dynamic decisions about when and how to use available functions. BFCL V3 uses state-based evaluation by verifying the actual state of API systems after function execution, providing more realistic assessment of function calling capabilities in agentic applications.", + "metric_id": "llm_stats.bfcl-v3-multiturn.score", + "metric_name": "BFCL_v3_MultiTurn score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bfcl-v3-multiturn", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BFCL_v3_MultiTurn", + "raw_categories": "[\"reasoning\",\"general\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.768, + "details": { + "raw_score": "0.768", + "raw_score_field": "score", + "raw_model_id": "minimax-m2.5", + "raw_benchmark_id": "bfcl-v3-multiturn", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m2.5\",\"https://llm-stats.com/benchmarks/bfcl-v3-multiturn\",\"https://api.llm-stats.com/leaderboard/benchmarks/bfcl-v3-multiturn\"]", + "raw_score_id": "bfcl-v3-multiturn::minimax-m2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "browsecomp::browsecomp-minimax-m2.5", + "evaluation_name": "llm_stats.browsecomp", + "source_data": { + "dataset_name": "BrowseComp", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m2.5", + "https://llm-stats.com/benchmarks/browsecomp", + "https://api.llm-stats.com/leaderboard/benchmarks/browsecomp" + ], + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_model_id": "minimax-m2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BrowseComp is a benchmark comprising 1,266 questions that challenge AI agents to persistently navigate the internet in search of hard-to-find, entangled information. The benchmark measures agents' ability to exercise persistence in information gathering, demonstrate creativity in web navigation, and find concise, verifiable answers. Despite the difficulty of the questions, BrowseComp is simple and easy-to-use, as predicted answers are short and easily verifiable against reference answers.", + "metric_id": "llm_stats.browsecomp.score", + "metric_name": "BrowseComp score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BrowseComp", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "52" + } + }, + "score_details": { + "score": 0.763, + "details": { + "raw_score": "0.763", + "raw_score_field": "score", + "raw_model_id": "minimax-m2.5", + "raw_benchmark_id": "browsecomp", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m2.5\",\"https://llm-stats.com/benchmarks/browsecomp\",\"https://api.llm-stats.com/leaderboard/benchmarks/browsecomp\"]", + "raw_score_id": "browsecomp::minimax-m2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gdpval-mm::gdpval-mm-minimax-m2.5", + "evaluation_name": "llm_stats.gdpval-mm", + "source_data": { + "dataset_name": "GDPval-MM", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m2.5", + "https://llm-stats.com/benchmarks/gdpval-mm", + "https://api.llm-stats.com/leaderboard/benchmarks/gdpval-mm" + ], + "additional_details": { + "raw_benchmark_id": "gdpval-mm", + "raw_model_id": "minimax-m2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "GDPval-MM is the multimodal variant of the GDPval benchmark, evaluating AI model performance on real-world economically valuable tasks that require processing and generating multimodal content including documents, slides, diagrams, spreadsheets, images, and other professional deliverables across diverse industries.", + "metric_id": "llm_stats.gdpval-mm.score", + "metric_name": "GDPval-MM score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gdpval-mm", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GDPval-MM", + "raw_categories": "[\"multimodal\",\"reasoning\",\"finance\",\"general\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.59, + "details": { + "raw_score": "0.59", + "raw_score_field": "score", + "raw_model_id": "minimax-m2.5", + "raw_benchmark_id": "gdpval-mm", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m2.5\",\"https://llm-stats.com/benchmarks/gdpval-mm\",\"https://api.llm-stats.com/leaderboard/benchmarks/gdpval-mm\"]", + "raw_score_id": "gdpval-mm::minimax-m2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mewc::mewc-minimax-m2.5", + "evaluation_name": "llm_stats.mewc", + "source_data": { + "dataset_name": "MEWC", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m2.5", + "https://llm-stats.com/benchmarks/mewc", + "https://api.llm-stats.com/leaderboard/benchmarks/mewc" + ], + "additional_details": { + "raw_benchmark_id": "mewc", + "raw_model_id": "minimax-m2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MEWC is a benchmark that evaluates AI model performance on multi-environment web challenges, testing agents' ability to navigate and complete complex tasks across diverse web environments.", + "metric_id": "llm_stats.mewc.score", + "metric_name": "MEWC score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mewc", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MEWC", + "raw_categories": "[\"reasoning\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.744, + "details": { + "raw_score": "0.744", + "raw_score_field": "score", + "raw_model_id": "minimax-m2.5", + "raw_benchmark_id": "mewc", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m2.5\",\"https://llm-stats.com/benchmarks/mewc\",\"https://api.llm-stats.com/leaderboard/benchmarks/mewc\"]", + "raw_score_id": "mewc::minimax-m2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "multi-swe-bench::multi-swe-bench-minimax-m2.5", + "evaluation_name": "llm_stats.multi-swe-bench", + "source_data": { + "dataset_name": "Multi-SWE-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m2.5", + "https://llm-stats.com/benchmarks/multi-swe-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/multi-swe-bench" + ], + "additional_details": { + "raw_benchmark_id": "multi-swe-bench", + "raw_model_id": "minimax-m2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multilingual benchmark for issue resolving that evaluates Large Language Models' ability to resolve software issues across diverse programming ecosystems. Covers 7 programming languages (Java, TypeScript, JavaScript, Go, Rust, C, and C++) with 1,632 high-quality instances carefully annotated by 68 expert annotators. Addresses limitations of existing benchmarks that focus almost exclusively on Python.", + "metric_id": "llm_stats.multi-swe-bench.score", + "metric_name": "Multi-SWE-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "multi-swe-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Multi-SWE-Bench", + "raw_categories": "[\"reasoning\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.513, + "details": { + "raw_score": "0.513", + "raw_score_field": "score", + "raw_model_id": "minimax-m2.5", + "raw_benchmark_id": "multi-swe-bench", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m2.5\",\"https://llm-stats.com/benchmarks/multi-swe-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/multi-swe-bench\"]", + "raw_score_id": "multi-swe-bench::minimax-m2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-verified::swe-bench-verified-minimax-m2.5", + "evaluation_name": "llm_stats.swe-bench-verified", + "source_data": { + "dataset_name": "SWE-Bench Verified", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m2.5", + "https://llm-stats.com/benchmarks/swe-bench-verified", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-verified" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-verified", + "raw_model_id": "minimax-m2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A verified subset of 500 software engineering problems from real GitHub issues, validated by human annotators for evaluating language models' ability to resolve real-world coding issues by generating patches for Python codebases.", + "metric_id": "llm_stats.swe-bench-verified.score", + "metric_name": "SWE-Bench Verified score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-verified", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Bench Verified", + "raw_categories": "[\"reasoning\",\"frontend_development\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "103" + } + }, + "score_details": { + "score": 0.802, + "details": { + "raw_score": "0.802", + "raw_score_field": "score", + "raw_model_id": "minimax-m2.5", + "raw_benchmark_id": "swe-bench-verified", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m2.5\",\"https://llm-stats.com/benchmarks/swe-bench-verified\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-verified\"]", + "raw_score_id": "swe-bench-verified::minimax-m2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "vibe-pro::vibe-pro-minimax-m2.5", + "evaluation_name": "llm_stats.vibe-pro", + "source_data": { + "dataset_name": "VIBE-Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m2.5", + "https://llm-stats.com/benchmarks/vibe-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/vibe-pro" + ], + "additional_details": { + "raw_benchmark_id": "vibe-pro", + "raw_model_id": "minimax-m2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VIBE-Pro is an advanced version of the VIBE (Visual & Interactive Benchmark for Execution) benchmark that evaluates LLMs on professional-grade full-stack application development tasks. It measures model performance across complex real-world development scenarios including web, mobile, and backend applications with higher difficulty than the standard VIBE benchmark.", + "metric_id": "llm_stats.vibe-pro.score", + "metric_name": "VIBE-Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "vibe-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VIBE-Pro", + "raw_categories": "[\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.542, + "details": { + "raw_score": "0.542", + "raw_score_field": "score", + "raw_model_id": "minimax-m2.5", + "raw_benchmark_id": "vibe-pro", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m2.5\",\"https://llm-stats.com/benchmarks/vibe-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/vibe-pro\"]", + "raw_score_id": "vibe-pro::minimax-m2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/minimax/minimax-m3/887875da-f4f8-4045-acf4-b0fe8c312069.json b/data/llm_stats/minimax/minimax-m3/887875da-f4f8-4045-acf4-b0fe8c312069.json new file mode 100644 index 0000000000000000000000000000000000000000..43d146ecb1e821915cda9236cc946d6be4f3f283 --- /dev/null +++ b/data/llm_stats/minimax/minimax-m3/887875da-f4f8-4045-acf4-b0fe8c312069.json @@ -0,0 +1,1712 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/minimax_minimax-m3/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "minimax-m3", + "id": "minimax/minimax-m3", + "developer": "minimax", + "additional_details": { + "raw_id": "minimax-m3", + "raw_model_name": "minimax-m3", + "raw_display_name": "MiniMax M3", + "raw_organization_id": "minimax", + "raw_organization_name": "MiniMax", + "raw_context_length": "1000000", + "raw_input_modalities": "[\"image\",\"text\",\"video\"]", + "raw_output_modalities": "[\"text\"]", + "raw_input_price": "0.3", + "raw_output_price": "1.2", + "raw_model_id": "minimax-m3", + "raw_provider_slug": "minimax", + "raw_provider_name": "ZeroEval" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "apex-agents::apex-agents-minimax-m3", + "evaluation_name": "llm_stats.apex-agents", + "source_data": { + "dataset_name": "APEX-Agents", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/apex-agents", + "https://api.llm-stats.com/leaderboard/benchmarks/apex-agents" + ], + "additional_details": { + "raw_benchmark_id": "apex-agents", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "APEX-Agents is a benchmark evaluating AI agents on long horizon professional tasks that require sustained reasoning, planning, and execution across complex multi-step workflows.", + "metric_id": "llm_stats.apex-agents.score", + "metric_name": "APEX-Agents score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "apex-agents", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "APEX-Agents", + "raw_categories": "[\"reasoning\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.277, + "details": { + "raw_score": "0.277", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "apex-agents", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/apex-agents\",\"https://api.llm-stats.com/leaderboard/benchmarks/apex-agents\"]", + "raw_score_id": "apex-agents::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bankertoolbench::bankertoolbench-minimax-m3", + "evaluation_name": "llm_stats.bankertoolbench", + "source_data": { + "dataset_name": "BankerToolBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/bankertoolbench", + "https://api.llm-stats.com/leaderboard/benchmarks/bankertoolbench" + ], + "additional_details": { + "raw_benchmark_id": "bankertoolbench", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BankerToolBench is a public benchmark that evaluates models on banking and finance tool-use tasks. Models are scored against dataset rubrics, measuring their ability to correctly invoke tools and complete multi-step financial workflows.", + "metric_id": "llm_stats.bankertoolbench.score", + "metric_name": "BankerToolBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bankertoolbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BankerToolBench", + "raw_categories": "[\"finance\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.7612, + "details": { + "raw_score": "0.7612", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "bankertoolbench", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/bankertoolbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/bankertoolbench\"]", + "raw_score_id": "bankertoolbench::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "browsecomp::browsecomp-minimax-m3", + "evaluation_name": "llm_stats.browsecomp", + "source_data": { + "dataset_name": "BrowseComp", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/browsecomp", + "https://api.llm-stats.com/leaderboard/benchmarks/browsecomp" + ], + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BrowseComp is a benchmark comprising 1,266 questions that challenge AI agents to persistently navigate the internet in search of hard-to-find, entangled information. The benchmark measures agents' ability to exercise persistence in information gathering, demonstrate creativity in web navigation, and find concise, verifiable answers. Despite the difficulty of the questions, BrowseComp is simple and easy-to-use, as predicted answers are short and easily verifiable against reference answers.", + "metric_id": "llm_stats.browsecomp.score", + "metric_name": "BrowseComp score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BrowseComp", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "52" + } + }, + "score_details": { + "score": 0.8352, + "details": { + "raw_score": "0.8352", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "browsecomp", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/browsecomp\",\"https://api.llm-stats.com/leaderboard/benchmarks/browsecomp\"]", + "raw_score_id": "browsecomp::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "cl-bench::cl-bench-minimax-m3", + "evaluation_name": "llm_stats.cl-bench", + "source_data": { + "dataset_name": "CL-bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/cl-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/cl-bench" + ], + "additional_details": { + "raw_benchmark_id": "cl-bench", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CL-bench is an open-source benchmark with its own data and rubrics for evaluating models on coding and agentic tasks, scored using a setup fully aligned with the official procedure.", + "metric_id": "llm_stats.cl-bench.score", + "metric_name": "CL-bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "cl-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CL-bench", + "raw_categories": "[\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.2048, + "details": { + "raw_score": "0.2048", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "cl-bench", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/cl-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/cl-bench\"]", + "raw_score_id": "cl-bench::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "claw-eval::claw-eval-minimax-m3", + "evaluation_name": "llm_stats.claw-eval", + "source_data": { + "dataset_name": "Claw-Eval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/claw-eval", + "https://api.llm-stats.com/leaderboard/benchmarks/claw-eval" + ], + "additional_details": { + "raw_benchmark_id": "claw-eval", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Claw-Eval tests real-world agentic task completion across complex multi-step scenarios, evaluating a model's ability to use tools, navigate environments, and complete end-to-end tasks autonomously.", + "metric_id": "llm_stats.claw-eval.score", + "metric_name": "Claw-Eval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "claw-eval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Claw-Eval", + "raw_categories": "[\"agents\",\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.745, + "details": { + "raw_score": "0.745", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "claw-eval", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/claw-eval\",\"https://api.llm-stats.com/leaderboard/benchmarks/claw-eval\"]", + "raw_score_id": "claw-eval::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "draco::draco-minimax-m3", + "evaluation_name": "llm_stats.draco", + "source_data": { + "dataset_name": "DRACO", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/draco", + "https://api.llm-stats.com/leaderboard/benchmarks/draco" + ], + "additional_details": { + "raw_benchmark_id": "draco", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "DRACO is a deep research benchmark that evaluates an agent's ability to gather, synthesize, and reason over information to answer complex research questions. Scores are based on official rubrics per question, with the final score being the average across all questions.", + "metric_id": "llm_stats.draco.score", + "metric_name": "DRACO score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "draco", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DRACO", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.7323, + "details": { + "raw_score": "0.7323", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "draco", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/draco\",\"https://api.llm-stats.com/leaderboard/benchmarks/draco\"]", + "raw_score_id": "draco::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gdpval-rubrics::gdpval-rubrics-minimax-m3", + "evaluation_name": "llm_stats.gdpval-rubrics", + "source_data": { + "dataset_name": "GDPval-Rubrics", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/gdpval-rubrics", + "https://api.llm-stats.com/leaderboard/benchmarks/gdpval-rubrics" + ], + "additional_details": { + "raw_benchmark_id": "gdpval-rubrics", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "GDPval-Rubrics evaluates AI model performance on economically valuable knowledge work tasks drawn from the public GDPval dataset. It uses pointwise scoring based on public rubrics, with the environment aligned to the GDPval-AA scaffolding.", + "metric_id": "llm_stats.gdpval-rubrics.score", + "metric_name": "GDPval-Rubrics score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gdpval-rubrics", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GDPval-Rubrics", + "raw_categories": "[\"legal\",\"reasoning\",\"finance\",\"general\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.7478, + "details": { + "raw_score": "0.7478", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "gdpval-rubrics", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/gdpval-rubrics\",\"https://api.llm-stats.com/leaderboard/benchmarks/gdpval-rubrics\"]", + "raw_score_id": "gdpval-rubrics::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "imo-2025::imo-2025-minimax-m3", + "evaluation_name": "llm_stats.imo-2025", + "source_data": { + "dataset_name": "IMO 2025", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/imo-2025", + "https://api.llm-stats.com/leaderboard/benchmarks/imo-2025" + ], + "additional_details": { + "raw_benchmark_id": "imo-2025", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "IMO 2025 evaluates models on the six problems from the 2025 International Mathematical Olympiad, requiring rigorous proof-based reasoning. Following the MathArena methodology, proofs are graded against human expert rubrics by dual strong judge models, with the minimum of the two scores taken as final. Maximum score is 42 points.", + "metric_id": "llm_stats.imo-2025.score", + "metric_name": "IMO 2025 score", + "metric_kind": "benchmark_score", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "raw_benchmark_id": "imo-2025", + "raw_score_field": "score", + "bound_strategy": "inferred_percent", + "raw_name": "IMO 2025", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 35.0, + "details": { + "raw_score": "35.0", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "imo-2025", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/imo-2025\",\"https://api.llm-stats.com/leaderboard/benchmarks/imo-2025\"]", + "raw_score_id": "imo-2025::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "kernelbench-hard::kernelbench-hard-minimax-m3", + "evaluation_name": "llm_stats.kernelbench-hard", + "source_data": { + "dataset_name": "KernelBench Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/kernelbench-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/kernelbench-hard" + ], + "additional_details": { + "raw_benchmark_id": "kernelbench-hard", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "KernelBench Hard evaluates agentic GPU kernel optimization on the hardest problem set. Each question is scored by the agent's submitted operator TFLOPs relative to the theoretical peak of the current hardware, with the benchmark score being the average across all questions.", + "metric_id": "llm_stats.kernelbench-hard.score", + "metric_name": "KernelBench Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "kernelbench-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "KernelBench Hard", + "raw_categories": "[\"agents\",\"code\",\"systems\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.288, + "details": { + "raw_score": "0.288", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "kernelbench-hard", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/kernelbench-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/kernelbench-hard\"]", + "raw_score_id": "kernelbench-hard::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "livesqlbench::livesqlbench-minimax-m3", + "evaluation_name": "llm_stats.livesqlbench", + "source_data": { + "dataset_name": "LiveSQLBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/livesqlbench", + "https://api.llm-stats.com/leaderboard/benchmarks/livesqlbench" + ], + "additional_details": { + "raw_benchmark_id": "livesqlbench", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LiveSQLBench evaluates models on generating correct SQL queries against live PostgreSQL databases. The LiveSQLBench-Base-Full v1 dataset contains 600 questions across 22 PostgreSQL databases, testing real-world database reasoning and query generation.", + "metric_id": "llm_stats.livesqlbench.score", + "metric_name": "LiveSQLBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "livesqlbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LiveSQLBench", + "raw_categories": "[\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.4017, + "details": { + "raw_score": "0.4017", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "livesqlbench", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/livesqlbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/livesqlbench\"]", + "raw_score_id": "livesqlbench::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "loca-bench-256k::loca-bench-256k-minimax-m3", + "evaluation_name": "llm_stats.loca-bench-256k", + "source_data": { + "dataset_name": "LOCA-Bench (256k)", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/loca-bench-256k", + "https://api.llm-stats.com/leaderboard/benchmarks/loca-bench-256k" + ], + "additional_details": { + "raw_benchmark_id": "loca-bench-256k", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LOCA-Bench is a long-context agentic benchmark. The 256k variant evaluates agents using the official ReAct mode with an environment description length of 256k tokens, measuring how well models reason and act over very long contexts.", + "metric_id": "llm_stats.loca-bench-256k.score", + "metric_name": "LOCA-Bench (256k) score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "loca-bench-256k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LOCA-Bench (256k)", + "raw_categories": "[\"reasoning\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.493, + "details": { + "raw_score": "0.493", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "loca-bench-256k", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/loca-bench-256k\",\"https://api.llm-stats.com/leaderboard/benchmarks/loca-bench-256k\"]", + "raw_score_id": "loca-bench-256k::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mcp-atlas::mcp-atlas-minimax-m3", + "evaluation_name": "llm_stats.mcp-atlas", + "source_data": { + "dataset_name": "MCP Atlas", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/mcp-atlas", + "https://api.llm-stats.com/leaderboard/benchmarks/mcp-atlas" + ], + "additional_details": { + "raw_benchmark_id": "mcp-atlas", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MCP Atlas is a benchmark for evaluating AI models on scaled tool use capabilities, measuring how well models can coordinate and utilize multiple tools across complex multi-step tasks.", + "metric_id": "llm_stats.mcp-atlas.score", + "metric_name": "MCP Atlas score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mcp-atlas", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MCP Atlas", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.742, + "details": { + "raw_score": "0.742", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "mcp-atlas", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/mcp-atlas\",\"https://api.llm-stats.com/leaderboard/benchmarks/mcp-atlas\"]", + "raw_score_id": "mcp-atlas::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmmu-pro::mmmu-pro-minimax-m3", + "evaluation_name": "llm_stats.mmmu-pro", + "source_data": { + "dataset_name": "MMMU-Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/mmmu-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/mmmu-pro" + ], + "additional_details": { + "raw_benchmark_id": "mmmu-pro", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A more robust multi-discipline multimodal understanding benchmark that enhances MMMU through a three-step process: filtering text-only answerable questions, augmenting candidate options, and introducing vision-only input settings. Achieves significantly lower model performance (16.8-26.9%) compared to original MMMU, providing more rigorous evaluation that closely mimics real-world scenarios.", + "metric_id": "llm_stats.mmmu-pro.score", + "metric_name": "MMMU-Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmmu-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMMU-Pro", + "raw_categories": "[\"multimodal\",\"reasoning\",\"general\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "60" + } + }, + "score_details": { + "score": 0.781, + "details": { + "raw_score": "0.781", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "mmmu-pro", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/mmmu-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmmu-pro\"]", + "raw_score_id": "mmmu-pro::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "nl2repo::nl2repo-minimax-m3", + "evaluation_name": "llm_stats.nl2repo", + "source_data": { + "dataset_name": "NL2Repo", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/nl2repo", + "https://api.llm-stats.com/leaderboard/benchmarks/nl2repo" + ], + "additional_details": { + "raw_benchmark_id": "nl2repo", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "NL2Repo evaluates long-horizon coding capabilities including repository-level understanding, where models must generate or modify code across entire repositories from natural language specifications.", + "metric_id": "llm_stats.nl2repo.score", + "metric_name": "NL2Repo score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "nl2repo", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "NL2Repo", + "raw_categories": "[\"agents\",\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.4213, + "details": { + "raw_score": "0.4213", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "nl2repo", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/nl2repo\",\"https://api.llm-stats.com/leaderboard/benchmarks/nl2repo\"]", + "raw_score_id": "nl2repo::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "officeqa-pro::officeqa-pro-minimax-m3", + "evaluation_name": "llm_stats.officeqa-pro", + "source_data": { + "dataset_name": "OfficeQA Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/officeqa-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/officeqa-pro" + ], + "additional_details": { + "raw_benchmark_id": "officeqa-pro", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OfficeQA Pro evaluates AI models on professional knowledge-work questions and tasks drawn from real office workflows, including document analysis, spreadsheet reasoning, and information synthesis across business domains.", + "metric_id": "llm_stats.officeqa-pro.score", + "metric_name": "OfficeQA Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "officeqa-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OfficeQA Pro", + "raw_categories": "[\"reasoning\",\"general\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.451, + "details": { + "raw_score": "0.451", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "officeqa-pro", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/officeqa-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/officeqa-pro\"]", + "raw_score_id": "officeqa-pro::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "omnidocbench-1.5::omnidocbench-1.5-minimax-m3", + "evaluation_name": "llm_stats.omnidocbench-1.5", + "source_data": { + "dataset_name": "OmniDocBench 1.5", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/omnidocbench-1.5", + "https://api.llm-stats.com/leaderboard/benchmarks/omnidocbench-1.5" + ], + "additional_details": { + "raw_benchmark_id": "omnidocbench-1.5", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OmniDocBench 1.5 is a comprehensive benchmark for evaluating multimodal large language models on document understanding tasks, including OCR, document parsing, information extraction, and visual question answering across diverse document types. Lower Overall Edit Distance scores are better.", + "metric_id": "llm_stats.omnidocbench-1.5.score", + "metric_name": "OmniDocBench 1.5 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "omnidocbench-1.5", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OmniDocBench 1.5", + "raw_categories": "[\"multimodal\",\"reasoning\",\"structured_output\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.916, + "details": { + "raw_score": "0.916", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "omnidocbench-1.5", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/omnidocbench-1.5\",\"https://api.llm-stats.com/leaderboard/benchmarks/omnidocbench-1.5\"]", + "raw_score_id": "omnidocbench-1.5::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "osworld-verified::osworld-verified-minimax-m3", + "evaluation_name": "llm_stats.osworld-verified", + "source_data": { + "dataset_name": "OSWorld-Verified", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/osworld-verified", + "https://api.llm-stats.com/leaderboard/benchmarks/osworld-verified" + ], + "additional_details": { + "raw_benchmark_id": "osworld-verified", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OSWorld-Verified is a verified subset of OSWorld, a scalable real computer environment for multimodal agents supporting task setup, execution-based evaluation, and interactive learning across Ubuntu, Windows, and macOS.", + "metric_id": "llm_stats.osworld-verified.score", + "metric_name": "OSWorld-Verified score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "osworld-verified", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OSWorld-Verified", + "raw_categories": "[\"multimodal\",\"general\",\"agents\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.7006, + "details": { + "raw_score": "0.7006", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "osworld-verified", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/osworld-verified\",\"https://api.llm-stats.com/leaderboard/benchmarks/osworld-verified\"]", + "raw_score_id": "osworld-verified::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "paperbench::paperbench-minimax-m3", + "evaluation_name": "llm_stats.paperbench", + "source_data": { + "dataset_name": "PaperBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/paperbench", + "https://api.llm-stats.com/leaderboard/benchmarks/paperbench" + ], + "additional_details": { + "raw_benchmark_id": "paperbench", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "PaperBench is a benchmark for evaluating AI agents on their ability to replicate research papers. It tests models on complex, multi-step workflows involving code implementation, experimentation, and reproducing scientific results from academic publications.", + "metric_id": "llm_stats.paperbench.score", + "metric_name": "PaperBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "paperbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PaperBench", + "raw_categories": "[\"reasoning\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.526, + "details": { + "raw_score": "0.526", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "paperbench", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/paperbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/paperbench\"]", + "raw_score_id": "paperbench::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "posttrainbench::posttrainbench-minimax-m3", + "evaluation_name": "llm_stats.posttrainbench", + "source_data": { + "dataset_name": "PostTrainBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/posttrainbench", + "https://api.llm-stats.com/leaderboard/benchmarks/posttrainbench" + ], + "additional_details": { + "raw_benchmark_id": "posttrainbench", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "PostTrainBench evaluates a model's ability to autonomously post-train base models. Given pretrain-only base models, the agent must complete the full pipeline of data synthesis, training, evaluation, and iteration within a time budget, scored across downstream benchmarks such as AIME2025, BFCL, GPQA Main, GSM8K, and HumanEval.", + "metric_id": "llm_stats.posttrainbench.score", + "metric_name": "PostTrainBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "posttrainbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PostTrainBench", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"systems\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.371, + "details": { + "raw_score": "0.371", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "posttrainbench", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/posttrainbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/posttrainbench\"]", + "raw_score_id": "posttrainbench::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "spreadsheetbench-v1::spreadsheetbench-v1-minimax-m3", + "evaluation_name": "llm_stats.spreadsheetbench-v1", + "source_data": { + "dataset_name": "SpreadSheetBench-v1", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/spreadsheetbench-v1", + "https://api.llm-stats.com/leaderboard/benchmarks/spreadsheetbench-v1" + ], + "additional_details": { + "raw_benchmark_id": "spreadsheetbench-v1", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SpreadSheetBench-v1 evaluates office automation agents on spreadsheet reasoning and manipulation tasks, measuring the ability to analyze, transform, and operate on spreadsheet data through tools.", + "metric_id": "llm_stats.spreadsheetbench-v1.score", + "metric_name": "SpreadSheetBench-v1 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "spreadsheetbench-v1", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SpreadSheetBench-v1", + "raw_categories": "[\"productivity\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.8935, + "details": { + "raw_score": "0.8935", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "spreadsheetbench-v1", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/spreadsheetbench-v1\",\"https://api.llm-stats.com/leaderboard/benchmarks/spreadsheetbench-v1\"]", + "raw_score_id": "spreadsheetbench-v1::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "svg-bench::svg-bench-minimax-m3", + "evaluation_name": "llm_stats.svg-bench", + "source_data": { + "dataset_name": "SVG-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/svg-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/svg-bench" + ], + "additional_details": { + "raw_benchmark_id": "svg-bench", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SVG-Bench is an internal benchmark that comprehensively evaluates SVG generation performance. It accepts text and image inputs across build-from-scratch and edit-based tasks, using a VLM to verify rendering accuracy of the generated outputs.", + "metric_id": "llm_stats.svg-bench.score", + "metric_name": "SVG-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "svg-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SVG-Bench", + "raw_categories": "[\"multimodal\",\"agents\",\"code\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.637, + "details": { + "raw_score": "0.637", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "svg-bench", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/svg-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/svg-bench\"]", + "raw_score_id": "svg-bench::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-atlas-codebase-qna::swe-atlas-codebase-qna-minimax-m3", + "evaluation_name": "llm_stats.swe-atlas-codebase-qna", + "source_data": { + "dataset_name": "SWE Atlas - Codebase QnA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/swe-atlas-codebase-qna", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-atlas-codebase-qna" + ], + "additional_details": { + "raw_benchmark_id": "swe-atlas-codebase-qna", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SWE Atlas - Codebase QnA evaluates a model's ability to answer questions about real codebases, measuring repository-level comprehension and the ability to reason about code structure, behavior, and intent across an entire project.", + "metric_id": "llm_stats.swe-atlas-codebase-qna.score", + "metric_name": "SWE Atlas - Codebase QnA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-atlas-codebase-qna", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE Atlas - Codebase QnA", + "raw_categories": "[\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.379, + "details": { + "raw_score": "0.379", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "swe-atlas-codebase-qna", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/swe-atlas-codebase-qna\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-atlas-codebase-qna\"]", + "raw_score_id": "swe-atlas-codebase-qna::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-atlas-test-writing::swe-atlas-test-writing-minimax-m3", + "evaluation_name": "llm_stats.swe-atlas-test-writing", + "source_data": { + "dataset_name": "SWE Atlas - Test Writing", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/swe-atlas-test-writing", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-atlas-test-writing" + ], + "additional_details": { + "raw_benchmark_id": "swe-atlas-test-writing", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SWE Atlas - Test Writing evaluates a model's ability to author meaningful tests for real-world software projects, measuring how well agents can understand code and produce correct, useful test coverage.", + "metric_id": "llm_stats.swe-atlas-test-writing.score", + "metric_name": "SWE Atlas - Test Writing score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-atlas-test-writing", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE Atlas - Test Writing", + "raw_categories": "[\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.3083, + "details": { + "raw_score": "0.3083", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "swe-atlas-test-writing", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/swe-atlas-test-writing\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-atlas-test-writing\"]", + "raw_score_id": "swe-atlas-test-writing::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-pro::swe-bench-pro-minimax-m3", + "evaluation_name": "llm_stats.swe-bench-pro", + "source_data": { + "dataset_name": "SWE-Bench Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/swe-bench-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-pro" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-pro", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SWE-Bench Pro is an advanced version of SWE-Bench that evaluates language models on complex, real-world software engineering tasks requiring extended reasoning and multi-step problem solving.", + "metric_id": "llm_stats.swe-bench-pro.score", + "metric_name": "SWE-Bench Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Bench Pro", + "raw_categories": "[\"reasoning\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "35" + } + }, + "score_details": { + "score": 0.59, + "details": { + "raw_score": "0.59", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "swe-bench-pro", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/swe-bench-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-pro\"]", + "raw_score_id": "swe-bench-pro::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-verified::swe-bench-verified-minimax-m3", + "evaluation_name": "llm_stats.swe-bench-verified", + "source_data": { + "dataset_name": "SWE-Bench Verified", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/swe-bench-verified", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-verified" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-verified", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A verified subset of 500 software engineering problems from real GitHub issues, validated by human annotators for evaluating language models' ability to resolve real-world coding issues by generating patches for Python codebases.", + "metric_id": "llm_stats.swe-bench-verified.score", + "metric_name": "SWE-Bench Verified score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-verified", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Bench Verified", + "raw_categories": "[\"reasoning\",\"frontend_development\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "103" + } + }, + "score_details": { + "score": 0.805, + "details": { + "raw_score": "0.805", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "swe-bench-verified", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/swe-bench-verified\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-verified\"]", + "raw_score_id": "swe-bench-verified::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-fficiency::swe-fficiency-minimax-m3", + "evaluation_name": "llm_stats.swe-fficiency", + "source_data": { + "dataset_name": "SWE-fficiency", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/swe-fficiency", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-fficiency" + ], + "additional_details": { + "raw_benchmark_id": "swe-fficiency", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SWE-fficiency is an open-source benchmark and workflow that evaluates language models on optimizing the runtime efficiency of real-world software engineering tasks, measuring how well agents can improve code performance autonomously.", + "metric_id": "llm_stats.swe-fficiency.score", + "metric_name": "SWE-fficiency score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-fficiency", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-fficiency", + "raw_categories": "[\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.348, + "details": { + "raw_score": "0.348", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "swe-fficiency", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/swe-fficiency\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-fficiency\"]", + "raw_score_id": "swe-fficiency::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2.1::terminal-bench-2.1-minimax-m3", + "evaluation_name": "llm_stats.terminal-bench-2.1", + "source_data": { + "dataset_name": "Terminal-Bench 2.1", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/terminal-bench-2.1", + "https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2.1" + ], + "additional_details": { + "raw_benchmark_id": "terminal-bench-2.1", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Terminal-Bench 2.1 is an updated release of the Terminal-Bench benchmark that tests AI agents' ability to operate a computer via the terminal. It evaluates how well models handle real-world, end-to-end tasks autonomously, including compiling code, training models, setting up servers, system administration, data science workflows, and security tasks.", + "metric_id": "llm_stats.terminal-bench-2.1.score", + "metric_name": "Terminal-Bench 2.1 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "terminal-bench-2.1", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Terminal-Bench 2.1", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.66, + "details": { + "raw_score": "0.66", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "terminal-bench-2.1", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/terminal-bench-2.1\",\"https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2.1\"]", + "raw_score_id": "terminal-bench-2.1::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "usamo-2026::usamo-2026-minimax-m3", + "evaluation_name": "llm_stats.usamo-2026", + "source_data": { + "dataset_name": "USAMO 2026", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/usamo-2026", + "https://api.llm-stats.com/leaderboard/benchmarks/usamo-2026" + ], + "additional_details": { + "raw_benchmark_id": "usamo-2026", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "USAMO 2026 evaluates models on the six problems from the 2026 United States of America Mathematical Olympiad, requiring rigorous proof-based reasoning. Following the MathArena methodology, proofs are graded against human expert rubrics by dual strong judge models, with the minimum of the two scores taken as final. Maximum score is 42 points.", + "metric_id": "llm_stats.usamo-2026.score", + "metric_name": "USAMO 2026 score", + "metric_kind": "benchmark_score", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "raw_benchmark_id": "usamo-2026", + "raw_score_field": "score", + "bound_strategy": "inferred_percent", + "raw_name": "USAMO 2026", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 36.0, + "details": { + "raw_score": "36.0", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "usamo-2026", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/usamo-2026\",\"https://api.llm-stats.com/leaderboard/benchmarks/usamo-2026\"]", + "raw_score_id": "usamo-2026::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "vibe-v2::vibe-v2-minimax-m3", + "evaluation_name": "llm_stats.vibe-v2", + "source_data": { + "dataset_name": "VIBE-V2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/vibe-v2", + "https://api.llm-stats.com/leaderboard/benchmarks/vibe-v2" + ], + "additional_details": { + "raw_benchmark_id": "vibe-v2", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VIBE-V2 is an internal benchmark covering pure front-end and full-stack Web, Android, and iOS projects with build-from-scratch tasks. It uses an Agent-as-a-Verifier paradigm to automatically verify program interaction logic and visual output, scoring models through a unified pipeline that includes a requirement set, containerized deployment, and a dynamic interaction environment.", + "metric_id": "llm_stats.vibe-v2.score", + "metric_name": "VIBE-V2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "vibe-v2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VIBE-V2", + "raw_categories": "[\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.5012, + "details": { + "raw_score": "0.5012", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "vibe-v2", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/vibe-v2\",\"https://api.llm-stats.com/leaderboard/benchmarks/vibe-v2\"]", + "raw_score_id": "vibe-v2::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "video-mme::video-mme-minimax-m3", + "evaluation_name": "llm_stats.video-mme", + "source_data": { + "dataset_name": "Video-MME", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/video-mme", + "https://api.llm-stats.com/leaderboard/benchmarks/video-mme" + ], + "additional_details": { + "raw_benchmark_id": "video-mme", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Video-MME is the first-ever comprehensive evaluation benchmark of Multi-modal Large Language Models (MLLMs) in video analysis. It features 900 videos totaling 254 hours with 2,700 human-annotated question-answer pairs across 6 primary visual domains (Knowledge, Film & Television, Sports Competition, Life Record, Multilingual, and others) and 30 subfields. The benchmark evaluates models across diverse temporal dimensions (11 seconds to 1 hour), integrates multi-modal inputs including video frames, subtitles, and audio, and uses rigorous manual labeling by expert annotators for precise assessment.", + "metric_id": "llm_stats.video-mme.score", + "metric_name": "Video-MME score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "video-mme", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Video-MME", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "17" + } + }, + "score_details": { + "score": 0.854, + "details": { + "raw_score": "0.854", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "video-mme", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/video-mme\",\"https://api.llm-stats.com/leaderboard/benchmarks/video-mme\"]", + "raw_score_id": "video-mme::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "videommmu::videommmu-minimax-m3", + "evaluation_name": "llm_stats.videommmu", + "source_data": { + "dataset_name": "VideoMMMU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/videommmu", + "https://api.llm-stats.com/leaderboard/benchmarks/videommmu" + ], + "additional_details": { + "raw_benchmark_id": "videommmu", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Video-MMMU evaluates Large Multimodal Models' ability to acquire knowledge from expert-level professional videos across six disciplines through three cognitive stages: perception, comprehension, and adaptation. Contains 300 videos and 900 human-annotated questions spanning Art, Business, Science, Medicine, Humanities, and Engineering.", + "metric_id": "llm_stats.videommmu.score", + "metric_name": "VideoMMMU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "videommmu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VideoMMMU", + "raw_categories": "[\"multimodal\",\"reasoning\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "26" + } + }, + "score_details": { + "score": 0.846, + "details": { + "raw_score": "0.846", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "videommmu", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/videommmu\",\"https://api.llm-stats.com/leaderboard/benchmarks/videommmu\"]", + "raw_score_id": "videommmu::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "yc-bench::yc-bench-minimax-m3", + "evaluation_name": "llm_stats.yc-bench", + "source_data": { + "dataset_name": "YC-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minimax-m3", + "https://llm-stats.com/benchmarks/yc-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/yc-bench" + ], + "additional_details": { + "raw_benchmark_id": "yc-bench", + "raw_model_id": "minimax-m3", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "YC-Bench evaluates agents on long-horizon, open-ended business and investment decision-making. The reported metric is the final assets (fund value, in US dollars) accumulated by the agent over the course of the simulation.", + "metric_id": "llm_stats.yc-bench.score", + "metric_name": "YC-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "points", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 2100000.0, + "additional_details": { + "raw_benchmark_id": "yc-bench", + "raw_score_field": "score", + "bound_strategy": "observed_value", + "raw_name": "YC-Bench", + "raw_categories": "[\"finance\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 2100000.0, + "details": { + "raw_score": "2100000.0", + "raw_score_field": "score", + "raw_model_id": "minimax-m3", + "raw_benchmark_id": "yc-bench", + "source_urls_json": "[\"https://llm-stats.com/models/minimax-m3\",\"https://llm-stats.com/benchmarks/yc-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/yc-bench\"]", + "raw_score_id": "yc-bench::minimax-m3", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/mistral/ministral-3-14b-base-2512/3c4d020c-72ee-42ca-8377-702997133804.json b/data/llm_stats/mistral/ministral-3-14b-base-2512/3c4d020c-72ee-42ca-8377-702997133804.json new file mode 100644 index 0000000000000000000000000000000000000000..a858b35fdf4ae6a31dfdf01ae3ca9b46a38dc62b --- /dev/null +++ b/data/llm_stats/mistral/ministral-3-14b-base-2512/3c4d020c-72ee-42ca-8377-702997133804.json @@ -0,0 +1,255 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/mistral_ministral-3-14b-base-2512/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Ministral 3 (14B Base 2512)", + "id": "mistral/ministral-3-14b-base-2512", + "developer": "mistral", + "additional_details": { + "raw_id": "ministral-3-14b-base-2512", + "raw_name": "Ministral 3 (14B Base 2512)", + "raw_model_id": "ministral-3-14b-base-2512", + "raw_model_name": "Ministral 3 (14B Base 2512)", + "raw_organization_id": "mistral", + "raw_organization_name": "Mistral AI", + "raw_release_date": "2025-12-04", + "raw_announcement_date": "2025-12-04", + "raw_multimodal": "true", + "raw_param_count": "14000000000", + "raw_provider_slug": "mistral", + "raw_provider_name": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "agieval::agieval-ministral-3-14b-base-2512", + "evaluation_name": "llm_stats.agieval", + "source_data": { + "dataset_name": "AGIEval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/ministral-3-14b-base-2512", + "https://llm-stats.com/benchmarks/agieval", + "https://api.llm-stats.com/leaderboard/benchmarks/agieval" + ], + "additional_details": { + "raw_benchmark_id": "agieval", + "raw_model_id": "ministral-3-14b-base-2512", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A human-centric benchmark for evaluating foundation models on standardized exams including college entrance exams (Gaokao, SAT), law school admission tests (LSAT), math competitions, lawyer qualification tests, and civil service exams. Contains 20 tasks (18 multiple-choice, 2 cloze) designed to assess understanding, knowledge, reasoning, and calculation abilities in real-world academic and professional contexts.", + "metric_id": "llm_stats.agieval.score", + "metric_name": "AGIEval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "agieval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AGIEval", + "raw_categories": "[\"legal\",\"math\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.648, + "details": { + "raw_score": "0.648", + "raw_score_field": "score", + "raw_model_id": "ministral-3-14b-base-2512", + "raw_benchmark_id": "agieval", + "source_urls_json": "[\"https://llm-stats.com/models/ministral-3-14b-base-2512\",\"https://llm-stats.com/benchmarks/agieval\",\"https://api.llm-stats.com/leaderboard/benchmarks/agieval\"]", + "raw_score_id": "agieval::ministral-3-14b-base-2512", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "math-cot::math-cot-ministral-3-14b-base-2512", + "evaluation_name": "llm_stats.math-cot", + "source_data": { + "dataset_name": "MATH (CoT)", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/ministral-3-14b-base-2512", + "https://llm-stats.com/benchmarks/math-cot", + "https://api.llm-stats.com/leaderboard/benchmarks/math-(cot)" + ], + "additional_details": { + "raw_benchmark_id": "math-(cot)", + "raw_model_id": "ministral-3-14b-base-2512", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MATH dataset contains 12,500 challenging competition mathematics problems from AMC 10, AMC 12, AIME, and other mathematics competitions. Each problem includes full step-by-step solutions and spans multiple difficulty levels (1-5) across seven mathematical subjects. This variant uses Chain-of-Thought prompting to encourage step-by-step reasoning.", + "metric_id": "llm_stats.math-cot.score", + "metric_name": "MATH (CoT) score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "math-(cot)", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MATH (CoT)", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.676, + "details": { + "raw_score": "0.676", + "raw_score_field": "score", + "raw_model_id": "ministral-3-14b-base-2512", + "raw_benchmark_id": "math-(cot)", + "source_urls_json": "[\"https://llm-stats.com/models/ministral-3-14b-base-2512\",\"https://llm-stats.com/benchmarks/math-cot\",\"https://api.llm-stats.com/leaderboard/benchmarks/math-(cot)\"]", + "raw_score_id": "math-(cot)::ministral-3-14b-base-2512", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "multilingual-mmlu::multilingual-mmlu-ministral-3-14b-base-2512", + "evaluation_name": "llm_stats.multilingual-mmlu", + "source_data": { + "dataset_name": "Multilingual MMLU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/ministral-3-14b-base-2512", + "https://llm-stats.com/benchmarks/multilingual-mmlu", + "https://api.llm-stats.com/leaderboard/benchmarks/multilingual-mmlu" + ], + "additional_details": { + "raw_benchmark_id": "multilingual-mmlu", + "raw_model_id": "ministral-3-14b-base-2512", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MMLU-ProX is a comprehensive multilingual benchmark covering 29 typologically diverse languages, building upon MMLU-Pro. Each language version consists of 11,829 identical questions enabling direct cross-linguistic comparisons. The benchmark evaluates large language models' reasoning capabilities across linguistic and cultural boundaries through challenging, reasoning-focused questions with 10 answer choices.", + "metric_id": "llm_stats.multilingual-mmlu.score", + "metric_name": "Multilingual MMLU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "multilingual-mmlu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Multilingual MMLU", + "raw_categories": "[\"language\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.742, + "details": { + "raw_score": "0.742", + "raw_score_field": "score", + "raw_model_id": "ministral-3-14b-base-2512", + "raw_benchmark_id": "multilingual-mmlu", + "source_urls_json": "[\"https://llm-stats.com/models/ministral-3-14b-base-2512\",\"https://llm-stats.com/benchmarks/multilingual-mmlu\",\"https://api.llm-stats.com/leaderboard/benchmarks/multilingual-mmlu\"]", + "raw_score_id": "multilingual-mmlu::ministral-3-14b-base-2512", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "triviaqa::triviaqa-ministral-3-14b-base-2512", + "evaluation_name": "llm_stats.triviaqa", + "source_data": { + "dataset_name": "TriviaQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/ministral-3-14b-base-2512", + "https://llm-stats.com/benchmarks/triviaqa", + "https://api.llm-stats.com/leaderboard/benchmarks/triviaqa" + ], + "additional_details": { + "raw_benchmark_id": "triviaqa", + "raw_model_id": "ministral-3-14b-base-2512", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A large-scale reading comprehension dataset containing over 650K question-answer-evidence triples. TriviaQA includes 95K question-answer pairs authored by trivia enthusiasts and independently gathered evidence documents (six per question on average) that provide high quality distant supervision for answering the questions. The dataset features relatively complex, compositional questions with considerable syntactic and lexical variability, requiring cross-sentence reasoning to find answers.", + "metric_id": "llm_stats.triviaqa.score", + "metric_name": "TriviaQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "triviaqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TriviaQA", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.749, + "details": { + "raw_score": "0.749", + "raw_score_field": "score", + "raw_model_id": "ministral-3-14b-base-2512", + "raw_benchmark_id": "triviaqa", + "source_urls_json": "[\"https://llm-stats.com/models/ministral-3-14b-base-2512\",\"https://llm-stats.com/benchmarks/triviaqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/triviaqa\"]", + "raw_score_id": "triviaqa::ministral-3-14b-base-2512", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/mistral/ministral-3-3b-base-2512/9ea5b8ac-d845-405b-909a-2ba15e1664be.json b/data/llm_stats/mistral/ministral-3-3b-base-2512/9ea5b8ac-d845-405b-909a-2ba15e1664be.json new file mode 100644 index 0000000000000000000000000000000000000000..2719d9e960a170a0155095e30dc365376ef5d346 --- /dev/null +++ b/data/llm_stats/mistral/ministral-3-3b-base-2512/9ea5b8ac-d845-405b-909a-2ba15e1664be.json @@ -0,0 +1,255 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/mistral_ministral-3-3b-base-2512/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Ministral 3 (3B Base 2512)", + "id": "mistral/ministral-3-3b-base-2512", + "developer": "mistral", + "additional_details": { + "raw_id": "ministral-3-3b-base-2512", + "raw_name": "Ministral 3 (3B Base 2512)", + "raw_model_id": "ministral-3-3b-base-2512", + "raw_model_name": "Ministral 3 (3B Base 2512)", + "raw_organization_id": "mistral", + "raw_organization_name": "Mistral AI", + "raw_release_date": "2025-12-04", + "raw_announcement_date": "2025-12-04", + "raw_multimodal": "true", + "raw_param_count": "3000000000", + "raw_provider_slug": "mistral", + "raw_provider_name": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "agieval::agieval-ministral-3-3b-base-2512", + "evaluation_name": "llm_stats.agieval", + "source_data": { + "dataset_name": "AGIEval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/ministral-3-3b-base-2512", + "https://llm-stats.com/benchmarks/agieval", + "https://api.llm-stats.com/leaderboard/benchmarks/agieval" + ], + "additional_details": { + "raw_benchmark_id": "agieval", + "raw_model_id": "ministral-3-3b-base-2512", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A human-centric benchmark for evaluating foundation models on standardized exams including college entrance exams (Gaokao, SAT), law school admission tests (LSAT), math competitions, lawyer qualification tests, and civil service exams. Contains 20 tasks (18 multiple-choice, 2 cloze) designed to assess understanding, knowledge, reasoning, and calculation abilities in real-world academic and professional contexts.", + "metric_id": "llm_stats.agieval.score", + "metric_name": "AGIEval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "agieval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AGIEval", + "raw_categories": "[\"legal\",\"math\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.511, + "details": { + "raw_score": "0.511", + "raw_score_field": "score", + "raw_model_id": "ministral-3-3b-base-2512", + "raw_benchmark_id": "agieval", + "source_urls_json": "[\"https://llm-stats.com/models/ministral-3-3b-base-2512\",\"https://llm-stats.com/benchmarks/agieval\",\"https://api.llm-stats.com/leaderboard/benchmarks/agieval\"]", + "raw_score_id": "agieval::ministral-3-3b-base-2512", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "math-cot::math-cot-ministral-3-3b-base-2512", + "evaluation_name": "llm_stats.math-cot", + "source_data": { + "dataset_name": "MATH (CoT)", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/ministral-3-3b-base-2512", + "https://llm-stats.com/benchmarks/math-cot", + "https://api.llm-stats.com/leaderboard/benchmarks/math-(cot)" + ], + "additional_details": { + "raw_benchmark_id": "math-(cot)", + "raw_model_id": "ministral-3-3b-base-2512", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MATH dataset contains 12,500 challenging competition mathematics problems from AMC 10, AMC 12, AIME, and other mathematics competitions. Each problem includes full step-by-step solutions and spans multiple difficulty levels (1-5) across seven mathematical subjects. This variant uses Chain-of-Thought prompting to encourage step-by-step reasoning.", + "metric_id": "llm_stats.math-cot.score", + "metric_name": "MATH (CoT) score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "math-(cot)", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MATH (CoT)", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.601, + "details": { + "raw_score": "0.601", + "raw_score_field": "score", + "raw_model_id": "ministral-3-3b-base-2512", + "raw_benchmark_id": "math-(cot)", + "source_urls_json": "[\"https://llm-stats.com/models/ministral-3-3b-base-2512\",\"https://llm-stats.com/benchmarks/math-cot\",\"https://api.llm-stats.com/leaderboard/benchmarks/math-(cot)\"]", + "raw_score_id": "math-(cot)::ministral-3-3b-base-2512", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "multilingual-mmlu::multilingual-mmlu-ministral-3-3b-base-2512", + "evaluation_name": "llm_stats.multilingual-mmlu", + "source_data": { + "dataset_name": "Multilingual MMLU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/ministral-3-3b-base-2512", + "https://llm-stats.com/benchmarks/multilingual-mmlu", + "https://api.llm-stats.com/leaderboard/benchmarks/multilingual-mmlu" + ], + "additional_details": { + "raw_benchmark_id": "multilingual-mmlu", + "raw_model_id": "ministral-3-3b-base-2512", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MMLU-ProX is a comprehensive multilingual benchmark covering 29 typologically diverse languages, building upon MMLU-Pro. Each language version consists of 11,829 identical questions enabling direct cross-linguistic comparisons. The benchmark evaluates large language models' reasoning capabilities across linguistic and cultural boundaries through challenging, reasoning-focused questions with 10 answer choices.", + "metric_id": "llm_stats.multilingual-mmlu.score", + "metric_name": "Multilingual MMLU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "multilingual-mmlu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Multilingual MMLU", + "raw_categories": "[\"language\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.652, + "details": { + "raw_score": "0.652", + "raw_score_field": "score", + "raw_model_id": "ministral-3-3b-base-2512", + "raw_benchmark_id": "multilingual-mmlu", + "source_urls_json": "[\"https://llm-stats.com/models/ministral-3-3b-base-2512\",\"https://llm-stats.com/benchmarks/multilingual-mmlu\",\"https://api.llm-stats.com/leaderboard/benchmarks/multilingual-mmlu\"]", + "raw_score_id": "multilingual-mmlu::ministral-3-3b-base-2512", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "triviaqa::triviaqa-ministral-3-3b-base-2512", + "evaluation_name": "llm_stats.triviaqa", + "source_data": { + "dataset_name": "TriviaQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/ministral-3-3b-base-2512", + "https://llm-stats.com/benchmarks/triviaqa", + "https://api.llm-stats.com/leaderboard/benchmarks/triviaqa" + ], + "additional_details": { + "raw_benchmark_id": "triviaqa", + "raw_model_id": "ministral-3-3b-base-2512", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A large-scale reading comprehension dataset containing over 650K question-answer-evidence triples. TriviaQA includes 95K question-answer pairs authored by trivia enthusiasts and independently gathered evidence documents (six per question on average) that provide high quality distant supervision for answering the questions. The dataset features relatively complex, compositional questions with considerable syntactic and lexical variability, requiring cross-sentence reasoning to find answers.", + "metric_id": "llm_stats.triviaqa.score", + "metric_name": "TriviaQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "triviaqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TriviaQA", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.592, + "details": { + "raw_score": "0.592", + "raw_score_field": "score", + "raw_model_id": "ministral-3-3b-base-2512", + "raw_benchmark_id": "triviaqa", + "source_urls_json": "[\"https://llm-stats.com/models/ministral-3-3b-base-2512\",\"https://llm-stats.com/benchmarks/triviaqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/triviaqa\"]", + "raw_score_id": "triviaqa::ministral-3-3b-base-2512", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/mistral/ministral-3-8b-base-2512/a71d950f-f9a4-4caa-9dfd-3ec17fc30cf6.json b/data/llm_stats/mistral/ministral-3-8b-base-2512/a71d950f-f9a4-4caa-9dfd-3ec17fc30cf6.json new file mode 100644 index 0000000000000000000000000000000000000000..59ff0623e6c72988afaf3da2725bc87c4fddf9ee --- /dev/null +++ b/data/llm_stats/mistral/ministral-3-8b-base-2512/a71d950f-f9a4-4caa-9dfd-3ec17fc30cf6.json @@ -0,0 +1,255 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/mistral_ministral-3-8b-base-2512/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Ministral 3 (8B Base 2512)", + "id": "mistral/ministral-3-8b-base-2512", + "developer": "mistral", + "additional_details": { + "raw_id": "ministral-3-8b-base-2512", + "raw_name": "Ministral 3 (8B Base 2512)", + "raw_model_id": "ministral-3-8b-base-2512", + "raw_model_name": "Ministral 3 (8B Base 2512)", + "raw_organization_id": "mistral", + "raw_organization_name": "Mistral AI", + "raw_release_date": "2025-12-04", + "raw_announcement_date": "2025-12-04", + "raw_multimodal": "true", + "raw_param_count": "8000000000", + "raw_provider_slug": "mistral", + "raw_provider_name": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "agieval::agieval-ministral-3-8b-base-2512", + "evaluation_name": "llm_stats.agieval", + "source_data": { + "dataset_name": "AGIEval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/ministral-3-8b-base-2512", + "https://llm-stats.com/benchmarks/agieval", + "https://api.llm-stats.com/leaderboard/benchmarks/agieval" + ], + "additional_details": { + "raw_benchmark_id": "agieval", + "raw_model_id": "ministral-3-8b-base-2512", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A human-centric benchmark for evaluating foundation models on standardized exams including college entrance exams (Gaokao, SAT), law school admission tests (LSAT), math competitions, lawyer qualification tests, and civil service exams. Contains 20 tasks (18 multiple-choice, 2 cloze) designed to assess understanding, knowledge, reasoning, and calculation abilities in real-world academic and professional contexts.", + "metric_id": "llm_stats.agieval.score", + "metric_name": "AGIEval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "agieval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AGIEval", + "raw_categories": "[\"legal\",\"math\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.591, + "details": { + "raw_score": "0.591", + "raw_score_field": "score", + "raw_model_id": "ministral-3-8b-base-2512", + "raw_benchmark_id": "agieval", + "source_urls_json": "[\"https://llm-stats.com/models/ministral-3-8b-base-2512\",\"https://llm-stats.com/benchmarks/agieval\",\"https://api.llm-stats.com/leaderboard/benchmarks/agieval\"]", + "raw_score_id": "agieval::ministral-3-8b-base-2512", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "math-cot::math-cot-ministral-3-8b-base-2512", + "evaluation_name": "llm_stats.math-cot", + "source_data": { + "dataset_name": "MATH (CoT)", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/ministral-3-8b-base-2512", + "https://llm-stats.com/benchmarks/math-cot", + "https://api.llm-stats.com/leaderboard/benchmarks/math-(cot)" + ], + "additional_details": { + "raw_benchmark_id": "math-(cot)", + "raw_model_id": "ministral-3-8b-base-2512", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MATH dataset contains 12,500 challenging competition mathematics problems from AMC 10, AMC 12, AIME, and other mathematics competitions. Each problem includes full step-by-step solutions and spans multiple difficulty levels (1-5) across seven mathematical subjects. This variant uses Chain-of-Thought prompting to encourage step-by-step reasoning.", + "metric_id": "llm_stats.math-cot.score", + "metric_name": "MATH (CoT) score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "math-(cot)", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MATH (CoT)", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.626, + "details": { + "raw_score": "0.626", + "raw_score_field": "score", + "raw_model_id": "ministral-3-8b-base-2512", + "raw_benchmark_id": "math-(cot)", + "source_urls_json": "[\"https://llm-stats.com/models/ministral-3-8b-base-2512\",\"https://llm-stats.com/benchmarks/math-cot\",\"https://api.llm-stats.com/leaderboard/benchmarks/math-(cot)\"]", + "raw_score_id": "math-(cot)::ministral-3-8b-base-2512", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "multilingual-mmlu::multilingual-mmlu-ministral-3-8b-base-2512", + "evaluation_name": "llm_stats.multilingual-mmlu", + "source_data": { + "dataset_name": "Multilingual MMLU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/ministral-3-8b-base-2512", + "https://llm-stats.com/benchmarks/multilingual-mmlu", + "https://api.llm-stats.com/leaderboard/benchmarks/multilingual-mmlu" + ], + "additional_details": { + "raw_benchmark_id": "multilingual-mmlu", + "raw_model_id": "ministral-3-8b-base-2512", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MMLU-ProX is a comprehensive multilingual benchmark covering 29 typologically diverse languages, building upon MMLU-Pro. Each language version consists of 11,829 identical questions enabling direct cross-linguistic comparisons. The benchmark evaluates large language models' reasoning capabilities across linguistic and cultural boundaries through challenging, reasoning-focused questions with 10 answer choices.", + "metric_id": "llm_stats.multilingual-mmlu.score", + "metric_name": "Multilingual MMLU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "multilingual-mmlu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Multilingual MMLU", + "raw_categories": "[\"language\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.706, + "details": { + "raw_score": "0.706", + "raw_score_field": "score", + "raw_model_id": "ministral-3-8b-base-2512", + "raw_benchmark_id": "multilingual-mmlu", + "source_urls_json": "[\"https://llm-stats.com/models/ministral-3-8b-base-2512\",\"https://llm-stats.com/benchmarks/multilingual-mmlu\",\"https://api.llm-stats.com/leaderboard/benchmarks/multilingual-mmlu\"]", + "raw_score_id": "multilingual-mmlu::ministral-3-8b-base-2512", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "triviaqa::triviaqa-ministral-3-8b-base-2512", + "evaluation_name": "llm_stats.triviaqa", + "source_data": { + "dataset_name": "TriviaQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/ministral-3-8b-base-2512", + "https://llm-stats.com/benchmarks/triviaqa", + "https://api.llm-stats.com/leaderboard/benchmarks/triviaqa" + ], + "additional_details": { + "raw_benchmark_id": "triviaqa", + "raw_model_id": "ministral-3-8b-base-2512", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A large-scale reading comprehension dataset containing over 650K question-answer-evidence triples. TriviaQA includes 95K question-answer pairs authored by trivia enthusiasts and independently gathered evidence documents (six per question on average) that provide high quality distant supervision for answering the questions. The dataset features relatively complex, compositional questions with considerable syntactic and lexical variability, requiring cross-sentence reasoning to find answers.", + "metric_id": "llm_stats.triviaqa.score", + "metric_name": "TriviaQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "triviaqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TriviaQA", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.681, + "details": { + "raw_score": "0.681", + "raw_score_field": "score", + "raw_model_id": "ministral-3-8b-base-2512", + "raw_benchmark_id": "triviaqa", + "source_urls_json": "[\"https://llm-stats.com/models/ministral-3-8b-base-2512\",\"https://llm-stats.com/benchmarks/triviaqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/triviaqa\"]", + "raw_score_id": "triviaqa::ministral-3-8b-base-2512", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/mistral/ministral-8b-instruct-2410/cd95419b-4b95-4a1e-ad13-d0ad34c22c33.json b/data/llm_stats/mistral/ministral-8b-instruct-2410/cd95419b-4b95-4a1e-ad13-d0ad34c22c33.json new file mode 100644 index 0000000000000000000000000000000000000000..f984a041983136b570b0ebc7b296c2090f8f2e6e --- /dev/null +++ b/data/llm_stats/mistral/ministral-8b-instruct-2410/cd95419b-4b95-4a1e-ad13-d0ad34c22c33.json @@ -0,0 +1,463 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/mistral_ministral-8b-instruct-2410/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Ministral 8B Instruct", + "id": "mistral/ministral-8b-instruct-2410", + "developer": "mistral", + "additional_details": { + "raw_id": "ministral-8b-instruct-2410", + "raw_name": "Ministral 8B Instruct", + "raw_model_id": "ministral-8b-instruct-2410", + "raw_model_name": "Ministral 8B Instruct", + "raw_organization_id": "mistral", + "raw_organization_name": "Mistral AI", + "raw_release_date": "2024-10-16", + "raw_announcement_date": "2024-10-16", + "raw_multimodal": "false", + "raw_param_count": "8019808256", + "raw_provider_slug": "mistral", + "raw_provider_name": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "agieval::agieval-ministral-8b-instruct-2410", + "evaluation_name": "llm_stats.agieval", + "source_data": { + "dataset_name": "AGIEval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/ministral-8b-instruct-2410", + "https://llm-stats.com/benchmarks/agieval", + "https://api.llm-stats.com/leaderboard/benchmarks/agieval" + ], + "additional_details": { + "raw_benchmark_id": "agieval", + "raw_model_id": "ministral-8b-instruct-2410", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A human-centric benchmark for evaluating foundation models on standardized exams including college entrance exams (Gaokao, SAT), law school admission tests (LSAT), math competitions, lawyer qualification tests, and civil service exams. Contains 20 tasks (18 multiple-choice, 2 cloze) designed to assess understanding, knowledge, reasoning, and calculation abilities in real-world academic and professional contexts.", + "metric_id": "llm_stats.agieval.score", + "metric_name": "AGIEval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "agieval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AGIEval", + "raw_categories": "[\"legal\",\"math\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.483, + "details": { + "raw_score": "0.483", + "raw_score_field": "score", + "raw_model_id": "ministral-8b-instruct-2410", + "raw_benchmark_id": "agieval", + "source_urls_json": "[\"https://llm-stats.com/models/ministral-8b-instruct-2410\",\"https://llm-stats.com/benchmarks/agieval\",\"https://api.llm-stats.com/leaderboard/benchmarks/agieval\"]", + "raw_score_id": "agieval::ministral-8b-instruct-2410", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "arc-c::arc-c-ministral-8b-instruct-2410", + "evaluation_name": "llm_stats.arc-c", + "source_data": { + "dataset_name": "ARC-C", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/ministral-8b-instruct-2410", + "https://llm-stats.com/benchmarks/arc-c", + "https://api.llm-stats.com/leaderboard/benchmarks/arc-c" + ], + "additional_details": { + "raw_benchmark_id": "arc-c", + "raw_model_id": "ministral-8b-instruct-2410", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The AI2 Reasoning Challenge (ARC) Challenge Set is a multiple-choice question-answering benchmark containing grade-school level science questions that require advanced reasoning capabilities. ARC-C specifically contains questions that were answered incorrectly by both retrieval-based and word co-occurrence algorithms, making it a particularly challenging subset designed to test commonsense reasoning abilities in AI systems.", + "metric_id": "llm_stats.arc-c.score", + "metric_name": "ARC-C score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arc-c", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ARC-C", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "34" + } + }, + "score_details": { + "score": 0.719, + "details": { + "raw_score": "0.719", + "raw_score_field": "score", + "raw_model_id": "ministral-8b-instruct-2410", + "raw_benchmark_id": "arc-c", + "source_urls_json": "[\"https://llm-stats.com/models/ministral-8b-instruct-2410\",\"https://llm-stats.com/benchmarks/arc-c\",\"https://api.llm-stats.com/leaderboard/benchmarks/arc-c\"]", + "raw_score_id": "arc-c::ministral-8b-instruct-2410", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "arena-hard::arena-hard-ministral-8b-instruct-2410", + "evaluation_name": "llm_stats.arena-hard", + "source_data": { + "dataset_name": "Arena Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/ministral-8b-instruct-2410", + "https://llm-stats.com/benchmarks/arena-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/arena-hard" + ], + "additional_details": { + "raw_benchmark_id": "arena-hard", + "raw_model_id": "ministral-8b-instruct-2410", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Arena-Hard-Auto is an automatic evaluation benchmark for instruction-tuned LLMs consisting of 500 challenging real-world prompts curated by BenchBuilder. It includes open-ended software engineering problems, mathematical questions, and creative writing tasks. The benchmark uses LLM-as-a-Judge methodology with GPT-4.1 and Gemini-2.5 as automatic judges to approximate human preference. Arena-Hard achieves 98.6% correlation with human preference rankings and provides 3x higher separation of model performances compared to MT-Bench, making it highly effective for distinguishing between models of similar quality.", + "metric_id": "llm_stats.arena-hard.score", + "metric_name": "Arena Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arena-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Arena Hard", + "raw_categories": "[\"reasoning\",\"general\",\"creativity\",\"writing\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "26" + } + }, + "score_details": { + "score": 0.709, + "details": { + "raw_score": "0.709", + "raw_score_field": "score", + "raw_model_id": "ministral-8b-instruct-2410", + "raw_benchmark_id": "arena-hard", + "source_urls_json": "[\"https://llm-stats.com/models/ministral-8b-instruct-2410\",\"https://llm-stats.com/benchmarks/arena-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/arena-hard\"]", + "raw_score_id": "arena-hard::ministral-8b-instruct-2410", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "french-mmlu::french-mmlu-ministral-8b-instruct-2410", + "evaluation_name": "llm_stats.french-mmlu", + "source_data": { + "dataset_name": "French MMLU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/ministral-8b-instruct-2410", + "https://llm-stats.com/benchmarks/french-mmlu", + "https://api.llm-stats.com/leaderboard/benchmarks/french-mmlu" + ], + "additional_details": { + "raw_benchmark_id": "french-mmlu", + "raw_model_id": "ministral-8b-instruct-2410", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "French version of MMLU-Pro, a multilingual benchmark for evaluating language models' cross-lingual reasoning capabilities across 14 diverse domains including mathematics, physics, chemistry, law, engineering, psychology, and health.", + "metric_id": "llm_stats.french-mmlu.score", + "metric_name": "French MMLU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "french-mmlu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "French MMLU", + "raw_categories": "[\"language\",\"legal\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.575, + "details": { + "raw_score": "0.575", + "raw_score_field": "score", + "raw_model_id": "ministral-8b-instruct-2410", + "raw_benchmark_id": "french-mmlu", + "source_urls_json": "[\"https://llm-stats.com/models/ministral-8b-instruct-2410\",\"https://llm-stats.com/benchmarks/french-mmlu\",\"https://api.llm-stats.com/leaderboard/benchmarks/french-mmlu\"]", + "raw_score_id": "french-mmlu::ministral-8b-instruct-2410", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mbpp-pass-1::mbpp-pass-1-ministral-8b-instruct-2410", + "evaluation_name": "llm_stats.mbpp-pass-1", + "source_data": { + "dataset_name": "MBPP pass@1", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/ministral-8b-instruct-2410", + "https://llm-stats.com/benchmarks/mbpp-pass-1", + "https://api.llm-stats.com/leaderboard/benchmarks/mbpp-pass@1" + ], + "additional_details": { + "raw_benchmark_id": "mbpp-pass@1", + "raw_model_id": "ministral-8b-instruct-2410", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MBPP (Mostly Basic Python Problems) is a benchmark of 974 crowd-sourced Python programming problems designed to be solvable by entry-level programmers. Each problem consists of a task description, code solution, and 3 automated test cases. This variant uses pass@1 evaluation metric measuring the percentage of problems solved correctly on the first attempt.", + "metric_id": "llm_stats.mbpp-pass-1.score", + "metric_name": "MBPP pass@1 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mbpp-pass@1", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MBPP pass@1", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.7, + "details": { + "raw_score": "0.7", + "raw_score_field": "score", + "raw_model_id": "ministral-8b-instruct-2410", + "raw_benchmark_id": "mbpp-pass@1", + "source_urls_json": "[\"https://llm-stats.com/models/ministral-8b-instruct-2410\",\"https://llm-stats.com/benchmarks/mbpp-pass-1\",\"https://api.llm-stats.com/leaderboard/benchmarks/mbpp-pass@1\"]", + "raw_score_id": "mbpp-pass@1::ministral-8b-instruct-2410", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mt-bench::mt-bench-ministral-8b-instruct-2410", + "evaluation_name": "llm_stats.mt-bench", + "source_data": { + "dataset_name": "MT-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/ministral-8b-instruct-2410", + "https://llm-stats.com/benchmarks/mt-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/mt-bench" + ], + "additional_details": { + "raw_benchmark_id": "mt-bench", + "raw_model_id": "ministral-8b-instruct-2410", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MT-Bench is a challenging multi-turn benchmark that measures the ability of large language models to engage in coherent, informative, and engaging conversations. It uses strong LLMs as judges for scalable and explainable evaluation of multi-turn dialogue capabilities.", + "metric_id": "llm_stats.mt-bench.score", + "metric_name": "MT-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mt-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MT-Bench", + "raw_categories": "[\"reasoning\",\"roleplay\",\"general\",\"communication\",\"creativity\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.83, + "details": { + "raw_score": "0.83", + "raw_score_field": "score", + "raw_model_id": "ministral-8b-instruct-2410", + "raw_benchmark_id": "mt-bench", + "source_urls_json": "[\"https://llm-stats.com/models/ministral-8b-instruct-2410\",\"https://llm-stats.com/benchmarks/mt-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/mt-bench\"]", + "raw_score_id": "mt-bench::ministral-8b-instruct-2410", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "triviaqa::triviaqa-ministral-8b-instruct-2410", + "evaluation_name": "llm_stats.triviaqa", + "source_data": { + "dataset_name": "TriviaQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/ministral-8b-instruct-2410", + "https://llm-stats.com/benchmarks/triviaqa", + "https://api.llm-stats.com/leaderboard/benchmarks/triviaqa" + ], + "additional_details": { + "raw_benchmark_id": "triviaqa", + "raw_model_id": "ministral-8b-instruct-2410", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A large-scale reading comprehension dataset containing over 650K question-answer-evidence triples. TriviaQA includes 95K question-answer pairs authored by trivia enthusiasts and independently gathered evidence documents (six per question on average) that provide high quality distant supervision for answering the questions. The dataset features relatively complex, compositional questions with considerable syntactic and lexical variability, requiring cross-sentence reasoning to find answers.", + "metric_id": "llm_stats.triviaqa.score", + "metric_name": "TriviaQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "triviaqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TriviaQA", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.655, + "details": { + "raw_score": "0.655", + "raw_score_field": "score", + "raw_model_id": "ministral-8b-instruct-2410", + "raw_benchmark_id": "triviaqa", + "source_urls_json": "[\"https://llm-stats.com/models/ministral-8b-instruct-2410\",\"https://llm-stats.com/benchmarks/triviaqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/triviaqa\"]", + "raw_score_id": "triviaqa::ministral-8b-instruct-2410", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "winogrande::winogrande-ministral-8b-instruct-2410", + "evaluation_name": "llm_stats.winogrande", + "source_data": { + "dataset_name": "Winogrande", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/ministral-8b-instruct-2410", + "https://llm-stats.com/benchmarks/winogrande", + "https://api.llm-stats.com/leaderboard/benchmarks/winogrande" + ], + "additional_details": { + "raw_benchmark_id": "winogrande", + "raw_model_id": "ministral-8b-instruct-2410", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WinoGrande: An Adversarial Winograd Schema Challenge at Scale. A large-scale dataset of 44,000 pronoun resolution problems designed to test machine commonsense reasoning. Uses adversarial filtering to reduce spurious biases and provides a more robust evaluation of whether AI systems truly understand commonsense or exploit statistical shortcuts. Current best AI methods achieve 59.4-79.1% accuracy, significantly below human performance of 94.0%.", + "metric_id": "llm_stats.winogrande.score", + "metric_name": "Winogrande score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "winogrande", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Winogrande", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "22" + } + }, + "score_details": { + "score": 0.753, + "details": { + "raw_score": "0.753", + "raw_score_field": "score", + "raw_model_id": "ministral-8b-instruct-2410", + "raw_benchmark_id": "winogrande", + "source_urls_json": "[\"https://llm-stats.com/models/ministral-8b-instruct-2410\",\"https://llm-stats.com/benchmarks/winogrande\",\"https://api.llm-stats.com/leaderboard/benchmarks/winogrande\"]", + "raw_score_id": "winogrande::ministral-8b-instruct-2410", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/mistral/mistral-small-24b-base-2501/8f4b0856-02ab-4caf-93a2-f632746319d8.json b/data/llm_stats/mistral/mistral-small-24b-base-2501/8f4b0856-02ab-4caf-93a2-f632746319d8.json new file mode 100644 index 0000000000000000000000000000000000000000..33394463bea008ed1350f685acdb608b0c547300 --- /dev/null +++ b/data/llm_stats/mistral/mistral-small-24b-base-2501/8f4b0856-02ab-4caf-93a2-f632746319d8.json @@ -0,0 +1,203 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/mistral_mistral-small-24b-base-2501/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Mistral Small 3 24B Base", + "id": "mistral/mistral-small-24b-base-2501", + "developer": "mistral", + "additional_details": { + "raw_id": "mistral-small-24b-base-2501", + "raw_name": "Mistral Small 3 24B Base", + "raw_model_id": "mistral-small-24b-base-2501", + "raw_model_name": "Mistral Small 3 24B Base", + "raw_organization_id": "mistral", + "raw_organization_name": "Mistral AI", + "raw_release_date": "2025-01-30", + "raw_announcement_date": "2025-01-30", + "raw_multimodal": "true", + "raw_param_count": "23600000000", + "raw_provider_slug": "mistral", + "raw_provider_name": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "agieval::agieval-mistral-small-24b-base-2501", + "evaluation_name": "llm_stats.agieval", + "source_data": { + "dataset_name": "AGIEval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mistral-small-24b-base-2501", + "https://llm-stats.com/benchmarks/agieval", + "https://api.llm-stats.com/leaderboard/benchmarks/agieval" + ], + "additional_details": { + "raw_benchmark_id": "agieval", + "raw_model_id": "mistral-small-24b-base-2501", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A human-centric benchmark for evaluating foundation models on standardized exams including college entrance exams (Gaokao, SAT), law school admission tests (LSAT), math competitions, lawyer qualification tests, and civil service exams. Contains 20 tasks (18 multiple-choice, 2 cloze) designed to assess understanding, knowledge, reasoning, and calculation abilities in real-world academic and professional contexts.", + "metric_id": "llm_stats.agieval.score", + "metric_name": "AGIEval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "agieval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AGIEval", + "raw_categories": "[\"legal\",\"math\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.658, + "details": { + "raw_score": "0.658", + "raw_score_field": "score", + "raw_model_id": "mistral-small-24b-base-2501", + "raw_benchmark_id": "agieval", + "source_urls_json": "[\"https://llm-stats.com/models/mistral-small-24b-base-2501\",\"https://llm-stats.com/benchmarks/agieval\",\"https://api.llm-stats.com/leaderboard/benchmarks/agieval\"]", + "raw_score_id": "agieval::mistral-small-24b-base-2501", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "arc-c::arc-c-mistral-small-24b-base-2501", + "evaluation_name": "llm_stats.arc-c", + "source_data": { + "dataset_name": "ARC-C", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mistral-small-24b-base-2501", + "https://llm-stats.com/benchmarks/arc-c", + "https://api.llm-stats.com/leaderboard/benchmarks/arc-c" + ], + "additional_details": { + "raw_benchmark_id": "arc-c", + "raw_model_id": "mistral-small-24b-base-2501", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The AI2 Reasoning Challenge (ARC) Challenge Set is a multiple-choice question-answering benchmark containing grade-school level science questions that require advanced reasoning capabilities. ARC-C specifically contains questions that were answered incorrectly by both retrieval-based and word co-occurrence algorithms, making it a particularly challenging subset designed to test commonsense reasoning abilities in AI systems.", + "metric_id": "llm_stats.arc-c.score", + "metric_name": "ARC-C score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arc-c", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ARC-C", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "34" + } + }, + "score_details": { + "score": 0.9129, + "details": { + "raw_score": "0.9129", + "raw_score_field": "score", + "raw_model_id": "mistral-small-24b-base-2501", + "raw_benchmark_id": "arc-c", + "source_urls_json": "[\"https://llm-stats.com/models/mistral-small-24b-base-2501\",\"https://llm-stats.com/benchmarks/arc-c\",\"https://api.llm-stats.com/leaderboard/benchmarks/arc-c\"]", + "raw_score_id": "arc-c::mistral-small-24b-base-2501", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "triviaqa::triviaqa-mistral-small-24b-base-2501", + "evaluation_name": "llm_stats.triviaqa", + "source_data": { + "dataset_name": "TriviaQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mistral-small-24b-base-2501", + "https://llm-stats.com/benchmarks/triviaqa", + "https://api.llm-stats.com/leaderboard/benchmarks/triviaqa" + ], + "additional_details": { + "raw_benchmark_id": "triviaqa", + "raw_model_id": "mistral-small-24b-base-2501", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A large-scale reading comprehension dataset containing over 650K question-answer-evidence triples. TriviaQA includes 95K question-answer pairs authored by trivia enthusiasts and independently gathered evidence documents (six per question on average) that provide high quality distant supervision for answering the questions. The dataset features relatively complex, compositional questions with considerable syntactic and lexical variability, requiring cross-sentence reasoning to find answers.", + "metric_id": "llm_stats.triviaqa.score", + "metric_name": "TriviaQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "triviaqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TriviaQA", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.8032, + "details": { + "raw_score": "0.8032", + "raw_score_field": "score", + "raw_model_id": "mistral-small-24b-base-2501", + "raw_benchmark_id": "triviaqa", + "source_urls_json": "[\"https://llm-stats.com/models/mistral-small-24b-base-2501\",\"https://llm-stats.com/benchmarks/triviaqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/triviaqa\"]", + "raw_score_id": "triviaqa::mistral-small-24b-base-2501", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/mistral/mistral-small-latest/e638cc22-6365-4ff9-a6f4-c3db9353538a.json b/data/llm_stats/mistral/mistral-small-latest/e638cc22-6365-4ff9-a6f4-c3db9353538a.json new file mode 100644 index 0000000000000000000000000000000000000000..4126341c91f70be08c330853ad452ff7b6516f11 --- /dev/null +++ b/data/llm_stats/mistral/mistral-small-latest/e638cc22-6365-4ff9-a6f4-c3db9353538a.json @@ -0,0 +1,204 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/mistral_mistral-small-latest/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "mistral-small-latest", + "id": "mistral/mistral-small-latest", + "developer": "mistral", + "additional_details": { + "raw_id": "mistral-small-latest", + "raw_model_name": "mistral-small-latest", + "raw_display_name": "Mistral Small 4", + "raw_organization_id": "mistral", + "raw_organization_name": "Mistral AI", + "raw_context_length": "256000", + "raw_input_modalities": "[\"image\",\"text\"]", + "raw_output_modalities": "[\"text\"]", + "raw_input_price": "0.15", + "raw_output_price": "0.6", + "raw_model_id": "mistral-small-latest", + "raw_provider_slug": "mistral", + "raw_provider_name": "ZeroEval" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aa-lcr::aa-lcr-mistral-small-latest", + "evaluation_name": "llm_stats.aa-lcr", + "source_data": { + "dataset_name": "AA-LCR", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mistral-small-latest", + "https://llm-stats.com/benchmarks/aa-lcr", + "https://api.llm-stats.com/leaderboard/benchmarks/aa-lcr" + ], + "additional_details": { + "raw_benchmark_id": "aa-lcr", + "raw_model_id": "mistral-small-latest", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Agent Arena Long Context Reasoning benchmark", + "metric_id": "llm_stats.aa-lcr.score", + "metric_name": "AA-LCR score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aa-lcr", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AA-LCR", + "raw_categories": "[\"long_context\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "14" + } + }, + "score_details": { + "score": 0.712, + "details": { + "raw_score": "0.712", + "raw_score_field": "score", + "raw_model_id": "mistral-small-latest", + "raw_benchmark_id": "aa-lcr", + "source_urls_json": "[\"https://llm-stats.com/models/mistral-small-latest\",\"https://llm-stats.com/benchmarks/aa-lcr\",\"https://api.llm-stats.com/leaderboard/benchmarks/aa-lcr\"]", + "raw_score_id": "aa-lcr::mistral-small-latest", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "arena-hard::arena-hard-mistral-small-latest", + "evaluation_name": "llm_stats.arena-hard", + "source_data": { + "dataset_name": "Arena Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mistral-small-latest", + "https://llm-stats.com/benchmarks/arena-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/arena-hard" + ], + "additional_details": { + "raw_benchmark_id": "arena-hard", + "raw_model_id": "mistral-small-latest", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Arena-Hard-Auto is an automatic evaluation benchmark for instruction-tuned LLMs consisting of 500 challenging real-world prompts curated by BenchBuilder. It includes open-ended software engineering problems, mathematical questions, and creative writing tasks. The benchmark uses LLM-as-a-Judge methodology with GPT-4.1 and Gemini-2.5 as automatic judges to approximate human preference. Arena-Hard achieves 98.6% correlation with human preference rankings and provides 3x higher separation of model performances compared to MT-Bench, making it highly effective for distinguishing between models of similar quality.", + "metric_id": "llm_stats.arena-hard.score", + "metric_name": "Arena Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arena-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Arena Hard", + "raw_categories": "[\"reasoning\",\"general\",\"creativity\",\"writing\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "26" + } + }, + "score_details": { + "score": 0.583, + "details": { + "raw_score": "0.583", + "raw_score_field": "score", + "raw_model_id": "mistral-small-latest", + "raw_benchmark_id": "arena-hard", + "source_urls_json": "[\"https://llm-stats.com/models/mistral-small-latest\",\"https://llm-stats.com/benchmarks/arena-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/arena-hard\"]", + "raw_score_id": "arena-hard::mistral-small-latest", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "collie::collie-mistral-small-latest", + "evaluation_name": "llm_stats.collie", + "source_data": { + "dataset_name": "COLLIE", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mistral-small-latest", + "https://llm-stats.com/benchmarks/collie", + "https://api.llm-stats.com/leaderboard/benchmarks/collie" + ], + "additional_details": { + "raw_benchmark_id": "collie", + "raw_model_id": "mistral-small-latest", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "COLLIE is a grammar-based framework for systematic construction of constrained text generation tasks. It allows specification of rich, compositional constraints across diverse generation levels and modeling challenges including language understanding, logical reasoning, and semantic planning. The COLLIE-v1 dataset contains 2,080 instances across 13 constraint structures.", + "metric_id": "llm_stats.collie.score", + "metric_name": "COLLIE score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "collie", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "COLLIE", + "raw_categories": "[\"language\",\"reasoning\",\"writing\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.629, + "details": { + "raw_score": "0.629", + "raw_score_field": "score", + "raw_model_id": "mistral-small-latest", + "raw_benchmark_id": "collie", + "source_urls_json": "[\"https://llm-stats.com/models/mistral-small-latest\",\"https://llm-stats.com/benchmarks/collie\",\"https://api.llm-stats.com/leaderboard/benchmarks/collie\"]", + "raw_score_id": "collie::mistral-small-latest", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/moonshotai/kimi-k2.5/5c01c774-83ba-450f-bd36-9d95013b7b4e.json b/data/llm_stats/moonshotai/kimi-k2.5/5c01c774-83ba-450f-bd36-9d95013b7b4e.json new file mode 100644 index 0000000000000000000000000000000000000000..9ba2dbeceee23ea9b6941dc0cfee2fab28fa6bd2 --- /dev/null +++ b/data/llm_stats/moonshotai/kimi-k2.5/5c01c774-83ba-450f-bd36-9d95013b7b4e.json @@ -0,0 +1,1763 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/moonshotai_kimi-k2.5/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Kimi K2.5", + "id": "moonshotai/kimi-k2.5", + "developer": "moonshotai", + "additional_details": { + "raw_id": "kimi-k2.5", + "raw_name": "Kimi K2.5", + "raw_model_id": "kimi-k2.5", + "raw_model_name": "Kimi K2.5", + "raw_organization_id": "moonshotai", + "raw_organization_name": "Moonshot AI", + "raw_release_date": "2026-01-27", + "raw_announcement_date": "2026-01-27", + "raw_multimodal": "true", + "raw_param_count": "1000000000000", + "raw_provider_slug": "moonshotai", + "raw_provider_name": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aa-lcr::aa-lcr-kimi-k2.5", + "evaluation_name": "llm_stats.aa-lcr", + "source_data": { + "dataset_name": "AA-LCR", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/aa-lcr", + "https://api.llm-stats.com/leaderboard/benchmarks/aa-lcr" + ], + "additional_details": { + "raw_benchmark_id": "aa-lcr", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Agent Arena Long Context Reasoning benchmark", + "metric_id": "llm_stats.aa-lcr.score", + "metric_name": "AA-LCR score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aa-lcr", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AA-LCR", + "raw_categories": "[\"long_context\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "14" + } + }, + "score_details": { + "score": 0.7, + "details": { + "raw_score": "0.7", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "aa-lcr", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/aa-lcr\",\"https://api.llm-stats.com/leaderboard/benchmarks/aa-lcr\"]", + "raw_score_id": "aa-lcr::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "aime-2025::aime-2025-kimi-k2.5", + "evaluation_name": "llm_stats.aime-2025", + "source_data": { + "dataset_name": "AIME 2025", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/aime-2025", + "https://api.llm-stats.com/leaderboard/benchmarks/aime-2025" + ], + "additional_details": { + "raw_benchmark_id": "aime-2025", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "All 30 problems from the 2025 American Invitational Mathematics Examination (AIME I and AIME II), testing olympiad-level mathematical reasoning with integer answers from 000-999. Used as an AI benchmark to evaluate large language models' ability to solve complex mathematical problems requiring multi-step logical deductions and structured symbolic reasoning.", + "metric_id": "llm_stats.aime-2025.score", + "metric_name": "AIME 2025 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aime-2025", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AIME 2025", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "114" + } + }, + "score_details": { + "score": 0.961, + "details": { + "raw_score": "0.961", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "aime-2025", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/aime-2025\",\"https://api.llm-stats.com/leaderboard/benchmarks/aime-2025\"]", + "raw_score_id": "aime-2025::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "browsecomp::browsecomp-kimi-k2.5", + "evaluation_name": "llm_stats.browsecomp", + "source_data": { + "dataset_name": "BrowseComp", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/browsecomp", + "https://api.llm-stats.com/leaderboard/benchmarks/browsecomp" + ], + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BrowseComp is a benchmark comprising 1,266 questions that challenge AI agents to persistently navigate the internet in search of hard-to-find, entangled information. The benchmark measures agents' ability to exercise persistence in information gathering, demonstrate creativity in web navigation, and find concise, verifiable answers. Despite the difficulty of the questions, BrowseComp is simple and easy-to-use, as predicted answers are short and easily verifiable against reference answers.", + "metric_id": "llm_stats.browsecomp.score", + "metric_name": "BrowseComp score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BrowseComp", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "52" + } + }, + "score_details": { + "score": 0.749, + "details": { + "raw_score": "0.749", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "browsecomp", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/browsecomp\",\"https://api.llm-stats.com/leaderboard/benchmarks/browsecomp\"]", + "raw_score_id": "browsecomp::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "cybergym::cybergym-kimi-k2.5", + "evaluation_name": "llm_stats.cybergym", + "source_data": { + "dataset_name": "CyberGym", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/cybergym", + "https://api.llm-stats.com/leaderboard/benchmarks/cybergym" + ], + "additional_details": { + "raw_benchmark_id": "cybergym", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CyberGym is a benchmark for evaluating AI agents on cybersecurity tasks, testing their ability to identify vulnerabilities, perform security analysis, and complete security-related challenges in a controlled environment.", + "metric_id": "llm_stats.cybergym.score", + "metric_name": "CyberGym score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "cybergym", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CyberGym", + "raw_categories": "[\"safety\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.413, + "details": { + "raw_score": "0.413", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "cybergym", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/cybergym\",\"https://api.llm-stats.com/leaderboard/benchmarks/cybergym\"]", + "raw_score_id": "cybergym::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "deepsearchqa::deepsearchqa-kimi-k2.5", + "evaluation_name": "llm_stats.deepsearchqa", + "source_data": { + "dataset_name": "DeepSearchQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/deepsearchqa", + "https://api.llm-stats.com/leaderboard/benchmarks/deepsearchqa" + ], + "additional_details": { + "raw_benchmark_id": "deepsearchqa", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "DeepSearchQA is a benchmark for evaluating deep search and question-answering capabilities, testing models' ability to perform multi-hop reasoning and information retrieval across complex knowledge domains.", + "metric_id": "llm_stats.deepsearchqa.score", + "metric_name": "DeepSearchQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "deepsearchqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DeepSearchQA", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.771, + "details": { + "raw_score": "0.771", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "deepsearchqa", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/deepsearchqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/deepsearchqa\"]", + "raw_score_id": "deepsearchqa::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "finsearchcomp-t2-t3::finsearchcomp-t2-t3-kimi-k2.5", + "evaluation_name": "llm_stats.finsearchcomp-t2-t3", + "source_data": { + "dataset_name": "FinSearchComp T2&T3", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/finsearchcomp-t2-t3", + "https://api.llm-stats.com/leaderboard/benchmarks/finsearchcomp-t2-t3" + ], + "additional_details": { + "raw_benchmark_id": "finsearchcomp-t2-t3", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "FinSearchComp T2&T3 is a combined benchmark for evaluating financial search and reasoning capabilities on Tier 2 and Tier 3 tasks, testing models' ability to retrieve and analyze complex financial information using tools.", + "metric_id": "llm_stats.finsearchcomp-t2-t3.score", + "metric_name": "FinSearchComp T2&T3 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "finsearchcomp-t2-t3", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FinSearchComp T2&T3", + "raw_categories": "[\"reasoning\",\"search\",\"finance\",\"economics\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.678, + "details": { + "raw_score": "0.678", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "finsearchcomp-t2-t3", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/finsearchcomp-t2-t3\",\"https://api.llm-stats.com/leaderboard/benchmarks/finsearchcomp-t2-t3\"]", + "raw_score_id": "finsearchcomp-t2-t3::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hmmt-2025::hmmt-2025-kimi-k2.5", + "evaluation_name": "llm_stats.hmmt-2025", + "source_data": { + "dataset_name": "HMMT 2025", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/hmmt-2025", + "https://api.llm-stats.com/leaderboard/benchmarks/hmmt-2025" + ], + "additional_details": { + "raw_benchmark_id": "hmmt-2025", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Harvard-MIT Mathematics Tournament 2025 - A prestigious student-organized mathematics competition for high school students featuring two tournaments (November 2025 at MIT and February 2026 at Harvard) with individual tests, team rounds, and guts rounds", + "metric_id": "llm_stats.hmmt-2025.score", + "metric_name": "HMMT 2025 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hmmt-2025", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HMMT 2025", + "raw_categories": "[\"math\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "33" + } + }, + "score_details": { + "score": 0.954, + "details": { + "raw_score": "0.954", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "hmmt-2025", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/hmmt-2025\",\"https://api.llm-stats.com/leaderboard/benchmarks/hmmt-2025\"]", + "raw_score_id": "hmmt-2025::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "humanity-s-last-exam::humanity-s-last-exam-kimi-k2.5", + "evaluation_name": "llm_stats.humanity-s-last-exam", + "source_data": { + "dataset_name": "Humanity's Last Exam", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/humanity-s-last-exam", + "https://api.llm-stats.com/leaderboard/benchmarks/humanity's-last-exam" + ], + "additional_details": { + "raw_benchmark_id": "humanity's-last-exam", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Humanity's Last Exam (HLE) is a multi-modal academic benchmark with 2,500 questions across mathematics, humanities, and natural sciences, designed to test LLM capabilities at the frontier of human knowledge with unambiguous, verifiable solutions", + "metric_id": "llm_stats.humanity-s-last-exam.score", + "metric_name": "Humanity's Last Exam score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "humanity's-last-exam", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Humanity's Last Exam", + "raw_categories": "[\"math\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "88" + } + }, + "score_details": { + "score": 0.502, + "details": { + "raw_score": "0.502", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "humanity's-last-exam", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/humanity-s-last-exam\",\"https://api.llm-stats.com/leaderboard/benchmarks/humanity's-last-exam\"]", + "raw_score_id": "humanity's-last-exam::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "imo-answerbench::imo-answerbench-kimi-k2.5", + "evaluation_name": "llm_stats.imo-answerbench", + "source_data": { + "dataset_name": "IMO-AnswerBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/imo-answerbench", + "https://api.llm-stats.com/leaderboard/benchmarks/imo-answerbench" + ], + "additional_details": { + "raw_benchmark_id": "imo-answerbench", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "IMO-AnswerBench is a benchmark for evaluating mathematical reasoning capabilities on International Mathematical Olympiad (IMO) problems, focusing on answer generation and verification.", + "metric_id": "llm_stats.imo-answerbench.score", + "metric_name": "IMO-AnswerBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "imo-answerbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IMO-AnswerBench", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.818, + "details": { + "raw_score": "0.818", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "imo-answerbench", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/imo-answerbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/imo-answerbench\"]", + "raw_score_id": "imo-answerbench::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "infovqatest::infovqatest-kimi-k2.5", + "evaluation_name": "llm_stats.infovqatest", + "source_data": { + "dataset_name": "InfoVQAtest", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/infovqatest", + "https://api.llm-stats.com/leaderboard/benchmarks/infovqatest" + ], + "additional_details": { + "raw_benchmark_id": "infovqatest", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "InfoVQA test set with infographic images requiring joint reasoning over document layout, textual content, graphical elements, and data visualizations with elementary reasoning and arithmetic skills", + "metric_id": "llm_stats.infovqatest.score", + "metric_name": "InfoVQAtest score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "infovqatest", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "InfoVQAtest", + "raw_categories": "[\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.926, + "details": { + "raw_score": "0.926", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "infovqatest", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/infovqatest\",\"https://api.llm-stats.com/leaderboard/benchmarks/infovqatest\"]", + "raw_score_id": "infovqatest::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "livecodebench-v6::livecodebench-v6-kimi-k2.5", + "evaluation_name": "llm_stats.livecodebench-v6", + "source_data": { + "dataset_name": "LiveCodeBench v6", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/livecodebench-v6", + "https://api.llm-stats.com/leaderboard/benchmarks/livecodebench-v6" + ], + "additional_details": { + "raw_benchmark_id": "livecodebench-v6", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LiveCodeBench is a holistic and contamination-free evaluation benchmark for large language models for code. It continuously collects new problems from programming contests (LeetCode, AtCoder, CodeForces) and evaluates four different scenarios: code generation, self-repair, code execution, and test output prediction. Problems are annotated with release dates to enable evaluation on unseen problems released after a model's training cutoff.", + "metric_id": "llm_stats.livecodebench-v6.score", + "metric_name": "LiveCodeBench v6 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "livecodebench-v6", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LiveCodeBench v6", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "53" + } + }, + "score_details": { + "score": 0.85, + "details": { + "raw_score": "0.85", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "livecodebench-v6", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/livecodebench-v6\",\"https://api.llm-stats.com/leaderboard/benchmarks/livecodebench-v6\"]", + "raw_score_id": "livecodebench-v6::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "longbench-v2::longbench-v2-kimi-k2.5", + "evaluation_name": "llm_stats.longbench-v2", + "source_data": { + "dataset_name": "LongBench v2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/longbench-v2", + "https://api.llm-stats.com/leaderboard/benchmarks/longbench-v2" + ], + "additional_details": { + "raw_benchmark_id": "longbench-v2", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LongBench v2 is a benchmark designed to assess the ability of LLMs to handle long-context problems requiring deep understanding and reasoning across real-world multitasks. It consists of 503 challenging multiple-choice questions with contexts ranging from 8k to 2M words across six major task categories: single-document QA, multi-document QA, long in-context learning, long-dialogue history understanding, code repository understanding, and long structured data understanding.", + "metric_id": "llm_stats.longbench-v2.score", + "metric_name": "LongBench v2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "longbench-v2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LongBench v2", + "raw_categories": "[\"long_context\",\"reasoning\",\"structured_output\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.61, + "details": { + "raw_score": "0.61", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "longbench-v2", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/longbench-v2\",\"https://api.llm-stats.com/leaderboard/benchmarks/longbench-v2\"]", + "raw_score_id": "longbench-v2::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "longvideobench::longvideobench-kimi-k2.5", + "evaluation_name": "llm_stats.longvideobench", + "source_data": { + "dataset_name": "LongVideoBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/longvideobench", + "https://api.llm-stats.com/leaderboard/benchmarks/longvideobench" + ], + "additional_details": { + "raw_benchmark_id": "longvideobench", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LongVideoBench is a question-answering benchmark featuring video-language interleaved inputs up to an hour long. It includes 3,763 varying-length web-collected videos with subtitles across diverse themes and 6,678 human-annotated multiple-choice questions in 17 fine-grained categories for comprehensive evaluation of long-term multimodal understanding.", + "metric_id": "llm_stats.longvideobench.score", + "metric_name": "LongVideoBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "longvideobench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LongVideoBench", + "raw_categories": "[\"long_context\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.798, + "details": { + "raw_score": "0.798", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "longvideobench", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/longvideobench\",\"https://api.llm-stats.com/leaderboard/benchmarks/longvideobench\"]", + "raw_score_id": "longvideobench::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "lvbench::lvbench-kimi-k2.5", + "evaluation_name": "llm_stats.lvbench", + "source_data": { + "dataset_name": "LVBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/lvbench", + "https://api.llm-stats.com/leaderboard/benchmarks/lvbench" + ], + "additional_details": { + "raw_benchmark_id": "lvbench", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LVBench is an extreme long video understanding benchmark designed to evaluate multimodal models on videos up to two hours in duration. It contains 6 major categories and 21 subcategories, with videos averaging five times longer than existing datasets. The benchmark addresses applications requiring comprehension of extremely long videos.", + "metric_id": "llm_stats.lvbench.score", + "metric_name": "LVBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "lvbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LVBench", + "raw_categories": "[\"long_context\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.759, + "details": { + "raw_score": "0.759", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "lvbench", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/lvbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/lvbench\"]", + "raw_score_id": "lvbench::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mathvision::mathvision-kimi-k2.5", + "evaluation_name": "llm_stats.mathvision", + "source_data": { + "dataset_name": "MathVision", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/mathvision", + "https://api.llm-stats.com/leaderboard/benchmarks/mathvision" + ], + "additional_details": { + "raw_benchmark_id": "mathvision", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MATH-Vision is a dataset designed to measure multimodal mathematical reasoning capabilities. It focuses on evaluating how well models can solve mathematical problems that require both visual understanding and mathematical reasoning, bridging the gap between visual and mathematical domains.", + "metric_id": "llm_stats.mathvision.score", + "metric_name": "MathVision score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mathvision", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MathVision", + "raw_categories": "[\"math\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "31" + } + }, + "score_details": { + "score": 0.842, + "details": { + "raw_score": "0.842", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "mathvision", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/mathvision\",\"https://api.llm-stats.com/leaderboard/benchmarks/mathvision\"]", + "raw_score_id": "mathvision::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mathvista-mini::mathvista-mini-kimi-k2.5", + "evaluation_name": "llm_stats.mathvista-mini", + "source_data": { + "dataset_name": "MathVista-Mini", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/mathvista-mini", + "https://api.llm-stats.com/leaderboard/benchmarks/mathvista-mini" + ], + "additional_details": { + "raw_benchmark_id": "mathvista-mini", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MathVista-Mini is a smaller version of the MathVista benchmark that evaluates mathematical reasoning in visual contexts. It consists of examples derived from multimodal datasets involving mathematics, combining challenges from diverse mathematical and visual tasks to assess foundation models' ability to solve problems requiring both visual understanding and mathematical reasoning.", + "metric_id": "llm_stats.mathvista-mini.score", + "metric_name": "MathVista-Mini score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mathvista-mini", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MathVista-Mini", + "raw_categories": "[\"math\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.901, + "details": { + "raw_score": "0.901", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "mathvista-mini", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/mathvista-mini\",\"https://api.llm-stats.com/leaderboard/benchmarks/mathvista-mini\"]", + "raw_score_id": "mathvista-mini::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu-pro::mmlu-pro-kimi-k2.5", + "evaluation_name": "llm_stats.mmlu-pro", + "source_data": { + "dataset_name": "MMLU-Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/mmlu-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu-pro" + ], + "additional_details": { + "raw_benchmark_id": "mmlu-pro", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A more robust and challenging multi-task language understanding benchmark that extends MMLU by expanding multiple-choice options from 4 to 10, eliminating trivial questions, and focusing on reasoning-intensive tasks. Features over 12,000 curated questions across 14 domains and causes a 16-33% accuracy drop compared to original MMLU.", + "metric_id": "llm_stats.mmlu-pro.score", + "metric_name": "MMLU-Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU-Pro", + "raw_categories": "[\"language\",\"legal\",\"math\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "129" + } + }, + "score_details": { + "score": 0.871, + "details": { + "raw_score": "0.871", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "mmlu-pro", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/mmlu-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu-pro\"]", + "raw_score_id": "mmlu-pro::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmmu-pro::mmmu-pro-kimi-k2.5", + "evaluation_name": "llm_stats.mmmu-pro", + "source_data": { + "dataset_name": "MMMU-Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/mmmu-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/mmmu-pro" + ], + "additional_details": { + "raw_benchmark_id": "mmmu-pro", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A more robust multi-discipline multimodal understanding benchmark that enhances MMMU through a three-step process: filtering text-only answerable questions, augmenting candidate options, and introducing vision-only input settings. Achieves significantly lower model performance (16.8-26.9%) compared to original MMMU, providing more rigorous evaluation that closely mimics real-world scenarios.", + "metric_id": "llm_stats.mmmu-pro.score", + "metric_name": "MMMU-Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmmu-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMMU-Pro", + "raw_categories": "[\"multimodal\",\"reasoning\",\"general\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "60" + } + }, + "score_details": { + "score": 0.785, + "details": { + "raw_score": "0.785", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "mmmu-pro", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/mmmu-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmmu-pro\"]", + "raw_score_id": "mmmu-pro::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmvu::mmvu-kimi-k2.5", + "evaluation_name": "llm_stats.mmvu", + "source_data": { + "dataset_name": "MMVU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/mmvu", + "https://api.llm-stats.com/leaderboard/benchmarks/mmvu" + ], + "additional_details": { + "raw_benchmark_id": "mmvu", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MMVU (Multimodal Multi-disciplinary Video Understanding) is a benchmark for evaluating multimodal models on video understanding tasks across multiple disciplines, testing comprehension and reasoning capabilities on video content.", + "metric_id": "llm_stats.mmvu.score", + "metric_name": "MMVU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmvu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMVU", + "raw_categories": "[\"multimodal\",\"reasoning\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.804, + "details": { + "raw_score": "0.804", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "mmvu", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/mmvu\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmvu\"]", + "raw_score_id": "mmvu::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "motionbench::motionbench-kimi-k2.5", + "evaluation_name": "llm_stats.motionbench", + "source_data": { + "dataset_name": "MotionBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/motionbench", + "https://api.llm-stats.com/leaderboard/benchmarks/motionbench" + ], + "additional_details": { + "raw_benchmark_id": "motionbench", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MotionBench is a benchmark for evaluating multimodal models on motion understanding in videos, testing the ability to comprehend temporal dynamics, movement patterns, and action sequences.", + "metric_id": "llm_stats.motionbench.score", + "metric_name": "MotionBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "motionbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MotionBench", + "raw_categories": "[\"multimodal\",\"reasoning\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.704, + "details": { + "raw_score": "0.704", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "motionbench", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/motionbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/motionbench\"]", + "raw_score_id": "motionbench::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ocrbench::ocrbench-kimi-k2.5", + "evaluation_name": "llm_stats.ocrbench", + "source_data": { + "dataset_name": "OCRBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/ocrbench", + "https://api.llm-stats.com/leaderboard/benchmarks/ocrbench" + ], + "additional_details": { + "raw_benchmark_id": "ocrbench", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OCRBench: Comprehensive evaluation benchmark for assessing Optical Character Recognition (OCR) capabilities in Large Multimodal Models across text recognition, scene text VQA, and document understanding tasks", + "metric_id": "llm_stats.ocrbench.score", + "metric_name": "OCRBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ocrbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OCRBench", + "raw_categories": "[\"image_to_text\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "22" + } + }, + "score_details": { + "score": 0.923, + "details": { + "raw_score": "0.923", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "ocrbench", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/ocrbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/ocrbench\"]", + "raw_score_id": "ocrbench::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ojbench-cpp::ojbench-cpp-kimi-k2.5", + "evaluation_name": "llm_stats.ojbench-cpp", + "source_data": { + "dataset_name": "OJBench (C++)", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/ojbench-cpp", + "https://api.llm-stats.com/leaderboard/benchmarks/ojbench-cpp" + ], + "additional_details": { + "raw_benchmark_id": "ojbench-cpp", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OJBench (C++) is the C++ subset of OJBench, a competition-level code benchmark that evaluates large language models on programming competition problems from NOI and ICPC using C++ as the implementation language.", + "metric_id": "llm_stats.ojbench-cpp.score", + "metric_name": "OJBench (C++) score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ojbench-cpp", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OJBench (C++)", + "raw_categories": "[\"reasoning\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.574, + "details": { + "raw_score": "0.574", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "ojbench-cpp", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/ojbench-cpp\",\"https://api.llm-stats.com/leaderboard/benchmarks/ojbench-cpp\"]", + "raw_score_id": "ojbench-cpp::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "omnidocbench-1.5::omnidocbench-1.5-kimi-k2.5", + "evaluation_name": "llm_stats.omnidocbench-1.5", + "source_data": { + "dataset_name": "OmniDocBench 1.5", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/omnidocbench-1.5", + "https://api.llm-stats.com/leaderboard/benchmarks/omnidocbench-1.5" + ], + "additional_details": { + "raw_benchmark_id": "omnidocbench-1.5", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OmniDocBench 1.5 is a comprehensive benchmark for evaluating multimodal large language models on document understanding tasks, including OCR, document parsing, information extraction, and visual question answering across diverse document types. Lower Overall Edit Distance scores are better.", + "metric_id": "llm_stats.omnidocbench-1.5.score", + "metric_name": "OmniDocBench 1.5 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "omnidocbench-1.5", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OmniDocBench 1.5", + "raw_categories": "[\"multimodal\",\"reasoning\",\"structured_output\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.888, + "details": { + "raw_score": "0.888", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "omnidocbench-1.5", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/omnidocbench-1.5\",\"https://api.llm-stats.com/leaderboard/benchmarks/omnidocbench-1.5\"]", + "raw_score_id": "omnidocbench-1.5::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "paperbench::paperbench-kimi-k2.5", + "evaluation_name": "llm_stats.paperbench", + "source_data": { + "dataset_name": "PaperBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/paperbench", + "https://api.llm-stats.com/leaderboard/benchmarks/paperbench" + ], + "additional_details": { + "raw_benchmark_id": "paperbench", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "PaperBench is a benchmark for evaluating AI agents on their ability to replicate research papers. It tests models on complex, multi-step workflows involving code implementation, experimentation, and reproducing scientific results from academic publications.", + "metric_id": "llm_stats.paperbench.score", + "metric_name": "PaperBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "paperbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PaperBench", + "raw_categories": "[\"reasoning\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.635, + "details": { + "raw_score": "0.635", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "paperbench", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/paperbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/paperbench\"]", + "raw_score_id": "paperbench::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "scicode::scicode-kimi-k2.5", + "evaluation_name": "llm_stats.scicode", + "source_data": { + "dataset_name": "SciCode", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/scicode", + "https://api.llm-stats.com/leaderboard/benchmarks/scicode" + ], + "additional_details": { + "raw_benchmark_id": "scicode", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SciCode is a research coding benchmark curated by scientists that challenges language models to code solutions for scientific problems. It contains 338 subproblems decomposed from 80 challenging main problems across 16 natural science sub-fields including mathematics, physics, chemistry, biology, and materials science. Problems require knowledge recall, reasoning, and code synthesis skills.", + "metric_id": "llm_stats.scicode.score", + "metric_name": "SciCode score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "scicode", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SciCode", + "raw_categories": "[\"math\",\"physics\",\"reasoning\",\"biology\",\"chemistry\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.487, + "details": { + "raw_score": "0.487", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "scicode", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/scicode\",\"https://api.llm-stats.com/leaderboard/benchmarks/scicode\"]", + "raw_score_id": "scicode::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "seal-0::seal-0-kimi-k2.5", + "evaluation_name": "llm_stats.seal-0", + "source_data": { + "dataset_name": "Seal-0", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/seal-0", + "https://api.llm-stats.com/leaderboard/benchmarks/seal-0" + ], + "additional_details": { + "raw_benchmark_id": "seal-0", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Seal-0 is a benchmark for evaluating agentic search capabilities, testing models' ability to navigate and retrieve information using tools.", + "metric_id": "llm_stats.seal-0.score", + "metric_name": "Seal-0 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "seal-0", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Seal-0", + "raw_categories": "[\"reasoning\",\"search\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.574, + "details": { + "raw_score": "0.574", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "seal-0", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/seal-0\",\"https://api.llm-stats.com/leaderboard/benchmarks/seal-0\"]", + "raw_score_id": "seal-0::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "simplevqa::simplevqa-kimi-k2.5", + "evaluation_name": "llm_stats.simplevqa", + "source_data": { + "dataset_name": "SimpleVQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/simplevqa", + "https://api.llm-stats.com/leaderboard/benchmarks/simplevqa" + ], + "additional_details": { + "raw_benchmark_id": "simplevqa", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SimpleVQA is a visual question answering benchmark focused on simple queries.", + "metric_id": "llm_stats.simplevqa.score", + "metric_name": "SimpleVQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "simplevqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SimpleVQA", + "raw_categories": "[\"image_to_text\",\"multimodal\",\"general\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.712, + "details": { + "raw_score": "0.712", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "simplevqa", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/simplevqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/simplevqa\"]", + "raw_score_id": "simplevqa::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-multilingual::swe-bench-multilingual-kimi-k2.5", + "evaluation_name": "llm_stats.swe-bench-multilingual", + "source_data": { + "dataset_name": "SWE-bench Multilingual", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/swe-bench-multilingual", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-multilingual" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-multilingual", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multilingual benchmark for issue resolving in software engineering that covers Java, TypeScript, JavaScript, Go, Rust, C, and C++. Contains 1,632 high-quality instances carefully annotated from 2,456 candidates by 68 expert annotators, designed to evaluate Large Language Models across diverse software ecosystems beyond Python.", + "metric_id": "llm_stats.swe-bench-multilingual.score", + "metric_name": "SWE-bench Multilingual score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-multilingual", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-bench Multilingual", + "raw_categories": "[\"reasoning\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "33" + } + }, + "score_details": { + "score": 0.73, + "details": { + "raw_score": "0.73", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "swe-bench-multilingual", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/swe-bench-multilingual\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-multilingual\"]", + "raw_score_id": "swe-bench-multilingual::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "video-mme::video-mme-kimi-k2.5", + "evaluation_name": "llm_stats.video-mme", + "source_data": { + "dataset_name": "Video-MME", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/video-mme", + "https://api.llm-stats.com/leaderboard/benchmarks/video-mme" + ], + "additional_details": { + "raw_benchmark_id": "video-mme", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Video-MME is the first-ever comprehensive evaluation benchmark of Multi-modal Large Language Models (MLLMs) in video analysis. It features 900 videos totaling 254 hours with 2,700 human-annotated question-answer pairs across 6 primary visual domains (Knowledge, Film & Television, Sports Competition, Life Record, Multilingual, and others) and 30 subfields. The benchmark evaluates models across diverse temporal dimensions (11 seconds to 1 hour), integrates multi-modal inputs including video frames, subtitles, and audio, and uses rigorous manual labeling by expert annotators for precise assessment.", + "metric_id": "llm_stats.video-mme.score", + "metric_name": "Video-MME score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "video-mme", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Video-MME", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "17" + } + }, + "score_details": { + "score": 0.874, + "details": { + "raw_score": "0.874", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "video-mme", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/video-mme\",\"https://api.llm-stats.com/leaderboard/benchmarks/video-mme\"]", + "raw_score_id": "video-mme::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "videommmu::videommmu-kimi-k2.5", + "evaluation_name": "llm_stats.videommmu", + "source_data": { + "dataset_name": "VideoMMMU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/videommmu", + "https://api.llm-stats.com/leaderboard/benchmarks/videommmu" + ], + "additional_details": { + "raw_benchmark_id": "videommmu", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Video-MMMU evaluates Large Multimodal Models' ability to acquire knowledge from expert-level professional videos across six disciplines through three cognitive stages: perception, comprehension, and adaptation. Contains 300 videos and 900 human-annotated questions spanning Art, Business, Science, Medicine, Humanities, and Engineering.", + "metric_id": "llm_stats.videommmu.score", + "metric_name": "VideoMMMU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "videommmu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VideoMMMU", + "raw_categories": "[\"multimodal\",\"reasoning\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "26" + } + }, + "score_details": { + "score": 0.866, + "details": { + "raw_score": "0.866", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "videommmu", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/videommmu\",\"https://api.llm-stats.com/leaderboard/benchmarks/videommmu\"]", + "raw_score_id": "videommmu::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "widesearch::widesearch-kimi-k2.5", + "evaluation_name": "llm_stats.widesearch", + "source_data": { + "dataset_name": "WideSearch", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/widesearch", + "https://api.llm-stats.com/leaderboard/benchmarks/widesearch" + ], + "additional_details": { + "raw_benchmark_id": "widesearch", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WideSearch is an agentic search benchmark that evaluates models' ability to perform broad, parallel search operations across multiple sources. It tests wide-coverage information retrieval and synthesis capabilities.", + "metric_id": "llm_stats.widesearch.score", + "metric_name": "WideSearch score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "widesearch", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WideSearch", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.79, + "details": { + "raw_score": "0.79", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "widesearch", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/widesearch\",\"https://api.llm-stats.com/leaderboard/benchmarks/widesearch\"]", + "raw_score_id": "widesearch::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "worldvqa::worldvqa-kimi-k2.5", + "evaluation_name": "llm_stats.worldvqa", + "source_data": { + "dataset_name": "WorldVQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/worldvqa", + "https://api.llm-stats.com/leaderboard/benchmarks/worldvqa" + ], + "additional_details": { + "raw_benchmark_id": "worldvqa", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WorldVQA is a benchmark designed to evaluate atomic vision-centric world knowledge. It assesses models' ability to understand and reason about visual elements representing real-world knowledge.", + "metric_id": "llm_stats.worldvqa.score", + "metric_name": "WorldVQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "worldvqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WorldVQA", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.463, + "details": { + "raw_score": "0.463", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "worldvqa", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/worldvqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/worldvqa\"]", + "raw_score_id": "worldvqa::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "zerobench::zerobench-kimi-k2.5", + "evaluation_name": "llm_stats.zerobench", + "source_data": { + "dataset_name": "ZEROBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.5", + "https://llm-stats.com/benchmarks/zerobench", + "https://api.llm-stats.com/leaderboard/benchmarks/zerobench" + ], + "additional_details": { + "raw_benchmark_id": "zerobench", + "raw_model_id": "kimi-k2.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ZEROBench is a challenging vision benchmark designed to test models on zero-shot visual understanding tasks.", + "metric_id": "llm_stats.zerobench.score", + "metric_name": "ZEROBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "zerobench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ZEROBench", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.11, + "details": { + "raw_score": "0.11", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.5", + "raw_benchmark_id": "zerobench", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.5\",\"https://llm-stats.com/benchmarks/zerobench\",\"https://api.llm-stats.com/leaderboard/benchmarks/zerobench\"]", + "raw_score_id": "zerobench::kimi-k2.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/moonshotai/kimi-k2.6/4f7c5490-2210-4433-9ece-6cbb0318f078.json b/data/llm_stats/moonshotai/kimi-k2.6/4f7c5490-2210-4433-9ece-6cbb0318f078.json new file mode 100644 index 0000000000000000000000000000000000000000..c600606493ab61c1b180b7bd3c540fcc977aa713 --- /dev/null +++ b/data/llm_stats/moonshotai/kimi-k2.6/4f7c5490-2210-4433-9ece-6cbb0318f078.json @@ -0,0 +1,1296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/moonshotai_kimi-k2.6/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.6", + "id": "moonshotai/kimi-k2.6", + "developer": "moonshotai", + "additional_details": { + "raw_id": "kimi-k2.6", + "raw_model_name": "kimi-k2.6", + "raw_display_name": "Kimi K2.6", + "raw_organization_id": "moonshotai", + "raw_organization_name": "Moonshot AI", + "raw_context_length": "262144", + "raw_input_modalities": "[\"image\",\"text\",\"video\"]", + "raw_output_modalities": "[\"text\"]", + "raw_input_price": "0.75", + "raw_output_price": "3.5", + "raw_model_id": "kimi-k2.6", + "raw_provider_slug": "moonshotai", + "raw_provider_name": "ZeroEval" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aime-2026::aime-2026-kimi-k2.6", + "evaluation_name": "llm_stats.aime-2026", + "source_data": { + "dataset_name": "AIME 2026", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.6", + "https://llm-stats.com/benchmarks/aime-2026", + "https://api.llm-stats.com/leaderboard/benchmarks/aime-2026" + ], + "additional_details": { + "raw_benchmark_id": "aime-2026", + "raw_model_id": "kimi-k2.6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "All 30 problems from the 2026 American Invitational Mathematics Examination (AIME I and AIME II), testing olympiad-level mathematical reasoning with integer answers from 000-999. Used as an AI benchmark to evaluate large language models' ability to solve complex mathematical problems requiring multi-step logical deductions and structured symbolic reasoning.", + "metric_id": "llm_stats.aime-2026.score", + "metric_name": "AIME 2026 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aime-2026", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AIME 2026", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "17" + } + }, + "score_details": { + "score": 0.964, + "details": { + "raw_score": "0.964", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.6", + "raw_benchmark_id": "aime-2026", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.6\",\"https://llm-stats.com/benchmarks/aime-2026\",\"https://api.llm-stats.com/leaderboard/benchmarks/aime-2026\"]", + "raw_score_id": "aime-2026::kimi-k2.6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "apex-agents::apex-agents-kimi-k2.6", + "evaluation_name": "llm_stats.apex-agents", + "source_data": { + "dataset_name": "APEX-Agents", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.6", + "https://llm-stats.com/benchmarks/apex-agents", + "https://api.llm-stats.com/leaderboard/benchmarks/apex-agents" + ], + "additional_details": { + "raw_benchmark_id": "apex-agents", + "raw_model_id": "kimi-k2.6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "APEX-Agents is a benchmark evaluating AI agents on long horizon professional tasks that require sustained reasoning, planning, and execution across complex multi-step workflows.", + "metric_id": "llm_stats.apex-agents.score", + "metric_name": "APEX-Agents score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "apex-agents", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "APEX-Agents", + "raw_categories": "[\"reasoning\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.279, + "details": { + "raw_score": "0.279", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.6", + "raw_benchmark_id": "apex-agents", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.6\",\"https://llm-stats.com/benchmarks/apex-agents\",\"https://api.llm-stats.com/leaderboard/benchmarks/apex-agents\"]", + "raw_score_id": "apex-agents::kimi-k2.6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "babyvision::babyvision-kimi-k2.6", + "evaluation_name": "llm_stats.babyvision", + "source_data": { + "dataset_name": "BabyVision", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.6", + "https://llm-stats.com/benchmarks/babyvision", + "https://api.llm-stats.com/leaderboard/benchmarks/babyvision" + ], + "additional_details": { + "raw_benchmark_id": "babyvision", + "raw_model_id": "kimi-k2.6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A benchmark for early-stage visual reasoning and perception on child-like vision tasks.", + "metric_id": "llm_stats.babyvision.score", + "metric_name": "BabyVision score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "babyvision", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BabyVision", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.685, + "details": { + "raw_score": "0.685", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.6", + "raw_benchmark_id": "babyvision", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.6\",\"https://llm-stats.com/benchmarks/babyvision\",\"https://api.llm-stats.com/leaderboard/benchmarks/babyvision\"]", + "raw_score_id": "babyvision::kimi-k2.6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "browsecomp::browsecomp-kimi-k2.6", + "evaluation_name": "llm_stats.browsecomp", + "source_data": { + "dataset_name": "BrowseComp", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.6", + "https://llm-stats.com/benchmarks/browsecomp", + "https://api.llm-stats.com/leaderboard/benchmarks/browsecomp" + ], + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_model_id": "kimi-k2.6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BrowseComp is a benchmark comprising 1,266 questions that challenge AI agents to persistently navigate the internet in search of hard-to-find, entangled information. The benchmark measures agents' ability to exercise persistence in information gathering, demonstrate creativity in web navigation, and find concise, verifiable answers. Despite the difficulty of the questions, BrowseComp is simple and easy-to-use, as predicted answers are short and easily verifiable against reference answers.", + "metric_id": "llm_stats.browsecomp.score", + "metric_name": "BrowseComp score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BrowseComp", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "52" + } + }, + "score_details": { + "score": 0.863, + "details": { + "raw_score": "0.863", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.6", + "raw_benchmark_id": "browsecomp", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.6\",\"https://llm-stats.com/benchmarks/browsecomp\",\"https://api.llm-stats.com/leaderboard/benchmarks/browsecomp\"]", + "raw_score_id": "browsecomp::kimi-k2.6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "charxiv-r::charxiv-r-kimi-k2.6", + "evaluation_name": "llm_stats.charxiv-r", + "source_data": { + "dataset_name": "CharXiv-R", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.6", + "https://llm-stats.com/benchmarks/charxiv-r", + "https://api.llm-stats.com/leaderboard/benchmarks/charxiv-r" + ], + "additional_details": { + "raw_benchmark_id": "charxiv-r", + "raw_model_id": "kimi-k2.6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CharXiv-R is the reasoning component of the CharXiv benchmark, focusing on complex reasoning questions that require synthesizing information across visual chart elements. It evaluates multimodal large language models on their ability to understand and reason about scientific charts from arXiv papers through various reasoning tasks.", + "metric_id": "llm_stats.charxiv-r.score", + "metric_name": "CharXiv-R score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "charxiv-r", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CharXiv-R", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "43" + } + }, + "score_details": { + "score": 0.867, + "details": { + "raw_score": "0.867", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.6", + "raw_benchmark_id": "charxiv-r", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.6\",\"https://llm-stats.com/benchmarks/charxiv-r\",\"https://api.llm-stats.com/leaderboard/benchmarks/charxiv-r\"]", + "raw_score_id": "charxiv-r::kimi-k2.6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "claw-eval::claw-eval-kimi-k2.6", + "evaluation_name": "llm_stats.claw-eval", + "source_data": { + "dataset_name": "Claw-Eval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.6", + "https://llm-stats.com/benchmarks/claw-eval", + "https://api.llm-stats.com/leaderboard/benchmarks/claw-eval" + ], + "additional_details": { + "raw_benchmark_id": "claw-eval", + "raw_model_id": "kimi-k2.6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Claw-Eval tests real-world agentic task completion across complex multi-step scenarios, evaluating a model's ability to use tools, navigate environments, and complete end-to-end tasks autonomously.", + "metric_id": "llm_stats.claw-eval.score", + "metric_name": "Claw-Eval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "claw-eval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Claw-Eval", + "raw_categories": "[\"agents\",\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.809, + "details": { + "raw_score": "0.809", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.6", + "raw_benchmark_id": "claw-eval", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.6\",\"https://llm-stats.com/benchmarks/claw-eval\",\"https://api.llm-stats.com/leaderboard/benchmarks/claw-eval\"]", + "raw_score_id": "claw-eval::kimi-k2.6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "deepsearchqa::deepsearchqa-kimi-k2.6", + "evaluation_name": "llm_stats.deepsearchqa", + "source_data": { + "dataset_name": "DeepSearchQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.6", + "https://llm-stats.com/benchmarks/deepsearchqa", + "https://api.llm-stats.com/leaderboard/benchmarks/deepsearchqa" + ], + "additional_details": { + "raw_benchmark_id": "deepsearchqa", + "raw_model_id": "kimi-k2.6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "DeepSearchQA is a benchmark for evaluating deep search and question-answering capabilities, testing models' ability to perform multi-hop reasoning and information retrieval across complex knowledge domains.", + "metric_id": "llm_stats.deepsearchqa.score", + "metric_name": "DeepSearchQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "deepsearchqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DeepSearchQA", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.83, + "details": { + "raw_score": "0.83", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.6", + "raw_benchmark_id": "deepsearchqa", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.6\",\"https://llm-stats.com/benchmarks/deepsearchqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/deepsearchqa\"]", + "raw_score_id": "deepsearchqa::kimi-k2.6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gpqa::gpqa-kimi-k2.6", + "evaluation_name": "llm_stats.gpqa", + "source_data": { + "dataset_name": "GPQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.6", + "https://llm-stats.com/benchmarks/gpqa", + "https://api.llm-stats.com/leaderboard/benchmarks/gpqa" + ], + "additional_details": { + "raw_benchmark_id": "gpqa", + "raw_model_id": "kimi-k2.6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A challenging dataset of 448 multiple-choice questions written by domain experts in biology, physics, and chemistry. Questions are Google-proof and extremely difficult, with PhD experts reaching 65% accuracy.", + "metric_id": "llm_stats.gpqa.score", + "metric_name": "GPQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gpqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GPQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\",\"biology\",\"chemistry\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "226" + } + }, + "score_details": { + "score": 0.905, + "details": { + "raw_score": "0.905", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.6", + "raw_benchmark_id": "gpqa", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.6\",\"https://llm-stats.com/benchmarks/gpqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/gpqa\"]", + "raw_score_id": "gpqa::kimi-k2.6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hmmt-feb-26::hmmt-feb-26-kimi-k2.6", + "evaluation_name": "llm_stats.hmmt-feb-26", + "source_data": { + "dataset_name": "HMMT Feb 26", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.6", + "https://llm-stats.com/benchmarks/hmmt-feb-26", + "https://api.llm-stats.com/leaderboard/benchmarks/hmmt-feb-26" + ], + "additional_details": { + "raw_benchmark_id": "hmmt-feb-26", + "raw_model_id": "kimi-k2.6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "HMMT February 2026 is a math competition benchmark based on problems from the Harvard-MIT Mathematics Tournament, testing advanced mathematical problem-solving and reasoning.", + "metric_id": "llm_stats.hmmt-feb-26.score", + "metric_name": "HMMT Feb 26 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hmmt-feb-26", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HMMT Feb 26", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.927, + "details": { + "raw_score": "0.927", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.6", + "raw_benchmark_id": "hmmt-feb-26", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.6\",\"https://llm-stats.com/benchmarks/hmmt-feb-26\",\"https://api.llm-stats.com/leaderboard/benchmarks/hmmt-feb-26\"]", + "raw_score_id": "hmmt-feb-26::kimi-k2.6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "imo-answerbench::imo-answerbench-kimi-k2.6", + "evaluation_name": "llm_stats.imo-answerbench", + "source_data": { + "dataset_name": "IMO-AnswerBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.6", + "https://llm-stats.com/benchmarks/imo-answerbench", + "https://api.llm-stats.com/leaderboard/benchmarks/imo-answerbench" + ], + "additional_details": { + "raw_benchmark_id": "imo-answerbench", + "raw_model_id": "kimi-k2.6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "IMO-AnswerBench is a benchmark for evaluating mathematical reasoning capabilities on International Mathematical Olympiad (IMO) problems, focusing on answer generation and verification.", + "metric_id": "llm_stats.imo-answerbench.score", + "metric_name": "IMO-AnswerBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "imo-answerbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IMO-AnswerBench", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.86, + "details": { + "raw_score": "0.86", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.6", + "raw_benchmark_id": "imo-answerbench", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.6\",\"https://llm-stats.com/benchmarks/imo-answerbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/imo-answerbench\"]", + "raw_score_id": "imo-answerbench::kimi-k2.6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "livecodebench-v6::livecodebench-v6-kimi-k2.6", + "evaluation_name": "llm_stats.livecodebench-v6", + "source_data": { + "dataset_name": "LiveCodeBench v6", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.6", + "https://llm-stats.com/benchmarks/livecodebench-v6", + "https://api.llm-stats.com/leaderboard/benchmarks/livecodebench-v6" + ], + "additional_details": { + "raw_benchmark_id": "livecodebench-v6", + "raw_model_id": "kimi-k2.6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LiveCodeBench is a holistic and contamination-free evaluation benchmark for large language models for code. It continuously collects new problems from programming contests (LeetCode, AtCoder, CodeForces) and evaluates four different scenarios: code generation, self-repair, code execution, and test output prediction. Problems are annotated with release dates to enable evaluation on unseen problems released after a model's training cutoff.", + "metric_id": "llm_stats.livecodebench-v6.score", + "metric_name": "LiveCodeBench v6 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "livecodebench-v6", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LiveCodeBench v6", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "53" + } + }, + "score_details": { + "score": 0.896, + "details": { + "raw_score": "0.896", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.6", + "raw_benchmark_id": "livecodebench-v6", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.6\",\"https://llm-stats.com/benchmarks/livecodebench-v6\",\"https://api.llm-stats.com/leaderboard/benchmarks/livecodebench-v6\"]", + "raw_score_id": "livecodebench-v6::kimi-k2.6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mathvision::mathvision-kimi-k2.6", + "evaluation_name": "llm_stats.mathvision", + "source_data": { + "dataset_name": "MathVision", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.6", + "https://llm-stats.com/benchmarks/mathvision", + "https://api.llm-stats.com/leaderboard/benchmarks/mathvision" + ], + "additional_details": { + "raw_benchmark_id": "mathvision", + "raw_model_id": "kimi-k2.6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MATH-Vision is a dataset designed to measure multimodal mathematical reasoning capabilities. It focuses on evaluating how well models can solve mathematical problems that require both visual understanding and mathematical reasoning, bridging the gap between visual and mathematical domains.", + "metric_id": "llm_stats.mathvision.score", + "metric_name": "MathVision score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mathvision", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MathVision", + "raw_categories": "[\"math\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "31" + } + }, + "score_details": { + "score": 0.932, + "details": { + "raw_score": "0.932", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.6", + "raw_benchmark_id": "mathvision", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.6\",\"https://llm-stats.com/benchmarks/mathvision\",\"https://api.llm-stats.com/leaderboard/benchmarks/mathvision\"]", + "raw_score_id": "mathvision::kimi-k2.6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mcp-mark::mcp-mark-kimi-k2.6", + "evaluation_name": "llm_stats.mcp-mark", + "source_data": { + "dataset_name": "MCP-Mark", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.6", + "https://llm-stats.com/benchmarks/mcp-mark", + "https://api.llm-stats.com/leaderboard/benchmarks/mcp-mark" + ], + "additional_details": { + "raw_benchmark_id": "mcp-mark", + "raw_model_id": "kimi-k2.6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MCP-Mark evaluates LLMs on their ability to use Model Context Protocol (MCP) tools effectively, testing tool discovery, selection, invocation, and result interpretation across diverse MCP server scenarios.", + "metric_id": "llm_stats.mcp-mark.score", + "metric_name": "MCP-Mark score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mcp-mark", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MCP-Mark", + "raw_categories": "[\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.559, + "details": { + "raw_score": "0.559", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.6", + "raw_benchmark_id": "mcp-mark", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.6\",\"https://llm-stats.com/benchmarks/mcp-mark\",\"https://api.llm-stats.com/leaderboard/benchmarks/mcp-mark\"]", + "raw_score_id": "mcp-mark::kimi-k2.6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmmu-pro::mmmu-pro-kimi-k2.6", + "evaluation_name": "llm_stats.mmmu-pro", + "source_data": { + "dataset_name": "MMMU-Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.6", + "https://llm-stats.com/benchmarks/mmmu-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/mmmu-pro" + ], + "additional_details": { + "raw_benchmark_id": "mmmu-pro", + "raw_model_id": "kimi-k2.6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A more robust multi-discipline multimodal understanding benchmark that enhances MMMU through a three-step process: filtering text-only answerable questions, augmenting candidate options, and introducing vision-only input settings. Achieves significantly lower model performance (16.8-26.9%) compared to original MMMU, providing more rigorous evaluation that closely mimics real-world scenarios.", + "metric_id": "llm_stats.mmmu-pro.score", + "metric_name": "MMMU-Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmmu-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMMU-Pro", + "raw_categories": "[\"multimodal\",\"reasoning\",\"general\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "60" + } + }, + "score_details": { + "score": 0.801, + "details": { + "raw_score": "0.801", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.6", + "raw_benchmark_id": "mmmu-pro", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.6\",\"https://llm-stats.com/benchmarks/mmmu-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmmu-pro\"]", + "raw_score_id": "mmmu-pro::kimi-k2.6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ojbench::ojbench-kimi-k2.6", + "evaluation_name": "llm_stats.ojbench", + "source_data": { + "dataset_name": "OJBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.6", + "https://llm-stats.com/benchmarks/ojbench", + "https://api.llm-stats.com/leaderboard/benchmarks/ojbench" + ], + "additional_details": { + "raw_benchmark_id": "ojbench", + "raw_model_id": "kimi-k2.6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OJBench is a competition-level code benchmark designed to assess the competitive-level code reasoning abilities of large language models. It comprises 232 programming competition problems from NOI and ICPC, categorized into Easy, Medium, and Hard difficulty levels. The benchmark evaluates models' ability to solve complex competitive programming challenges using Python and C++.", + "metric_id": "llm_stats.ojbench.score", + "metric_name": "OJBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ojbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OJBench", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.606, + "details": { + "raw_score": "0.606", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.6", + "raw_benchmark_id": "ojbench", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.6\",\"https://llm-stats.com/benchmarks/ojbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/ojbench\"]", + "raw_score_id": "ojbench::kimi-k2.6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "osworld-verified::osworld-verified-kimi-k2.6", + "evaluation_name": "llm_stats.osworld-verified", + "source_data": { + "dataset_name": "OSWorld-Verified", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.6", + "https://llm-stats.com/benchmarks/osworld-verified", + "https://api.llm-stats.com/leaderboard/benchmarks/osworld-verified" + ], + "additional_details": { + "raw_benchmark_id": "osworld-verified", + "raw_model_id": "kimi-k2.6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OSWorld-Verified is a verified subset of OSWorld, a scalable real computer environment for multimodal agents supporting task setup, execution-based evaluation, and interactive learning across Ubuntu, Windows, and macOS.", + "metric_id": "llm_stats.osworld-verified.score", + "metric_name": "OSWorld-Verified score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "osworld-verified", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OSWorld-Verified", + "raw_categories": "[\"multimodal\",\"general\",\"agents\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.731, + "details": { + "raw_score": "0.731", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.6", + "raw_benchmark_id": "osworld-verified", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.6\",\"https://llm-stats.com/benchmarks/osworld-verified\",\"https://api.llm-stats.com/leaderboard/benchmarks/osworld-verified\"]", + "raw_score_id": "osworld-verified::kimi-k2.6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "scicode::scicode-kimi-k2.6", + "evaluation_name": "llm_stats.scicode", + "source_data": { + "dataset_name": "SciCode", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.6", + "https://llm-stats.com/benchmarks/scicode", + "https://api.llm-stats.com/leaderboard/benchmarks/scicode" + ], + "additional_details": { + "raw_benchmark_id": "scicode", + "raw_model_id": "kimi-k2.6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SciCode is a research coding benchmark curated by scientists that challenges language models to code solutions for scientific problems. It contains 338 subproblems decomposed from 80 challenging main problems across 16 natural science sub-fields including mathematics, physics, chemistry, biology, and materials science. Problems require knowledge recall, reasoning, and code synthesis skills.", + "metric_id": "llm_stats.scicode.score", + "metric_name": "SciCode score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "scicode", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SciCode", + "raw_categories": "[\"math\",\"physics\",\"reasoning\",\"biology\",\"chemistry\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.522, + "details": { + "raw_score": "0.522", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.6", + "raw_benchmark_id": "scicode", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.6\",\"https://llm-stats.com/benchmarks/scicode\",\"https://api.llm-stats.com/leaderboard/benchmarks/scicode\"]", + "raw_score_id": "scicode::kimi-k2.6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-multilingual::swe-bench-multilingual-kimi-k2.6", + "evaluation_name": "llm_stats.swe-bench-multilingual", + "source_data": { + "dataset_name": "SWE-bench Multilingual", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.6", + "https://llm-stats.com/benchmarks/swe-bench-multilingual", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-multilingual" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-multilingual", + "raw_model_id": "kimi-k2.6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multilingual benchmark for issue resolving in software engineering that covers Java, TypeScript, JavaScript, Go, Rust, C, and C++. Contains 1,632 high-quality instances carefully annotated from 2,456 candidates by 68 expert annotators, designed to evaluate Large Language Models across diverse software ecosystems beyond Python.", + "metric_id": "llm_stats.swe-bench-multilingual.score", + "metric_name": "SWE-bench Multilingual score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-multilingual", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-bench Multilingual", + "raw_categories": "[\"reasoning\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "33" + } + }, + "score_details": { + "score": 0.767, + "details": { + "raw_score": "0.767", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.6", + "raw_benchmark_id": "swe-bench-multilingual", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.6\",\"https://llm-stats.com/benchmarks/swe-bench-multilingual\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-multilingual\"]", + "raw_score_id": "swe-bench-multilingual::kimi-k2.6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-pro::swe-bench-pro-kimi-k2.6", + "evaluation_name": "llm_stats.swe-bench-pro", + "source_data": { + "dataset_name": "SWE-Bench Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.6", + "https://llm-stats.com/benchmarks/swe-bench-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-pro" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-pro", + "raw_model_id": "kimi-k2.6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SWE-Bench Pro is an advanced version of SWE-Bench that evaluates language models on complex, real-world software engineering tasks requiring extended reasoning and multi-step problem solving.", + "metric_id": "llm_stats.swe-bench-pro.score", + "metric_name": "SWE-Bench Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Bench Pro", + "raw_categories": "[\"reasoning\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "35" + } + }, + "score_details": { + "score": 0.586, + "details": { + "raw_score": "0.586", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.6", + "raw_benchmark_id": "swe-bench-pro", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.6\",\"https://llm-stats.com/benchmarks/swe-bench-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-pro\"]", + "raw_score_id": "swe-bench-pro::kimi-k2.6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-verified::swe-bench-verified-kimi-k2.6", + "evaluation_name": "llm_stats.swe-bench-verified", + "source_data": { + "dataset_name": "SWE-Bench Verified", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.6", + "https://llm-stats.com/benchmarks/swe-bench-verified", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-verified" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-verified", + "raw_model_id": "kimi-k2.6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A verified subset of 500 software engineering problems from real GitHub issues, validated by human annotators for evaluating language models' ability to resolve real-world coding issues by generating patches for Python codebases.", + "metric_id": "llm_stats.swe-bench-verified.score", + "metric_name": "SWE-Bench Verified score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-verified", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Bench Verified", + "raw_categories": "[\"reasoning\",\"frontend_development\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "103" + } + }, + "score_details": { + "score": 0.802, + "details": { + "raw_score": "0.802", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.6", + "raw_benchmark_id": "swe-bench-verified", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.6\",\"https://llm-stats.com/benchmarks/swe-bench-verified\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-verified\"]", + "raw_score_id": "swe-bench-verified::kimi-k2.6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2::terminal-bench-2-kimi-k2.6", + "evaluation_name": "llm_stats.terminal-bench-2", + "source_data": { + "dataset_name": "Terminal-Bench 2.0", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.6", + "https://llm-stats.com/benchmarks/terminal-bench-2", + "https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2" + ], + "additional_details": { + "raw_benchmark_id": "terminal-bench-2", + "raw_model_id": "kimi-k2.6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Terminal-Bench 2.0 is an updated benchmark for testing AI agents' tool use ability to operate a computer via terminal. It evaluates how well models can handle real-world, end-to-end tasks autonomously, including compiling code, training models, setting up servers, system administration, security tasks, data science workflows, and cybersecurity vulnerabilities.", + "metric_id": "llm_stats.terminal-bench-2.score", + "metric_name": "Terminal-Bench 2.0 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "terminal-bench-2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Terminal-Bench 2.0", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "49" + } + }, + "score_details": { + "score": 0.667, + "details": { + "raw_score": "0.667", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.6", + "raw_benchmark_id": "terminal-bench-2", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.6\",\"https://llm-stats.com/benchmarks/terminal-bench-2\",\"https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2\"]", + "raw_score_id": "terminal-bench-2::kimi-k2.6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "toolathlon::toolathlon-kimi-k2.6", + "evaluation_name": "llm_stats.toolathlon", + "source_data": { + "dataset_name": "Toolathlon", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.6", + "https://llm-stats.com/benchmarks/toolathlon", + "https://api.llm-stats.com/leaderboard/benchmarks/toolathlon" + ], + "additional_details": { + "raw_benchmark_id": "toolathlon", + "raw_model_id": "kimi-k2.6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Tool Decathlon is a comprehensive benchmark for evaluating AI agents' ability to use multiple tools across diverse task categories. It measures proficiency in tool selection, sequencing, and execution across ten different tool-use scenarios.", + "metric_id": "llm_stats.toolathlon.score", + "metric_name": "Toolathlon score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "toolathlon", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Toolathlon", + "raw_categories": "[\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "24" + } + }, + "score_details": { + "score": 0.5, + "details": { + "raw_score": "0.5", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.6", + "raw_benchmark_id": "toolathlon", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.6\",\"https://llm-stats.com/benchmarks/toolathlon\",\"https://api.llm-stats.com/leaderboard/benchmarks/toolathlon\"]", + "raw_score_id": "toolathlon::kimi-k2.6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "v-star::v-star-kimi-k2.6", + "evaluation_name": "llm_stats.v-star", + "source_data": { + "dataset_name": "V*", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.6", + "https://llm-stats.com/benchmarks/v-star", + "https://api.llm-stats.com/leaderboard/benchmarks/v-star" + ], + "additional_details": { + "raw_benchmark_id": "v-star", + "raw_model_id": "kimi-k2.6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A visual reasoning benchmark evaluating multimodal inference under challenging spatial and grounded tasks.", + "metric_id": "llm_stats.v-star.score", + "metric_name": "V* score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "v-star", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "V*", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.969, + "details": { + "raw_score": "0.969", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.6", + "raw_benchmark_id": "v-star", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.6\",\"https://llm-stats.com/benchmarks/v-star\",\"https://api.llm-stats.com/leaderboard/benchmarks/v-star\"]", + "raw_score_id": "v-star::kimi-k2.6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "widesearch::widesearch-kimi-k2.6", + "evaluation_name": "llm_stats.widesearch", + "source_data": { + "dataset_name": "WideSearch", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/kimi-k2.6", + "https://llm-stats.com/benchmarks/widesearch", + "https://api.llm-stats.com/leaderboard/benchmarks/widesearch" + ], + "additional_details": { + "raw_benchmark_id": "widesearch", + "raw_model_id": "kimi-k2.6", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WideSearch is an agentic search benchmark that evaluates models' ability to perform broad, parallel search operations across multiple sources. It tests wide-coverage information retrieval and synthesis capabilities.", + "metric_id": "llm_stats.widesearch.score", + "metric_name": "WideSearch score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "widesearch", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WideSearch", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.808, + "details": { + "raw_score": "0.808", + "raw_score_field": "score", + "raw_model_id": "kimi-k2.6", + "raw_benchmark_id": "widesearch", + "source_urls_json": "[\"https://llm-stats.com/models/kimi-k2.6\",\"https://llm-stats.com/benchmarks/widesearch\",\"https://api.llm-stats.com/leaderboard/benchmarks/widesearch\"]", + "raw_score_id": "widesearch::kimi-k2.6", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/nous-research/hermes-3-70b/7ddced61-03ed-4d30-adf6-8558ef30a6c3.json b/data/llm_stats/nous-research/hermes-3-70b/7ddced61-03ed-4d30-adf6-8558ef30a6c3.json new file mode 100644 index 0000000000000000000000000000000000000000..a96bacd434d4c9a1862d92d53f1068447dd0ae2e --- /dev/null +++ b/data/llm_stats/nous-research/hermes-3-70b/7ddced61-03ed-4d30-adf6-8558ef30a6c3.json @@ -0,0 +1,671 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/nous-research_hermes-3-70b/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Hermes 3 70B", + "id": "nous-research/hermes-3-70b", + "developer": "nous-research", + "additional_details": { + "raw_id": "hermes-3-70b", + "raw_name": "Hermes 3 70B", + "raw_model_id": "hermes-3-70b", + "raw_model_name": "Hermes 3 70B", + "raw_organization_id": "nous-research", + "raw_organization_name": "Nous Research", + "raw_release_date": "2024-08-15", + "raw_announcement_date": "2024-08-15", + "raw_multimodal": "false", + "raw_param_count": "70000000000", + "raw_provider_slug": "nous-research", + "raw_provider_name": "Nous Research" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "agieval::agieval-hermes-3-70b", + "evaluation_name": "llm_stats.agieval", + "source_data": { + "dataset_name": "AGIEval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/hermes-3-70b", + "https://llm-stats.com/benchmarks/agieval", + "https://api.llm-stats.com/leaderboard/benchmarks/agieval" + ], + "additional_details": { + "raw_benchmark_id": "agieval", + "raw_model_id": "hermes-3-70b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A human-centric benchmark for evaluating foundation models on standardized exams including college entrance exams (Gaokao, SAT), law school admission tests (LSAT), math competitions, lawyer qualification tests, and civil service exams. Contains 20 tasks (18 multiple-choice, 2 cloze) designed to assess understanding, knowledge, reasoning, and calculation abilities in real-world academic and professional contexts.", + "metric_id": "llm_stats.agieval.score", + "metric_name": "AGIEval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "agieval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AGIEval", + "raw_categories": "[\"legal\",\"math\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.5618, + "details": { + "raw_score": "0.5618", + "raw_score_field": "score", + "raw_model_id": "hermes-3-70b", + "raw_benchmark_id": "agieval", + "source_urls_json": "[\"https://llm-stats.com/models/hermes-3-70b\",\"https://llm-stats.com/benchmarks/agieval\",\"https://api.llm-stats.com/leaderboard/benchmarks/agieval\"]", + "raw_score_id": "agieval::hermes-3-70b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "arc-e::arc-e-hermes-3-70b", + "evaluation_name": "llm_stats.arc-e", + "source_data": { + "dataset_name": "ARC-E", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/hermes-3-70b", + "https://llm-stats.com/benchmarks/arc-e", + "https://api.llm-stats.com/leaderboard/benchmarks/arc-e" + ], + "additional_details": { + "raw_benchmark_id": "arc-e", + "raw_model_id": "hermes-3-70b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ARC-E (AI2 Reasoning Challenge - Easy Set) is a subset of grade-school level, multiple-choice science questions that requires knowledge and reasoning capabilities. Part of the AI2 Reasoning Challenge dataset containing 5,197 questions that test scientific reasoning and factual knowledge. The Easy Set contains questions that are answerable by retrieval-based and word co-occurrence algorithms, making them more accessible than the Challenge Set.", + "metric_id": "llm_stats.arc-e.score", + "metric_name": "ARC-E score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arc-e", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ARC-E", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.8295, + "details": { + "raw_score": "0.8295", + "raw_score_field": "score", + "raw_model_id": "hermes-3-70b", + "raw_benchmark_id": "arc-e", + "source_urls_json": "[\"https://llm-stats.com/models/hermes-3-70b\",\"https://llm-stats.com/benchmarks/arc-e\",\"https://api.llm-stats.com/leaderboard/benchmarks/arc-e\"]", + "raw_score_id": "arc-e::hermes-3-70b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bbh::bbh-hermes-3-70b", + "evaluation_name": "llm_stats.bbh", + "source_data": { + "dataset_name": "BBH", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/hermes-3-70b", + "https://llm-stats.com/benchmarks/bbh", + "https://api.llm-stats.com/leaderboard/benchmarks/bbh" + ], + "additional_details": { + "raw_benchmark_id": "bbh", + "raw_model_id": "hermes-3-70b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Big-Bench Hard (BBH) is a suite of 23 challenging tasks selected from BIG-Bench for which prior language model evaluations did not outperform the average human-rater. These tasks require multi-step reasoning across diverse domains including arithmetic, logical reasoning, reading comprehension, and commonsense reasoning. The benchmark was designed to test capabilities believed to be beyond current language models and focuses on evaluating complex reasoning skills including temporal understanding, spatial reasoning, causal understanding, and deductive logical reasoning.", + "metric_id": "llm_stats.bbh.score", + "metric_name": "BBH score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bbh", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BBH", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.6782, + "details": { + "raw_score": "0.6782", + "raw_score_field": "score", + "raw_model_id": "hermes-3-70b", + "raw_benchmark_id": "bbh", + "source_urls_json": "[\"https://llm-stats.com/models/hermes-3-70b\",\"https://llm-stats.com/benchmarks/bbh\",\"https://api.llm-stats.com/leaderboard/benchmarks/bbh\"]", + "raw_score_id": "bbh::hermes-3-70b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "boolq::boolq-hermes-3-70b", + "evaluation_name": "llm_stats.boolq", + "source_data": { + "dataset_name": "BoolQ", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/hermes-3-70b", + "https://llm-stats.com/benchmarks/boolq", + "https://api.llm-stats.com/leaderboard/benchmarks/boolq" + ], + "additional_details": { + "raw_benchmark_id": "boolq", + "raw_model_id": "hermes-3-70b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BoolQ is a reading comprehension dataset for yes/no questions containing 15,942 naturally occurring examples. Each example consists of a question, passage, and boolean answer, where questions are generated in unprompted and unconstrained settings. The dataset challenges models with complex, non-factoid information requiring entailment-like inference to solve.", + "metric_id": "llm_stats.boolq.score", + "metric_name": "BoolQ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "boolq", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BoolQ", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.8804, + "details": { + "raw_score": "0.8804", + "raw_score_field": "score", + "raw_model_id": "hermes-3-70b", + "raw_benchmark_id": "boolq", + "source_urls_json": "[\"https://llm-stats.com/models/hermes-3-70b\",\"https://llm-stats.com/benchmarks/boolq\",\"https://api.llm-stats.com/leaderboard/benchmarks/boolq\"]", + "raw_score_id": "boolq::hermes-3-70b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hellaswag::hellaswag-hermes-3-70b", + "evaluation_name": "llm_stats.hellaswag", + "source_data": { + "dataset_name": "HellaSwag", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/hermes-3-70b", + "https://llm-stats.com/benchmarks/hellaswag", + "https://api.llm-stats.com/leaderboard/benchmarks/hellaswag" + ], + "additional_details": { + "raw_benchmark_id": "hellaswag", + "raw_model_id": "hermes-3-70b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A challenging commonsense natural language inference dataset that uses Adversarial Filtering to create questions trivial for humans (>95% accuracy) but difficult for state-of-the-art models, requiring completion of sentence endings based on physical situations and everyday activities", + "metric_id": "llm_stats.hellaswag.score", + "metric_name": "HellaSwag score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hellaswag", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HellaSwag", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.8819, + "details": { + "raw_score": "0.8819", + "raw_score_field": "score", + "raw_model_id": "hermes-3-70b", + "raw_benchmark_id": "hellaswag", + "source_urls_json": "[\"https://llm-stats.com/models/hermes-3-70b\",\"https://llm-stats.com/benchmarks/hellaswag\",\"https://api.llm-stats.com/leaderboard/benchmarks/hellaswag\"]", + "raw_score_id": "hellaswag::hermes-3-70b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ifbench::ifbench-hermes-3-70b", + "evaluation_name": "llm_stats.ifbench", + "source_data": { + "dataset_name": "IFBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/hermes-3-70b", + "https://llm-stats.com/benchmarks/ifbench", + "https://api.llm-stats.com/leaderboard/benchmarks/ifbench" + ], + "additional_details": { + "raw_benchmark_id": "ifbench", + "raw_model_id": "hermes-3-70b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Instruction Following Benchmark evaluating model's ability to follow complex instructions", + "metric_id": "llm_stats.ifbench.score", + "metric_name": "IFBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ifbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IFBench", + "raw_categories": "[\"instruction_following\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.8121, + "details": { + "raw_score": "0.8121", + "raw_score_field": "score", + "raw_model_id": "hermes-3-70b", + "raw_benchmark_id": "ifbench", + "source_urls_json": "[\"https://llm-stats.com/models/hermes-3-70b\",\"https://llm-stats.com/benchmarks/ifbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/ifbench\"]", + "raw_score_id": "ifbench::hermes-3-70b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mt-bench::mt-bench-hermes-3-70b", + "evaluation_name": "llm_stats.mt-bench", + "source_data": { + "dataset_name": "MT-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/hermes-3-70b", + "https://llm-stats.com/benchmarks/mt-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/mt-bench" + ], + "additional_details": { + "raw_benchmark_id": "mt-bench", + "raw_model_id": "hermes-3-70b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MT-Bench is a challenging multi-turn benchmark that measures the ability of large language models to engage in coherent, informative, and engaging conversations. It uses strong LLMs as judges for scalable and explainable evaluation of multi-turn dialogue capabilities.", + "metric_id": "llm_stats.mt-bench.score", + "metric_name": "MT-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "raw_benchmark_id": "mt-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_percent", + "raw_name": "MT-Bench", + "raw_categories": "[\"reasoning\",\"roleplay\",\"general\",\"communication\",\"creativity\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 8.99, + "details": { + "raw_score": "8.99", + "raw_score_field": "score", + "raw_model_id": "hermes-3-70b", + "raw_benchmark_id": "mt-bench", + "source_urls_json": "[\"https://llm-stats.com/models/hermes-3-70b\",\"https://llm-stats.com/benchmarks/mt-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/mt-bench\"]", + "raw_score_id": "mt-bench::hermes-3-70b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "musr::musr-hermes-3-70b", + "evaluation_name": "llm_stats.musr", + "source_data": { + "dataset_name": "MuSR", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/hermes-3-70b", + "https://llm-stats.com/benchmarks/musr", + "https://api.llm-stats.com/leaderboard/benchmarks/musr" + ], + "additional_details": { + "raw_benchmark_id": "musr", + "raw_model_id": "hermes-3-70b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MuSR (Multistep Soft Reasoning) is a benchmark for evaluating language models on multistep soft reasoning tasks specified in natural language narratives. Created through a neurosymbolic synthetic-to-natural generation algorithm, it generates complex reasoning scenarios like murder mysteries roughly 1000 words in length that challenge current LLMs including GPT-4. The benchmark tests chain-of-thought reasoning capabilities across domains involving commonsense reasoning about physical and social situations.", + "metric_id": "llm_stats.musr.score", + "metric_name": "MuSR score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "musr", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MuSR", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.5067, + "details": { + "raw_score": "0.5067", + "raw_score_field": "score", + "raw_model_id": "hermes-3-70b", + "raw_benchmark_id": "musr", + "source_urls_json": "[\"https://llm-stats.com/models/hermes-3-70b\",\"https://llm-stats.com/benchmarks/musr\",\"https://api.llm-stats.com/leaderboard/benchmarks/musr\"]", + "raw_score_id": "musr::hermes-3-70b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "openbookqa::openbookqa-hermes-3-70b", + "evaluation_name": "llm_stats.openbookqa", + "source_data": { + "dataset_name": "OpenBookQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/hermes-3-70b", + "https://llm-stats.com/benchmarks/openbookqa", + "https://api.llm-stats.com/leaderboard/benchmarks/openbookqa" + ], + "additional_details": { + "raw_benchmark_id": "openbookqa", + "raw_model_id": "hermes-3-70b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OpenBookQA is a question-answering dataset modeled after open book exams for assessing human understanding. It contains 5,957 multiple-choice elementary-level science questions that probe understanding of 1,326 core science facts and their application to novel situations, requiring combination of open book facts with broad common knowledge through multi-hop reasoning.", + "metric_id": "llm_stats.openbookqa.score", + "metric_name": "OpenBookQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "openbookqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OpenBookQA", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.494, + "details": { + "raw_score": "0.494", + "raw_score_field": "score", + "raw_model_id": "hermes-3-70b", + "raw_benchmark_id": "openbookqa", + "source_urls_json": "[\"https://llm-stats.com/models/hermes-3-70b\",\"https://llm-stats.com/benchmarks/openbookqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/openbookqa\"]", + "raw_score_id": "openbookqa::hermes-3-70b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "piqa::piqa-hermes-3-70b", + "evaluation_name": "llm_stats.piqa", + "source_data": { + "dataset_name": "PIQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/hermes-3-70b", + "https://llm-stats.com/benchmarks/piqa", + "https://api.llm-stats.com/leaderboard/benchmarks/piqa" + ], + "additional_details": { + "raw_benchmark_id": "piqa", + "raw_model_id": "hermes-3-70b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "PIQA (Physical Interaction: Question Answering) is a benchmark dataset for physical commonsense reasoning in natural language. It tests AI systems' ability to answer questions requiring physical world knowledge through multiple choice questions with everyday situations, focusing on atypical solutions inspired by instructables.com. The dataset contains 21,000 multiple choice questions where models must choose the most appropriate solution for physical interactions.", + "metric_id": "llm_stats.piqa.score", + "metric_name": "PIQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "piqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PIQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.8444, + "details": { + "raw_score": "0.8444", + "raw_score_field": "score", + "raw_model_id": "hermes-3-70b", + "raw_benchmark_id": "piqa", + "source_urls_json": "[\"https://llm-stats.com/models/hermes-3-70b\",\"https://llm-stats.com/benchmarks/piqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/piqa\"]", + "raw_score_id": "piqa::hermes-3-70b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "truthfulqa::truthfulqa-hermes-3-70b", + "evaluation_name": "llm_stats.truthfulqa", + "source_data": { + "dataset_name": "TruthfulQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/hermes-3-70b", + "https://llm-stats.com/benchmarks/truthfulqa", + "https://api.llm-stats.com/leaderboard/benchmarks/truthfulqa" + ], + "additional_details": { + "raw_benchmark_id": "truthfulqa", + "raw_model_id": "hermes-3-70b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "TruthfulQA is a benchmark to measure whether language models are truthful in generating answers to questions. It comprises 817 questions that span 38 categories, including health, law, finance and politics. The questions are crafted such that some humans would answer falsely due to a false belief or misconception, testing models' ability to avoid generating false answers learned from human texts.", + "metric_id": "llm_stats.truthfulqa.score", + "metric_name": "TruthfulQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "truthfulqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TruthfulQA", + "raw_categories": "[\"legal\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.6329, + "details": { + "raw_score": "0.6329", + "raw_score_field": "score", + "raw_model_id": "hermes-3-70b", + "raw_benchmark_id": "truthfulqa", + "source_urls_json": "[\"https://llm-stats.com/models/hermes-3-70b\",\"https://llm-stats.com/benchmarks/truthfulqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/truthfulqa\"]", + "raw_score_id": "truthfulqa::hermes-3-70b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "winogrande::winogrande-hermes-3-70b", + "evaluation_name": "llm_stats.winogrande", + "source_data": { + "dataset_name": "Winogrande", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/hermes-3-70b", + "https://llm-stats.com/benchmarks/winogrande", + "https://api.llm-stats.com/leaderboard/benchmarks/winogrande" + ], + "additional_details": { + "raw_benchmark_id": "winogrande", + "raw_model_id": "hermes-3-70b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WinoGrande: An Adversarial Winograd Schema Challenge at Scale. A large-scale dataset of 44,000 pronoun resolution problems designed to test machine commonsense reasoning. Uses adversarial filtering to reduce spurious biases and provides a more robust evaluation of whether AI systems truly understand commonsense or exploit statistical shortcuts. Current best AI methods achieve 59.4-79.1% accuracy, significantly below human performance of 94.0%.", + "metric_id": "llm_stats.winogrande.score", + "metric_name": "Winogrande score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "winogrande", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Winogrande", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "22" + } + }, + "score_details": { + "score": 0.8319, + "details": { + "raw_score": "0.8319", + "raw_score_field": "score", + "raw_model_id": "hermes-3-70b", + "raw_benchmark_id": "winogrande", + "source_urls_json": "[\"https://llm-stats.com/models/hermes-3-70b\",\"https://llm-stats.com/benchmarks/winogrande\",\"https://api.llm-stats.com/leaderboard/benchmarks/winogrande\"]", + "raw_score_id": "winogrande::hermes-3-70b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/nvidia/nemotron-3-super-120b-a12b/a1018b14-4c4e-4d50-9212-e2929f418aca.json b/data/llm_stats/nvidia/nemotron-3-super-120b-a12b/a1018b14-4c4e-4d50-9212-e2929f418aca.json new file mode 100644 index 0000000000000000000000000000000000000000..1873d731058284e8e6bc101ce1f6e43e79d25f75 --- /dev/null +++ b/data/llm_stats/nvidia/nemotron-3-super-120b-a12b/a1018b14-4c4e-4d50-9212-e2929f418aca.json @@ -0,0 +1,671 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/nvidia_nemotron-3-super-120b-a12b/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Nemotron 3 Super (120B A12B)", + "id": "nvidia/nemotron-3-super-120b-a12b", + "developer": "nvidia", + "additional_details": { + "raw_id": "nemotron-3-super-120b-a12b", + "raw_name": "Nemotron 3 Super (120B A12B)", + "raw_model_id": "nemotron-3-super-120b-a12b", + "raw_model_name": "Nemotron 3 Super (120B A12B)", + "raw_organization_id": "nvidia", + "raw_organization_name": "NVIDIA", + "raw_release_date": "2026-03-11", + "raw_announcement_date": "2026-03-11", + "raw_multimodal": "false", + "raw_param_count": "120000000000", + "raw_provider_slug": "nvidia", + "raw_provider_name": "NVIDIA" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aa-lcr::aa-lcr-nemotron-3-super-120b-a12b", + "evaluation_name": "llm_stats.aa-lcr", + "source_data": { + "dataset_name": "AA-LCR", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-super-120b-a12b", + "https://llm-stats.com/benchmarks/aa-lcr", + "https://api.llm-stats.com/leaderboard/benchmarks/aa-lcr" + ], + "additional_details": { + "raw_benchmark_id": "aa-lcr", + "raw_model_id": "nemotron-3-super-120b-a12b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Agent Arena Long Context Reasoning benchmark", + "metric_id": "llm_stats.aa-lcr.score", + "metric_name": "AA-LCR score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aa-lcr", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AA-LCR", + "raw_categories": "[\"long_context\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "14" + } + }, + "score_details": { + "score": 0.5831, + "details": { + "raw_score": "0.5831", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-super-120b-a12b", + "raw_benchmark_id": "aa-lcr", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-super-120b-a12b\",\"https://llm-stats.com/benchmarks/aa-lcr\",\"https://api.llm-stats.com/leaderboard/benchmarks/aa-lcr\"]", + "raw_score_id": "aa-lcr::nemotron-3-super-120b-a12b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "arena-hard-v2::arena-hard-v2-nemotron-3-super-120b-a12b", + "evaluation_name": "llm_stats.arena-hard-v2", + "source_data": { + "dataset_name": "Arena-Hard v2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-super-120b-a12b", + "https://llm-stats.com/benchmarks/arena-hard-v2", + "https://api.llm-stats.com/leaderboard/benchmarks/arena-hard-v2" + ], + "additional_details": { + "raw_benchmark_id": "arena-hard-v2", + "raw_model_id": "nemotron-3-super-120b-a12b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Arena-Hard-Auto v2 is a challenging benchmark consisting of 500 carefully curated prompts sourced from Chatbot Arena and WildChat-1M, designed to evaluate large language models on real-world user queries. The benchmark covers diverse domains including open-ended software engineering problems, mathematics, creative writing, and technical problem-solving. It uses LLM-as-a-Judge for automatic evaluation, achieving 98.6% correlation with human preference rankings while providing 3x higher separation of model performances compared to MT-Bench. The benchmark emphasizes prompt specificity, complexity, and domain knowledge to better distinguish between model capabilities.", + "metric_id": "llm_stats.arena-hard-v2.score", + "metric_name": "Arena-Hard v2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arena-hard-v2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Arena-Hard v2", + "raw_categories": "[\"reasoning\",\"general\",\"creativity\",\"writing\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.7388, + "details": { + "raw_score": "0.7388", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-super-120b-a12b", + "raw_benchmark_id": "arena-hard-v2", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-super-120b-a12b\",\"https://llm-stats.com/benchmarks/arena-hard-v2\",\"https://api.llm-stats.com/leaderboard/benchmarks/arena-hard-v2\"]", + "raw_score_id": "arena-hard-v2::nemotron-3-super-120b-a12b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bird-sql-dev::bird-sql-dev-nemotron-3-super-120b-a12b", + "evaluation_name": "llm_stats.bird-sql-dev", + "source_data": { + "dataset_name": "Bird-SQL (dev)", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-super-120b-a12b", + "https://llm-stats.com/benchmarks/bird-sql-dev", + "https://api.llm-stats.com/leaderboard/benchmarks/bird-sql-(dev)" + ], + "additional_details": { + "raw_benchmark_id": "bird-sql-(dev)", + "raw_model_id": "nemotron-3-super-120b-a12b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BIRD (BIg Bench for LaRge-scale Database Grounded Text-to-SQLs) is a comprehensive text-to-SQL benchmark containing 12,751 question-SQL pairs across 95 databases (33.4 GB total) spanning 37+ professional domains. It evaluates large language models' ability to convert natural language to executable SQL queries in real-world scenarios with complex database schemas and dirty data.", + "metric_id": "llm_stats.bird-sql-dev.score", + "metric_name": "Bird-SQL (dev) score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bird-sql-(dev)", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Bird-SQL (dev)", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.418, + "details": { + "raw_score": "0.418", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-super-120b-a12b", + "raw_benchmark_id": "bird-sql-(dev)", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-super-120b-a12b\",\"https://llm-stats.com/benchmarks/bird-sql-dev\",\"https://api.llm-stats.com/leaderboard/benchmarks/bird-sql-(dev)\"]", + "raw_score_id": "bird-sql-(dev)::nemotron-3-super-120b-a12b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hmmt-2025::hmmt-2025-nemotron-3-super-120b-a12b", + "evaluation_name": "llm_stats.hmmt-2025", + "source_data": { + "dataset_name": "HMMT 2025", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-super-120b-a12b", + "https://llm-stats.com/benchmarks/hmmt-2025", + "https://api.llm-stats.com/leaderboard/benchmarks/hmmt-2025" + ], + "additional_details": { + "raw_benchmark_id": "hmmt-2025", + "raw_model_id": "nemotron-3-super-120b-a12b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Harvard-MIT Mathematics Tournament 2025 - A prestigious student-organized mathematics competition for high school students featuring two tournaments (November 2025 at MIT and February 2026 at Harvard) with individual tests, team rounds, and guts rounds", + "metric_id": "llm_stats.hmmt-2025.score", + "metric_name": "HMMT 2025 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hmmt-2025", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HMMT 2025", + "raw_categories": "[\"math\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "33" + } + }, + "score_details": { + "score": 0.9473, + "details": { + "raw_score": "0.9473", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-super-120b-a12b", + "raw_benchmark_id": "hmmt-2025", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-super-120b-a12b\",\"https://llm-stats.com/benchmarks/hmmt-2025\",\"https://api.llm-stats.com/leaderboard/benchmarks/hmmt-2025\"]", + "raw_score_id": "hmmt-2025::nemotron-3-super-120b-a12b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ifbench::ifbench-nemotron-3-super-120b-a12b", + "evaluation_name": "llm_stats.ifbench", + "source_data": { + "dataset_name": "IFBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-super-120b-a12b", + "https://llm-stats.com/benchmarks/ifbench", + "https://api.llm-stats.com/leaderboard/benchmarks/ifbench" + ], + "additional_details": { + "raw_benchmark_id": "ifbench", + "raw_model_id": "nemotron-3-super-120b-a12b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Instruction Following Benchmark evaluating model's ability to follow complex instructions", + "metric_id": "llm_stats.ifbench.score", + "metric_name": "IFBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ifbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IFBench", + "raw_categories": "[\"instruction_following\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.7256, + "details": { + "raw_score": "0.7256", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-super-120b-a12b", + "raw_benchmark_id": "ifbench", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-super-120b-a12b\",\"https://llm-stats.com/benchmarks/ifbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/ifbench\"]", + "raw_score_id": "ifbench::nemotron-3-super-120b-a12b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "livecodebench::livecodebench-nemotron-3-super-120b-a12b", + "evaluation_name": "llm_stats.livecodebench", + "source_data": { + "dataset_name": "LiveCodeBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-super-120b-a12b", + "https://llm-stats.com/benchmarks/livecodebench", + "https://api.llm-stats.com/leaderboard/benchmarks/livecodebench" + ], + "additional_details": { + "raw_benchmark_id": "livecodebench", + "raw_model_id": "nemotron-3-super-120b-a12b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LiveCodeBench is a holistic and contamination-free evaluation benchmark for large language models for code. It continuously collects new problems from programming contests (LeetCode, AtCoder, CodeForces) and evaluates four different scenarios: code generation, self-repair, code execution, and test output prediction. Problems are annotated with release dates to enable evaluation on unseen problems released after a model's training cutoff.", + "metric_id": "llm_stats.livecodebench.score", + "metric_name": "LiveCodeBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "livecodebench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LiveCodeBench", + "raw_categories": "[\"reasoning\",\"general\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "73" + } + }, + "score_details": { + "score": 0.8119, + "details": { + "raw_score": "0.8119", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-super-120b-a12b", + "raw_benchmark_id": "livecodebench", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-super-120b-a12b\",\"https://llm-stats.com/benchmarks/livecodebench\",\"https://api.llm-stats.com/leaderboard/benchmarks/livecodebench\"]", + "raw_score_id": "livecodebench::nemotron-3-super-120b-a12b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu-prox::mmlu-prox-nemotron-3-super-120b-a12b", + "evaluation_name": "llm_stats.mmlu-prox", + "source_data": { + "dataset_name": "MMLU-ProX", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-super-120b-a12b", + "https://llm-stats.com/benchmarks/mmlu-prox", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu-prox" + ], + "additional_details": { + "raw_benchmark_id": "mmlu-prox", + "raw_model_id": "nemotron-3-super-120b-a12b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Extended version of MMLU-Pro providing additional challenging multiple-choice questions for evaluating language models across diverse academic and professional domains. Built on the foundation of the Massive Multitask Language Understanding benchmark framework.", + "metric_id": "llm_stats.mmlu-prox.score", + "metric_name": "MMLU-ProX score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu-prox", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU-ProX", + "raw_categories": "[\"language\",\"legal\",\"math\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "32" + } + }, + "score_details": { + "score": 0.7936, + "details": { + "raw_score": "0.7936", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-super-120b-a12b", + "raw_benchmark_id": "mmlu-prox", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-super-120b-a12b\",\"https://llm-stats.com/benchmarks/mmlu-prox\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu-prox\"]", + "raw_score_id": "mmlu-prox::nemotron-3-super-120b-a12b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "multichallenge::multichallenge-nemotron-3-super-120b-a12b", + "evaluation_name": "llm_stats.multichallenge", + "source_data": { + "dataset_name": "Multi-Challenge", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-super-120b-a12b", + "https://llm-stats.com/benchmarks/multichallenge", + "https://api.llm-stats.com/leaderboard/benchmarks/multichallenge" + ], + "additional_details": { + "raw_benchmark_id": "multichallenge", + "raw_model_id": "nemotron-3-super-120b-a12b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MultiChallenge is a realistic multi-turn conversation evaluation benchmark that challenges frontier LLMs across four key categories: instruction retention (maintaining instructions throughout conversations), inference memory (recalling and connecting details from previous turns), reliable versioned editing (adapting to evolving instructions during collaborative editing), and self-coherence (avoiding contradictions in responses). The benchmark evaluates models on sustained, contextually complex dialogues across diverse topics including travel planning, technical documentation, and professional communication.", + "metric_id": "llm_stats.multichallenge.score", + "metric_name": "Multi-Challenge score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "multichallenge", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Multi-Challenge", + "raw_categories": "[\"reasoning\",\"communication\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "29" + } + }, + "score_details": { + "score": 0.5523, + "details": { + "raw_score": "0.5523", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-super-120b-a12b", + "raw_benchmark_id": "multichallenge", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-super-120b-a12b\",\"https://llm-stats.com/benchmarks/multichallenge\",\"https://api.llm-stats.com/leaderboard/benchmarks/multichallenge\"]", + "raw_score_id": "multichallenge::nemotron-3-super-120b-a12b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ruler::ruler-nemotron-3-super-120b-a12b", + "evaluation_name": "llm_stats.ruler", + "source_data": { + "dataset_name": "RULER", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-super-120b-a12b", + "https://llm-stats.com/benchmarks/ruler", + "https://api.llm-stats.com/leaderboard/benchmarks/ruler" + ], + "additional_details": { + "raw_benchmark_id": "ruler", + "raw_model_id": "nemotron-3-super-120b-a12b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "RULER v1 is a synthetic long-context benchmark for measuring how model quality degrades as input length increases. This packaging follows the public standalone NVIDIA RULER implementation with 13 official tasks spanning retrieval, multi-hop tracing, aggregation, and QA.", + "metric_id": "llm_stats.ruler.score", + "metric_name": "RULER score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ruler", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "RULER", + "raw_categories": "[\"long_context\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.9175, + "details": { + "raw_score": "0.9175", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-super-120b-a12b", + "raw_benchmark_id": "ruler", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-super-120b-a12b\",\"https://llm-stats.com/benchmarks/ruler\",\"https://api.llm-stats.com/leaderboard/benchmarks/ruler\"]", + "raw_score_id": "ruler::nemotron-3-super-120b-a12b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "scicode::scicode-nemotron-3-super-120b-a12b", + "evaluation_name": "llm_stats.scicode", + "source_data": { + "dataset_name": "SciCode", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-super-120b-a12b", + "https://llm-stats.com/benchmarks/scicode", + "https://api.llm-stats.com/leaderboard/benchmarks/scicode" + ], + "additional_details": { + "raw_benchmark_id": "scicode", + "raw_model_id": "nemotron-3-super-120b-a12b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SciCode is a research coding benchmark curated by scientists that challenges language models to code solutions for scientific problems. It contains 338 subproblems decomposed from 80 challenging main problems across 16 natural science sub-fields including mathematics, physics, chemistry, biology, and materials science. Problems require knowledge recall, reasoning, and code synthesis skills.", + "metric_id": "llm_stats.scicode.score", + "metric_name": "SciCode score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "scicode", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SciCode", + "raw_categories": "[\"math\",\"physics\",\"reasoning\",\"biology\",\"chemistry\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.4205, + "details": { + "raw_score": "0.4205", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-super-120b-a12b", + "raw_benchmark_id": "scicode", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-super-120b-a12b\",\"https://llm-stats.com/benchmarks/scicode\",\"https://api.llm-stats.com/leaderboard/benchmarks/scicode\"]", + "raw_score_id": "scicode::nemotron-3-super-120b-a12b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "tau2-airline::tau2-airline-nemotron-3-super-120b-a12b", + "evaluation_name": "llm_stats.tau2-airline", + "source_data": { + "dataset_name": "Tau2 Airline", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-super-120b-a12b", + "https://llm-stats.com/benchmarks/tau2-airline", + "https://api.llm-stats.com/leaderboard/benchmarks/tau2-airline" + ], + "additional_details": { + "raw_benchmark_id": "tau2-airline", + "raw_model_id": "nemotron-3-super-120b-a12b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "TAU2 airline domain benchmark for evaluating conversational agents in dual-control environments where both AI agents and users interact with tools in airline customer service scenarios. Tests agent coordination, communication, and ability to guide user actions in tasks like flight booking, modifications, cancellations, and refunds.", + "metric_id": "llm_stats.tau2-airline.score", + "metric_name": "Tau2 Airline score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "tau2-airline", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Tau2 Airline", + "raw_categories": "[\"reasoning\",\"communication\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.5625, + "details": { + "raw_score": "0.5625", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-super-120b-a12b", + "raw_benchmark_id": "tau2-airline", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-super-120b-a12b\",\"https://llm-stats.com/benchmarks/tau2-airline\",\"https://api.llm-stats.com/leaderboard/benchmarks/tau2-airline\"]", + "raw_score_id": "tau2-airline::nemotron-3-super-120b-a12b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "wmt24::wmt24-nemotron-3-super-120b-a12b", + "evaluation_name": "llm_stats.wmt24", + "source_data": { + "dataset_name": "WMT24++", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-super-120b-a12b", + "https://llm-stats.com/benchmarks/wmt24", + "https://api.llm-stats.com/leaderboard/benchmarks/wmt24++" + ], + "additional_details": { + "raw_benchmark_id": "wmt24++", + "raw_model_id": "nemotron-3-super-120b-a12b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WMT24++ is a comprehensive multilingual machine translation benchmark that expands the WMT24 dataset to cover 55 languages and dialects. It includes human-written references and post-edits across four domains (literary, news, social, and speech) to evaluate machine translation systems and large language models across diverse linguistic contexts.", + "metric_id": "llm_stats.wmt24.score", + "metric_name": "WMT24++ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "wmt24++", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WMT24++", + "raw_categories": "[\"language\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.8667, + "details": { + "raw_score": "0.8667", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-super-120b-a12b", + "raw_benchmark_id": "wmt24++", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-super-120b-a12b\",\"https://llm-stats.com/benchmarks/wmt24\",\"https://api.llm-stats.com/leaderboard/benchmarks/wmt24++\"]", + "raw_score_id": "wmt24++::nemotron-3-super-120b-a12b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/nvidia/nemotron-3-ultra-550b-a55b/07bd2825-afe0-46be-9dd0-3f207cfa047b.json b/data/llm_stats/nvidia/nemotron-3-ultra-550b-a55b/07bd2825-afe0-46be-9dd0-3f207cfa047b.json new file mode 100644 index 0000000000000000000000000000000000000000..1875cd26c0f99bdc4d724386542b4594a6a6cfc4 --- /dev/null +++ b/data/llm_stats/nvidia/nemotron-3-ultra-550b-a55b/07bd2825-afe0-46be-9dd0-3f207cfa047b.json @@ -0,0 +1,1139 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/nvidia_nemotron-3-ultra-550b-a55b/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Nemotron 3 Ultra (550B A55B)", + "id": "nvidia/nemotron-3-ultra-550b-a55b", + "developer": "nvidia", + "additional_details": { + "raw_id": "nemotron-3-ultra-550b-a55b", + "raw_name": "Nemotron 3 Ultra (550B A55B)", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "raw_model_name": "Nemotron 3 Ultra (550B A55B)", + "raw_organization_id": "nvidia", + "raw_organization_name": "NVIDIA", + "raw_release_date": "2026-06-04", + "raw_announcement_date": "2026-06-04", + "raw_multimodal": "false", + "raw_param_count": "550000000000", + "raw_provider_slug": "nvidia", + "raw_provider_name": "NVIDIA" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aa-lcr::aa-lcr-nemotron-3-ultra-550b-a55b", + "evaluation_name": "llm_stats.aa-lcr", + "source_data": { + "dataset_name": "AA-LCR", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-ultra-550b-a55b", + "https://llm-stats.com/benchmarks/aa-lcr", + "https://api.llm-stats.com/leaderboard/benchmarks/aa-lcr" + ], + "additional_details": { + "raw_benchmark_id": "aa-lcr", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Agent Arena Long Context Reasoning benchmark", + "metric_id": "llm_stats.aa-lcr.score", + "metric_name": "AA-LCR score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aa-lcr", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AA-LCR", + "raw_categories": "[\"long_context\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "14" + } + }, + "score_details": { + "score": 0.654, + "details": { + "raw_score": "0.654", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "raw_benchmark_id": "aa-lcr", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-ultra-550b-a55b\",\"https://llm-stats.com/benchmarks/aa-lcr\",\"https://api.llm-stats.com/leaderboard/benchmarks/aa-lcr\"]", + "raw_score_id": "aa-lcr::nemotron-3-ultra-550b-a55b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "apex::apex-nemotron-3-ultra-550b-a55b", + "evaluation_name": "llm_stats.apex", + "source_data": { + "dataset_name": "Apex", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-ultra-550b-a55b", + "https://llm-stats.com/benchmarks/apex", + "https://api.llm-stats.com/leaderboard/benchmarks/apex" + ], + "additional_details": { + "raw_benchmark_id": "apex", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Apex is a challenging frontier reasoning benchmark testing advanced multi-step problem solving across difficult STEM and logical tasks.", + "metric_id": "llm_stats.apex.score", + "metric_name": "Apex score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "apex", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Apex", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.848, + "details": { + "raw_score": "0.848", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "raw_benchmark_id": "apex", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-ultra-550b-a55b\",\"https://llm-stats.com/benchmarks/apex\",\"https://api.llm-stats.com/leaderboard/benchmarks/apex\"]", + "raw_score_id": "apex::nemotron-3-ultra-550b-a55b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "critpt::critpt-nemotron-3-ultra-550b-a55b", + "evaluation_name": "llm_stats.critpt", + "source_data": { + "dataset_name": "CritPT", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-ultra-550b-a55b", + "https://llm-stats.com/benchmarks/critpt", + "https://api.llm-stats.com/leaderboard/benchmarks/critpt" + ], + "additional_details": { + "raw_benchmark_id": "critpt", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CritPT is a challenging reasoning benchmark reported by Qwen for evaluating frontier mathematical and critical problem-solving capability.", + "metric_id": "llm_stats.critpt.score", + "metric_name": "CritPT score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "critpt", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CritPT", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.031, + "details": { + "raw_score": "0.031", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "raw_benchmark_id": "critpt", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-ultra-550b-a55b\",\"https://llm-stats.com/benchmarks/critpt\",\"https://api.llm-stats.com/leaderboard/benchmarks/critpt\"]", + "raw_score_id": "critpt::nemotron-3-ultra-550b-a55b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "finance-agent::finance-agent-nemotron-3-ultra-550b-a55b", + "evaluation_name": "llm_stats.finance-agent", + "source_data": { + "dataset_name": "Finance Agent", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-ultra-550b-a55b", + "https://llm-stats.com/benchmarks/finance-agent", + "https://api.llm-stats.com/leaderboard/benchmarks/finance-agent" + ], + "additional_details": { + "raw_benchmark_id": "finance-agent", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Finance Agent is a benchmark for evaluating AI models on agentic financial analysis tasks, testing their ability to process financial data, perform calculations, and generate accurate analyses across various financial domains.", + "metric_id": "llm_stats.finance-agent.score", + "metric_name": "Finance Agent score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "finance-agent", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Finance Agent", + "raw_categories": "[\"reasoning\",\"finance\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.537, + "details": { + "raw_score": "0.537", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "raw_benchmark_id": "finance-agent", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-ultra-550b-a55b\",\"https://llm-stats.com/benchmarks/finance-agent\",\"https://api.llm-stats.com/leaderboard/benchmarks/finance-agent\"]", + "raw_score_id": "finance-agent::nemotron-3-ultra-550b-a55b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gdpval::gdpval-nemotron-3-ultra-550b-a55b", + "evaluation_name": "llm_stats.gdpval", + "source_data": { + "dataset_name": "GDPval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-ultra-550b-a55b", + "https://llm-stats.com/benchmarks/gdpval", + "https://api.llm-stats.com/leaderboard/benchmarks/gdpval" + ], + "additional_details": { + "raw_benchmark_id": "gdpval", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "GDPval is an OpenAI benchmark evaluating AI models on economically valuable, real-world knowledge-work tasks spanning many professional occupations and industries.", + "metric_id": "llm_stats.gdpval.score", + "metric_name": "GDPval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gdpval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GDPval", + "raw_categories": "[\"legal\",\"reasoning\",\"finance\",\"general\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.467, + "details": { + "raw_score": "0.467", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "raw_benchmark_id": "gdpval", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-ultra-550b-a55b\",\"https://llm-stats.com/benchmarks/gdpval\",\"https://api.llm-stats.com/leaderboard/benchmarks/gdpval\"]", + "raw_score_id": "gdpval::nemotron-3-ultra-550b-a55b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ifbench::ifbench-nemotron-3-ultra-550b-a55b", + "evaluation_name": "llm_stats.ifbench", + "source_data": { + "dataset_name": "IFBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-ultra-550b-a55b", + "https://llm-stats.com/benchmarks/ifbench", + "https://api.llm-stats.com/leaderboard/benchmarks/ifbench" + ], + "additional_details": { + "raw_benchmark_id": "ifbench", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Instruction Following Benchmark evaluating model's ability to follow complex instructions", + "metric_id": "llm_stats.ifbench.score", + "metric_name": "IFBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ifbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IFBench", + "raw_categories": "[\"instruction_following\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.817, + "details": { + "raw_score": "0.817", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "raw_benchmark_id": "ifbench", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-ultra-550b-a55b\",\"https://llm-stats.com/benchmarks/ifbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/ifbench\"]", + "raw_score_id": "ifbench::nemotron-3-ultra-550b-a55b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "imo-answerbench::imo-answerbench-nemotron-3-ultra-550b-a55b", + "evaluation_name": "llm_stats.imo-answerbench", + "source_data": { + "dataset_name": "IMO-AnswerBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-ultra-550b-a55b", + "https://llm-stats.com/benchmarks/imo-answerbench", + "https://api.llm-stats.com/leaderboard/benchmarks/imo-answerbench" + ], + "additional_details": { + "raw_benchmark_id": "imo-answerbench", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "IMO-AnswerBench is a benchmark for evaluating mathematical reasoning capabilities on International Mathematical Olympiad (IMO) problems, focusing on answer generation and verification.", + "metric_id": "llm_stats.imo-answerbench.score", + "metric_name": "IMO-AnswerBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "imo-answerbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IMO-AnswerBench", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.923, + "details": { + "raw_score": "0.923", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "raw_benchmark_id": "imo-answerbench", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-ultra-550b-a55b\",\"https://llm-stats.com/benchmarks/imo-answerbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/imo-answerbench\"]", + "raw_score_id": "imo-answerbench::nemotron-3-ultra-550b-a55b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "livecodebench-v6::livecodebench-v6-nemotron-3-ultra-550b-a55b", + "evaluation_name": "llm_stats.livecodebench-v6", + "source_data": { + "dataset_name": "LiveCodeBench v6", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-ultra-550b-a55b", + "https://llm-stats.com/benchmarks/livecodebench-v6", + "https://api.llm-stats.com/leaderboard/benchmarks/livecodebench-v6" + ], + "additional_details": { + "raw_benchmark_id": "livecodebench-v6", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LiveCodeBench is a holistic and contamination-free evaluation benchmark for large language models for code. It continuously collects new problems from programming contests (LeetCode, AtCoder, CodeForces) and evaluates four different scenarios: code generation, self-repair, code execution, and test output prediction. Problems are annotated with release dates to enable evaluation on unseen problems released after a model's training cutoff.", + "metric_id": "llm_stats.livecodebench-v6.score", + "metric_name": "LiveCodeBench v6 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "livecodebench-v6", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LiveCodeBench v6", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "53" + } + }, + "score_details": { + "score": 0.89, + "details": { + "raw_score": "0.89", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "raw_benchmark_id": "livecodebench-v6", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-ultra-550b-a55b\",\"https://llm-stats.com/benchmarks/livecodebench-v6\",\"https://api.llm-stats.com/leaderboard/benchmarks/livecodebench-v6\"]", + "raw_score_id": "livecodebench-v6::nemotron-3-ultra-550b-a55b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "longbench-v2::longbench-v2-nemotron-3-ultra-550b-a55b", + "evaluation_name": "llm_stats.longbench-v2", + "source_data": { + "dataset_name": "LongBench v2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-ultra-550b-a55b", + "https://llm-stats.com/benchmarks/longbench-v2", + "https://api.llm-stats.com/leaderboard/benchmarks/longbench-v2" + ], + "additional_details": { + "raw_benchmark_id": "longbench-v2", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LongBench v2 is a benchmark designed to assess the ability of LLMs to handle long-context problems requiring deep understanding and reasoning across real-world multitasks. It consists of 503 challenging multiple-choice questions with contexts ranging from 8k to 2M words across six major task categories: single-document QA, multi-document QA, long in-context learning, long-dialogue history understanding, code repository understanding, and long structured data understanding.", + "metric_id": "llm_stats.longbench-v2.score", + "metric_name": "LongBench v2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "longbench-v2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LongBench v2", + "raw_categories": "[\"long_context\",\"reasoning\",\"structured_output\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.619, + "details": { + "raw_score": "0.619", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "raw_benchmark_id": "longbench-v2", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-ultra-550b-a55b\",\"https://llm-stats.com/benchmarks/longbench-v2\",\"https://api.llm-stats.com/leaderboard/benchmarks/longbench-v2\"]", + "raw_score_id": "longbench-v2::nemotron-3-ultra-550b-a55b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu-pro::mmlu-pro-nemotron-3-ultra-550b-a55b", + "evaluation_name": "llm_stats.mmlu-pro", + "source_data": { + "dataset_name": "MMLU-Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-ultra-550b-a55b", + "https://llm-stats.com/benchmarks/mmlu-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu-pro" + ], + "additional_details": { + "raw_benchmark_id": "mmlu-pro", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A more robust and challenging multi-task language understanding benchmark that extends MMLU by expanding multiple-choice options from 4 to 10, eliminating trivial questions, and focusing on reasoning-intensive tasks. Features over 12,000 curated questions across 14 domains and causes a 16-33% accuracy drop compared to original MMLU.", + "metric_id": "llm_stats.mmlu-pro.score", + "metric_name": "MMLU-Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU-Pro", + "raw_categories": "[\"language\",\"legal\",\"math\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "129" + } + }, + "score_details": { + "score": 0.868, + "details": { + "raw_score": "0.868", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "raw_benchmark_id": "mmlu-pro", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-ultra-550b-a55b\",\"https://llm-stats.com/benchmarks/mmlu-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu-pro\"]", + "raw_score_id": "mmlu-pro::nemotron-3-ultra-550b-a55b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu-prox::mmlu-prox-nemotron-3-ultra-550b-a55b", + "evaluation_name": "llm_stats.mmlu-prox", + "source_data": { + "dataset_name": "MMLU-ProX", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-ultra-550b-a55b", + "https://llm-stats.com/benchmarks/mmlu-prox", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu-prox" + ], + "additional_details": { + "raw_benchmark_id": "mmlu-prox", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Extended version of MMLU-Pro providing additional challenging multiple-choice questions for evaluating language models across diverse academic and professional domains. Built on the foundation of the Massive Multitask Language Understanding benchmark framework.", + "metric_id": "llm_stats.mmlu-prox.score", + "metric_name": "MMLU-ProX score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu-prox", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU-ProX", + "raw_categories": "[\"language\",\"legal\",\"math\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "32" + } + }, + "score_details": { + "score": 0.83, + "details": { + "raw_score": "0.83", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "raw_benchmark_id": "mmlu-prox", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-ultra-550b-a55b\",\"https://llm-stats.com/benchmarks/mmlu-prox\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu-prox\"]", + "raw_score_id": "mmlu-prox::nemotron-3-ultra-550b-a55b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "multichallenge::multichallenge-nemotron-3-ultra-550b-a55b", + "evaluation_name": "llm_stats.multichallenge", + "source_data": { + "dataset_name": "Multi-Challenge", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-ultra-550b-a55b", + "https://llm-stats.com/benchmarks/multichallenge", + "https://api.llm-stats.com/leaderboard/benchmarks/multichallenge" + ], + "additional_details": { + "raw_benchmark_id": "multichallenge", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MultiChallenge is a realistic multi-turn conversation evaluation benchmark that challenges frontier LLMs across four key categories: instruction retention (maintaining instructions throughout conversations), inference memory (recalling and connecting details from previous turns), reliable versioned editing (adapting to evolving instructions during collaborative editing), and self-coherence (avoiding contradictions in responses). The benchmark evaluates models on sustained, contextually complex dialogues across diverse topics including travel planning, technical documentation, and professional communication.", + "metric_id": "llm_stats.multichallenge.score", + "metric_name": "Multi-Challenge score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "multichallenge", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Multi-Challenge", + "raw_categories": "[\"reasoning\",\"communication\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "29" + } + }, + "score_details": { + "score": 0.638, + "details": { + "raw_score": "0.638", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "raw_benchmark_id": "multichallenge", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-ultra-550b-a55b\",\"https://llm-stats.com/benchmarks/multichallenge\",\"https://api.llm-stats.com/leaderboard/benchmarks/multichallenge\"]", + "raw_score_id": "multichallenge::nemotron-3-ultra-550b-a55b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "omniscience::omniscience-nemotron-3-ultra-550b-a55b", + "evaluation_name": "llm_stats.omniscience", + "source_data": { + "dataset_name": "OmniScience", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-ultra-550b-a55b", + "https://llm-stats.com/benchmarks/omniscience", + "https://api.llm-stats.com/leaderboard/benchmarks/omniscience" + ], + "additional_details": { + "raw_benchmark_id": "omniscience", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OmniScience is a broad scientific knowledge and reasoning benchmark that measures both answer accuracy and non-hallucination (calibrated abstention) across science domains.", + "metric_id": "llm_stats.omniscience.score", + "metric_name": "OmniScience score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "omniscience", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OmniScience", + "raw_categories": "[\"knowledge\",\"reasoning\",\"science\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.787, + "details": { + "raw_score": "0.787", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "raw_benchmark_id": "omniscience", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-ultra-550b-a55b\",\"https://llm-stats.com/benchmarks/omniscience\",\"https://api.llm-stats.com/leaderboard/benchmarks/omniscience\"]", + "raw_score_id": "omniscience::nemotron-3-ultra-550b-a55b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "pinchbench::pinchbench-nemotron-3-ultra-550b-a55b", + "evaluation_name": "llm_stats.pinchbench", + "source_data": { + "dataset_name": "PinchBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-ultra-550b-a55b", + "https://llm-stats.com/benchmarks/pinchbench", + "https://api.llm-stats.com/leaderboard/benchmarks/pinchbench" + ], + "additional_details": { + "raw_benchmark_id": "pinchbench", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "PinchBench evaluates coding agents on real-world agentic coding tasks, measuring both best-case and average performance across complex software engineering scenarios.", + "metric_id": "llm_stats.pinchbench.score", + "metric_name": "PinchBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "pinchbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PinchBench", + "raw_categories": "[\"agents\",\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.9, + "details": { + "raw_score": "0.9", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "raw_benchmark_id": "pinchbench", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-ultra-550b-a55b\",\"https://llm-stats.com/benchmarks/pinchbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/pinchbench\"]", + "raw_score_id": "pinchbench::nemotron-3-ultra-550b-a55b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "profbench::profbench-nemotron-3-ultra-550b-a55b", + "evaluation_name": "llm_stats.profbench", + "source_data": { + "dataset_name": "ProfBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-ultra-550b-a55b", + "https://llm-stats.com/benchmarks/profbench", + "https://api.llm-stats.com/leaderboard/benchmarks/profbench" + ], + "additional_details": { + "raw_benchmark_id": "profbench", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ProfBench evaluates models on professional-domain reasoning and knowledge-work tasks, including search-augmented question answering across expert fields.", + "metric_id": "llm_stats.profbench.score", + "metric_name": "ProfBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "profbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ProfBench", + "raw_categories": "[\"knowledge\",\"reasoning\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.56, + "details": { + "raw_score": "0.56", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "raw_benchmark_id": "profbench", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-ultra-550b-a55b\",\"https://llm-stats.com/benchmarks/profbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/profbench\"]", + "raw_score_id": "profbench::nemotron-3-ultra-550b-a55b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ruler::ruler-nemotron-3-ultra-550b-a55b", + "evaluation_name": "llm_stats.ruler", + "source_data": { + "dataset_name": "RULER", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-ultra-550b-a55b", + "https://llm-stats.com/benchmarks/ruler", + "https://api.llm-stats.com/leaderboard/benchmarks/ruler" + ], + "additional_details": { + "raw_benchmark_id": "ruler", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "RULER v1 is a synthetic long-context benchmark for measuring how model quality degrades as input length increases. This packaging follows the public standalone NVIDIA RULER implementation with 13 official tasks spanning retrieval, multi-hop tracing, aggregation, and QA.", + "metric_id": "llm_stats.ruler.score", + "metric_name": "RULER score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ruler", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "RULER", + "raw_categories": "[\"long_context\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.947, + "details": { + "raw_score": "0.947", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "raw_benchmark_id": "ruler", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-ultra-550b-a55b\",\"https://llm-stats.com/benchmarks/ruler\",\"https://api.llm-stats.com/leaderboard/benchmarks/ruler\"]", + "raw_score_id": "ruler::nemotron-3-ultra-550b-a55b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "scicode::scicode-nemotron-3-ultra-550b-a55b", + "evaluation_name": "llm_stats.scicode", + "source_data": { + "dataset_name": "SciCode", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-ultra-550b-a55b", + "https://llm-stats.com/benchmarks/scicode", + "https://api.llm-stats.com/leaderboard/benchmarks/scicode" + ], + "additional_details": { + "raw_benchmark_id": "scicode", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SciCode is a research coding benchmark curated by scientists that challenges language models to code solutions for scientific problems. It contains 338 subproblems decomposed from 80 challenging main problems across 16 natural science sub-fields including mathematics, physics, chemistry, biology, and materials science. Problems require knowledge recall, reasoning, and code synthesis skills.", + "metric_id": "llm_stats.scicode.score", + "metric_name": "SciCode score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "scicode", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SciCode", + "raw_categories": "[\"math\",\"physics\",\"reasoning\",\"biology\",\"chemistry\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.446, + "details": { + "raw_score": "0.446", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "raw_benchmark_id": "scicode", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-ultra-550b-a55b\",\"https://llm-stats.com/benchmarks/scicode\",\"https://api.llm-stats.com/leaderboard/benchmarks/scicode\"]", + "raw_score_id": "scicode::nemotron-3-ultra-550b-a55b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-multilingual::swe-bench-multilingual-nemotron-3-ultra-550b-a55b", + "evaluation_name": "llm_stats.swe-bench-multilingual", + "source_data": { + "dataset_name": "SWE-bench Multilingual", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-ultra-550b-a55b", + "https://llm-stats.com/benchmarks/swe-bench-multilingual", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-multilingual" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-multilingual", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multilingual benchmark for issue resolving in software engineering that covers Java, TypeScript, JavaScript, Go, Rust, C, and C++. Contains 1,632 high-quality instances carefully annotated from 2,456 candidates by 68 expert annotators, designed to evaluate Large Language Models across diverse software ecosystems beyond Python.", + "metric_id": "llm_stats.swe-bench-multilingual.score", + "metric_name": "SWE-bench Multilingual score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-multilingual", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-bench Multilingual", + "raw_categories": "[\"reasoning\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "33" + } + }, + "score_details": { + "score": 0.677, + "details": { + "raw_score": "0.677", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "raw_benchmark_id": "swe-bench-multilingual", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-ultra-550b-a55b\",\"https://llm-stats.com/benchmarks/swe-bench-multilingual\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-multilingual\"]", + "raw_score_id": "swe-bench-multilingual::nemotron-3-ultra-550b-a55b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "tau3-bench::tau3-bench-nemotron-3-ultra-550b-a55b", + "evaluation_name": "llm_stats.tau3-bench", + "source_data": { + "dataset_name": "TAU3-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-ultra-550b-a55b", + "https://llm-stats.com/benchmarks/tau3-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/tau3-bench" + ], + "additional_details": { + "raw_benchmark_id": "tau3-bench", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "TAU3-Bench is a benchmark for evaluating general-purpose agent capabilities, testing models on multi-turn interactions with simulated user models, retrieval, and complex decision-making scenarios.", + "metric_id": "llm_stats.tau3-bench.score", + "metric_name": "TAU3-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "tau3-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TAU3-Bench", + "raw_categories": "[\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.226, + "details": { + "raw_score": "0.226", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "raw_benchmark_id": "tau3-bench", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-ultra-550b-a55b\",\"https://llm-stats.com/benchmarks/tau3-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/tau3-bench\"]", + "raw_score_id": "tau3-bench::nemotron-3-ultra-550b-a55b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2.1::terminal-bench-2.1-nemotron-3-ultra-550b-a55b", + "evaluation_name": "llm_stats.terminal-bench-2.1", + "source_data": { + "dataset_name": "Terminal-Bench 2.1", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-ultra-550b-a55b", + "https://llm-stats.com/benchmarks/terminal-bench-2.1", + "https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2.1" + ], + "additional_details": { + "raw_benchmark_id": "terminal-bench-2.1", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Terminal-Bench 2.1 is an updated release of the Terminal-Bench benchmark that tests AI agents' ability to operate a computer via the terminal. It evaluates how well models handle real-world, end-to-end tasks autonomously, including compiling code, training models, setting up servers, system administration, data science workflows, and security tasks.", + "metric_id": "llm_stats.terminal-bench-2.1.score", + "metric_name": "Terminal-Bench 2.1 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "terminal-bench-2.1", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Terminal-Bench 2.1", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.564, + "details": { + "raw_score": "0.564", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "raw_benchmark_id": "terminal-bench-2.1", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-ultra-550b-a55b\",\"https://llm-stats.com/benchmarks/terminal-bench-2.1\",\"https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2.1\"]", + "raw_score_id": "terminal-bench-2.1::nemotron-3-ultra-550b-a55b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "wmt24::wmt24-nemotron-3-ultra-550b-a55b", + "evaluation_name": "llm_stats.wmt24", + "source_data": { + "dataset_name": "WMT24++", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/nemotron-3-ultra-550b-a55b", + "https://llm-stats.com/benchmarks/wmt24", + "https://api.llm-stats.com/leaderboard/benchmarks/wmt24++" + ], + "additional_details": { + "raw_benchmark_id": "wmt24++", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WMT24++ is a comprehensive multilingual machine translation benchmark that expands the WMT24 dataset to cover 55 languages and dialects. It includes human-written references and post-edits across four domains (literary, news, social, and speech) to evaluate machine translation systems and large language models across diverse linguistic contexts.", + "metric_id": "llm_stats.wmt24.score", + "metric_name": "WMT24++ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "wmt24++", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WMT24++", + "raw_categories": "[\"language\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.837, + "details": { + "raw_score": "0.837", + "raw_score_field": "score", + "raw_model_id": "nemotron-3-ultra-550b-a55b", + "raw_benchmark_id": "wmt24++", + "source_urls_json": "[\"https://llm-stats.com/models/nemotron-3-ultra-550b-a55b\",\"https://llm-stats.com/benchmarks/wmt24\",\"https://api.llm-stats.com/leaderboard/benchmarks/wmt24++\"]", + "raw_score_id": "wmt24++::nemotron-3-ultra-550b-a55b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/openai/gpt-5.2-pro-2025-12-11/66cbb424-d136-4bc2-a1bc-d328712b0199.json b/data/llm_stats/openai/gpt-5.2-pro-2025-12-11/66cbb424-d136-4bc2-a1bc-d328712b0199.json new file mode 100644 index 0000000000000000000000000000000000000000..fa559e2fa62415149f4955fc075efc2c90a64283 --- /dev/null +++ b/data/llm_stats/openai/gpt-5.2-pro-2025-12-11/66cbb424-d136-4bc2-a1bc-d328712b0199.json @@ -0,0 +1,358 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/openai_gpt-5.2-pro-2025-12-11/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "GPT-5.2 Pro", + "id": "openai/gpt-5.2-pro-2025-12-11", + "developer": "openai", + "additional_details": { + "raw_id": "gpt-5.2-pro-2025-12-11", + "raw_name": "GPT-5.2 Pro", + "raw_model_id": "gpt-5.2-pro-2025-12-11", + "raw_model_name": "GPT-5.2 Pro", + "raw_organization_id": "openai", + "raw_organization_name": "OpenAI", + "raw_release_date": "2025-12-11", + "raw_announcement_date": "2025-12-11", + "raw_multimodal": "true", + "raw_provider_slug": "openai", + "raw_provider_name": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aime-2025::aime-2025-gpt-5.2-pro-2025-12-11", + "evaluation_name": "llm_stats.aime-2025", + "source_data": { + "dataset_name": "AIME 2025", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.2-pro-2025-12-11", + "https://llm-stats.com/benchmarks/aime-2025", + "https://api.llm-stats.com/leaderboard/benchmarks/aime-2025" + ], + "additional_details": { + "raw_benchmark_id": "aime-2025", + "raw_model_id": "gpt-5.2-pro-2025-12-11", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "All 30 problems from the 2025 American Invitational Mathematics Examination (AIME I and AIME II), testing olympiad-level mathematical reasoning with integer answers from 000-999. Used as an AI benchmark to evaluate large language models' ability to solve complex mathematical problems requiring multi-step logical deductions and structured symbolic reasoning.", + "metric_id": "llm_stats.aime-2025.score", + "metric_name": "AIME 2025 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aime-2025", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AIME 2025", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "114" + } + }, + "score_details": { + "score": 1.0, + "details": { + "raw_score": "1.0", + "raw_score_field": "score", + "raw_model_id": "gpt-5.2-pro-2025-12-11", + "raw_benchmark_id": "aime-2025", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.2-pro-2025-12-11\",\"https://llm-stats.com/benchmarks/aime-2025\",\"https://api.llm-stats.com/leaderboard/benchmarks/aime-2025\"]", + "raw_score_id": "aime-2025::gpt-5.2-pro-2025-12-11", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "arc-agi-v2::arc-agi-v2-gpt-5.2-pro-2025-12-11", + "evaluation_name": "llm_stats.arc-agi-v2", + "source_data": { + "dataset_name": "ARC-AGI v2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.2-pro-2025-12-11", + "https://llm-stats.com/benchmarks/arc-agi-v2", + "https://api.llm-stats.com/leaderboard/benchmarks/arc-agi-v2" + ], + "additional_details": { + "raw_benchmark_id": "arc-agi-v2", + "raw_model_id": "gpt-5.2-pro-2025-12-11", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ARC-AGI-2 is an upgraded benchmark for measuring abstract reasoning and problem-solving abilities in AI systems through visual grid transformation tasks. It evaluates fluid intelligence via input-output grid pairs (1x1 to 30x30) using colored cells (0-9), requiring models to identify underlying transformation rules from demonstration examples and apply them to test cases. Designed to be easy for humans but challenging for AI, focusing on core cognitive abilities like spatial reasoning, pattern recognition, and compositional generalization.", + "metric_id": "llm_stats.arc-agi-v2.score", + "metric_name": "ARC-AGI v2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arc-agi-v2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ARC-AGI v2", + "raw_categories": "[\"reasoning\",\"spatial_reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.542, + "details": { + "raw_score": "0.542", + "raw_score_field": "score", + "raw_model_id": "gpt-5.2-pro-2025-12-11", + "raw_benchmark_id": "arc-agi-v2", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.2-pro-2025-12-11\",\"https://llm-stats.com/benchmarks/arc-agi-v2\",\"https://api.llm-stats.com/leaderboard/benchmarks/arc-agi-v2\"]", + "raw_score_id": "arc-agi-v2::gpt-5.2-pro-2025-12-11", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "arc-agi::arc-agi-gpt-5.2-pro-2025-12-11", + "evaluation_name": "llm_stats.arc-agi", + "source_data": { + "dataset_name": "ARC-AGI", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.2-pro-2025-12-11", + "https://llm-stats.com/benchmarks/arc-agi", + "https://api.llm-stats.com/leaderboard/benchmarks/arc-agi" + ], + "additional_details": { + "raw_benchmark_id": "arc-agi", + "raw_model_id": "gpt-5.2-pro-2025-12-11", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The Abstraction and Reasoning Corpus for Artificial General Intelligence (ARC-AGI) is a benchmark designed to test general intelligence and abstract reasoning capabilities through visual grid-based transformation tasks. Each task consists of 2-5 demonstration pairs showing input grids transformed into output grids according to underlying rules, with test-takers required to infer these rules and apply them to novel test inputs. The benchmark uses colored grids (up to 30x30) with 10 discrete colors/symbols, designed to measure human-like general fluid intelligence and skill-acquisition efficiency with minimal prior knowledge.", + "metric_id": "llm_stats.arc-agi.score", + "metric_name": "ARC-AGI score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arc-agi", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ARC-AGI", + "raw_categories": "[\"reasoning\",\"spatial_reasoning\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.905, + "details": { + "raw_score": "0.905", + "raw_score_field": "score", + "raw_model_id": "gpt-5.2-pro-2025-12-11", + "raw_benchmark_id": "arc-agi", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.2-pro-2025-12-11\",\"https://llm-stats.com/benchmarks/arc-agi\",\"https://api.llm-stats.com/leaderboard/benchmarks/arc-agi\"]", + "raw_score_id": "arc-agi::gpt-5.2-pro-2025-12-11", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "browsecomp::browsecomp-gpt-5.2-pro-2025-12-11", + "evaluation_name": "llm_stats.browsecomp", + "source_data": { + "dataset_name": "BrowseComp", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.2-pro-2025-12-11", + "https://llm-stats.com/benchmarks/browsecomp", + "https://api.llm-stats.com/leaderboard/benchmarks/browsecomp" + ], + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_model_id": "gpt-5.2-pro-2025-12-11", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BrowseComp is a benchmark comprising 1,266 questions that challenge AI agents to persistently navigate the internet in search of hard-to-find, entangled information. The benchmark measures agents' ability to exercise persistence in information gathering, demonstrate creativity in web navigation, and find concise, verifiable answers. Despite the difficulty of the questions, BrowseComp is simple and easy-to-use, as predicted answers are short and easily verifiable against reference answers.", + "metric_id": "llm_stats.browsecomp.score", + "metric_name": "BrowseComp score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BrowseComp", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "52" + } + }, + "score_details": { + "score": 0.779, + "details": { + "raw_score": "0.779", + "raw_score_field": "score", + "raw_model_id": "gpt-5.2-pro-2025-12-11", + "raw_benchmark_id": "browsecomp", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.2-pro-2025-12-11\",\"https://llm-stats.com/benchmarks/browsecomp\",\"https://api.llm-stats.com/leaderboard/benchmarks/browsecomp\"]", + "raw_score_id": "browsecomp::gpt-5.2-pro-2025-12-11", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gpqa::gpqa-gpt-5.2-pro-2025-12-11", + "evaluation_name": "llm_stats.gpqa", + "source_data": { + "dataset_name": "GPQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.2-pro-2025-12-11", + "https://llm-stats.com/benchmarks/gpqa", + "https://api.llm-stats.com/leaderboard/benchmarks/gpqa" + ], + "additional_details": { + "raw_benchmark_id": "gpqa", + "raw_model_id": "gpt-5.2-pro-2025-12-11", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A challenging dataset of 448 multiple-choice questions written by domain experts in biology, physics, and chemistry. Questions are Google-proof and extremely difficult, with PhD experts reaching 65% accuracy.", + "metric_id": "llm_stats.gpqa.score", + "metric_name": "GPQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gpqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GPQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\",\"biology\",\"chemistry\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "226" + } + }, + "score_details": { + "score": 0.932, + "details": { + "raw_score": "0.932", + "raw_score_field": "score", + "raw_model_id": "gpt-5.2-pro-2025-12-11", + "raw_benchmark_id": "gpqa", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.2-pro-2025-12-11\",\"https://llm-stats.com/benchmarks/gpqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/gpqa\"]", + "raw_score_id": "gpqa::gpt-5.2-pro-2025-12-11", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hmmt-2025::hmmt-2025-gpt-5.2-pro-2025-12-11", + "evaluation_name": "llm_stats.hmmt-2025", + "source_data": { + "dataset_name": "HMMT 2025", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.2-pro-2025-12-11", + "https://llm-stats.com/benchmarks/hmmt-2025", + "https://api.llm-stats.com/leaderboard/benchmarks/hmmt-2025" + ], + "additional_details": { + "raw_benchmark_id": "hmmt-2025", + "raw_model_id": "gpt-5.2-pro-2025-12-11", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Harvard-MIT Mathematics Tournament 2025 - A prestigious student-organized mathematics competition for high school students featuring two tournaments (November 2025 at MIT and February 2026 at Harvard) with individual tests, team rounds, and guts rounds", + "metric_id": "llm_stats.hmmt-2025.score", + "metric_name": "HMMT 2025 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hmmt-2025", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HMMT 2025", + "raw_categories": "[\"math\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "33" + } + }, + "score_details": { + "score": 1.0, + "details": { + "raw_score": "1.0", + "raw_score_field": "score", + "raw_model_id": "gpt-5.2-pro-2025-12-11", + "raw_benchmark_id": "hmmt-2025", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.2-pro-2025-12-11\",\"https://llm-stats.com/benchmarks/hmmt-2025\",\"https://api.llm-stats.com/leaderboard/benchmarks/hmmt-2025\"]", + "raw_score_id": "hmmt-2025::gpt-5.2-pro-2025-12-11", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/openai/gpt-5.4/0f126083-7ce8-4b51-a58b-a1569414a46d.json b/data/llm_stats/openai/gpt-5.4/0f126083-7ce8-4b51-a58b-a1569414a46d.json new file mode 100644 index 0000000000000000000000000000000000000000..587552791ec98e92958de1d5252083cd5b90d014 --- /dev/null +++ b/data/llm_stats/openai/gpt-5.4/0f126083-7ce8-4b51-a58b-a1569414a46d.json @@ -0,0 +1,984 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/openai_gpt-5.4/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4", + "id": "openai/gpt-5.4", + "developer": "openai", + "additional_details": { + "raw_id": "gpt-5.4", + "raw_model_name": "gpt-5.4", + "raw_display_name": "GPT-5.4", + "raw_organization_id": "openai", + "raw_organization_name": "OpenAI", + "raw_context_length": "1000000", + "raw_input_modalities": "[\"image\",\"text\"]", + "raw_output_modalities": "[\"text\"]", + "raw_input_price": "2.5", + "raw_output_price": "15.0", + "raw_model_id": "gpt-5.4", + "raw_provider_slug": "openai", + "raw_provider_name": "ZeroEval" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "arc-agi-v2::arc-agi-v2-gpt-5.4", + "evaluation_name": "llm_stats.arc-agi-v2", + "source_data": { + "dataset_name": "ARC-AGI v2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.4", + "https://llm-stats.com/benchmarks/arc-agi-v2", + "https://api.llm-stats.com/leaderboard/benchmarks/arc-agi-v2" + ], + "additional_details": { + "raw_benchmark_id": "arc-agi-v2", + "raw_model_id": "gpt-5.4", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ARC-AGI-2 is an upgraded benchmark for measuring abstract reasoning and problem-solving abilities in AI systems through visual grid transformation tasks. It evaluates fluid intelligence via input-output grid pairs (1x1 to 30x30) using colored cells (0-9), requiring models to identify underlying transformation rules from demonstration examples and apply them to test cases. Designed to be easy for humans but challenging for AI, focusing on core cognitive abilities like spatial reasoning, pattern recognition, and compositional generalization.", + "metric_id": "llm_stats.arc-agi-v2.score", + "metric_name": "ARC-AGI v2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arc-agi-v2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ARC-AGI v2", + "raw_categories": "[\"reasoning\",\"spatial_reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.733, + "details": { + "raw_score": "0.733", + "raw_score_field": "score", + "raw_model_id": "gpt-5.4", + "raw_benchmark_id": "arc-agi-v2", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.4\",\"https://llm-stats.com/benchmarks/arc-agi-v2\",\"https://api.llm-stats.com/leaderboard/benchmarks/arc-agi-v2\"]", + "raw_score_id": "arc-agi-v2::gpt-5.4", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "arc-agi::arc-agi-gpt-5.4", + "evaluation_name": "llm_stats.arc-agi", + "source_data": { + "dataset_name": "ARC-AGI", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.4", + "https://llm-stats.com/benchmarks/arc-agi", + "https://api.llm-stats.com/leaderboard/benchmarks/arc-agi" + ], + "additional_details": { + "raw_benchmark_id": "arc-agi", + "raw_model_id": "gpt-5.4", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The Abstraction and Reasoning Corpus for Artificial General Intelligence (ARC-AGI) is a benchmark designed to test general intelligence and abstract reasoning capabilities through visual grid-based transformation tasks. Each task consists of 2-5 demonstration pairs showing input grids transformed into output grids according to underlying rules, with test-takers required to infer these rules and apply them to novel test inputs. The benchmark uses colored grids (up to 30x30) with 10 discrete colors/symbols, designed to measure human-like general fluid intelligence and skill-acquisition efficiency with minimal prior knowledge.", + "metric_id": "llm_stats.arc-agi.score", + "metric_name": "ARC-AGI score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arc-agi", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ARC-AGI", + "raw_categories": "[\"reasoning\",\"spatial_reasoning\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.937, + "details": { + "raw_score": "0.937", + "raw_score_field": "score", + "raw_model_id": "gpt-5.4", + "raw_benchmark_id": "arc-agi", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.4\",\"https://llm-stats.com/benchmarks/arc-agi\",\"https://api.llm-stats.com/leaderboard/benchmarks/arc-agi\"]", + "raw_score_id": "arc-agi::gpt-5.4", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "browsecomp::browsecomp-gpt-5.4", + "evaluation_name": "llm_stats.browsecomp", + "source_data": { + "dataset_name": "BrowseComp", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.4", + "https://llm-stats.com/benchmarks/browsecomp", + "https://api.llm-stats.com/leaderboard/benchmarks/browsecomp" + ], + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_model_id": "gpt-5.4", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BrowseComp is a benchmark comprising 1,266 questions that challenge AI agents to persistently navigate the internet in search of hard-to-find, entangled information. The benchmark measures agents' ability to exercise persistence in information gathering, demonstrate creativity in web navigation, and find concise, verifiable answers. Despite the difficulty of the questions, BrowseComp is simple and easy-to-use, as predicted answers are short and easily verifiable against reference answers.", + "metric_id": "llm_stats.browsecomp.score", + "metric_name": "BrowseComp score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BrowseComp", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "52" + } + }, + "score_details": { + "score": 0.827, + "details": { + "raw_score": "0.827", + "raw_score_field": "score", + "raw_model_id": "gpt-5.4", + "raw_benchmark_id": "browsecomp", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.4\",\"https://llm-stats.com/benchmarks/browsecomp\",\"https://api.llm-stats.com/leaderboard/benchmarks/browsecomp\"]", + "raw_score_id": "browsecomp::gpt-5.4", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "finance-agent::finance-agent-gpt-5.4", + "evaluation_name": "llm_stats.finance-agent", + "source_data": { + "dataset_name": "Finance Agent", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.4", + "https://llm-stats.com/benchmarks/finance-agent", + "https://api.llm-stats.com/leaderboard/benchmarks/finance-agent" + ], + "additional_details": { + "raw_benchmark_id": "finance-agent", + "raw_model_id": "gpt-5.4", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Finance Agent is a benchmark for evaluating AI models on agentic financial analysis tasks, testing their ability to process financial data, perform calculations, and generate accurate analyses across various financial domains.", + "metric_id": "llm_stats.finance-agent.score", + "metric_name": "Finance Agent score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "finance-agent", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Finance Agent", + "raw_categories": "[\"reasoning\",\"finance\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.56, + "details": { + "raw_score": "0.56", + "raw_score_field": "score", + "raw_model_id": "gpt-5.4", + "raw_benchmark_id": "finance-agent", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.4\",\"https://llm-stats.com/benchmarks/finance-agent\",\"https://api.llm-stats.com/leaderboard/benchmarks/finance-agent\"]", + "raw_score_id": "finance-agent::gpt-5.4", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "frontiermath::frontiermath-gpt-5.4", + "evaluation_name": "llm_stats.frontiermath", + "source_data": { + "dataset_name": "FrontierMath", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.4", + "https://llm-stats.com/benchmarks/frontiermath", + "https://api.llm-stats.com/leaderboard/benchmarks/frontiermath" + ], + "additional_details": { + "raw_benchmark_id": "frontiermath", + "raw_model_id": "gpt-5.4", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A benchmark of hundreds of original, exceptionally challenging mathematics problems crafted and vetted by expert mathematicians, covering most major branches of modern mathematics from number theory and real analysis to algebraic geometry and category theory.", + "metric_id": "llm_stats.frontiermath.score", + "metric_name": "FrontierMath score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "frontiermath", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FrontierMath", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "14" + } + }, + "score_details": { + "score": 0.476, + "details": { + "raw_score": "0.476", + "raw_score_field": "score", + "raw_model_id": "gpt-5.4", + "raw_benchmark_id": "frontiermath", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.4\",\"https://llm-stats.com/benchmarks/frontiermath\",\"https://api.llm-stats.com/leaderboard/benchmarks/frontiermath\"]", + "raw_score_id": "frontiermath::gpt-5.4", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gpqa::gpqa-gpt-5.4", + "evaluation_name": "llm_stats.gpqa", + "source_data": { + "dataset_name": "GPQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.4", + "https://llm-stats.com/benchmarks/gpqa", + "https://api.llm-stats.com/leaderboard/benchmarks/gpqa" + ], + "additional_details": { + "raw_benchmark_id": "gpqa", + "raw_model_id": "gpt-5.4", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A challenging dataset of 448 multiple-choice questions written by domain experts in biology, physics, and chemistry. Questions are Google-proof and extremely difficult, with PhD experts reaching 65% accuracy.", + "metric_id": "llm_stats.gpqa.score", + "metric_name": "GPQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gpqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GPQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\",\"biology\",\"chemistry\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "226" + } + }, + "score_details": { + "score": 0.928, + "details": { + "raw_score": "0.928", + "raw_score_field": "score", + "raw_model_id": "gpt-5.4", + "raw_benchmark_id": "gpqa", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.4\",\"https://llm-stats.com/benchmarks/gpqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/gpqa\"]", + "raw_score_id": "gpqa::gpt-5.4", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "graphwalks-bfs-128k::graphwalks-bfs-128k-gpt-5.4", + "evaluation_name": "llm_stats.graphwalks-bfs-128k", + "source_data": { + "dataset_name": "Graphwalks BFS <128k", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.4", + "https://llm-stats.com/benchmarks/graphwalks-bfs-128k", + "https://api.llm-stats.com/leaderboard/benchmarks/graphwalks-bfs-<128k" + ], + "additional_details": { + "raw_benchmark_id": "graphwalks-bfs-<128k", + "raw_model_id": "gpt-5.4", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A graph reasoning benchmark that evaluates language models' ability to perform breadth-first search (BFS) operations on graphs with context length under 128k tokens, returning nodes reachable at specified depths.", + "metric_id": "llm_stats.graphwalks-bfs-128k.score", + "metric_name": "Graphwalks BFS <128k score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "graphwalks-bfs-<128k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Graphwalks BFS <128k", + "raw_categories": "[\"reasoning\",\"spatial_reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.93, + "details": { + "raw_score": "0.93", + "raw_score_field": "score", + "raw_model_id": "gpt-5.4", + "raw_benchmark_id": "graphwalks-bfs-<128k", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.4\",\"https://llm-stats.com/benchmarks/graphwalks-bfs-128k\",\"https://api.llm-stats.com/leaderboard/benchmarks/graphwalks-bfs-<128k\"]", + "raw_score_id": "graphwalks-bfs-<128k::gpt-5.4", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "graphwalks-bfs-128k::graphwalks-bfs-128k-gpt-5.4", + "evaluation_name": "llm_stats.graphwalks-bfs-128k", + "source_data": { + "dataset_name": "Graphwalks BFS >128k", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.4", + "https://llm-stats.com/benchmarks/graphwalks-bfs-128k", + "https://api.llm-stats.com/leaderboard/benchmarks/graphwalks-bfs->128k" + ], + "additional_details": { + "raw_benchmark_id": "graphwalks-bfs->128k", + "raw_model_id": "gpt-5.4", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A graph reasoning benchmark that evaluates language models' ability to perform breadth-first search (BFS) operations on graphs with context length over 128k tokens, testing long-context reasoning capabilities.", + "metric_id": "llm_stats.graphwalks-bfs-128k.score", + "metric_name": "Graphwalks BFS >128k score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "graphwalks-bfs->128k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Graphwalks BFS >128k", + "raw_categories": "[\"long_context\",\"reasoning\",\"spatial_reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.214, + "details": { + "raw_score": "0.214", + "raw_score_field": "score", + "raw_model_id": "gpt-5.4", + "raw_benchmark_id": "graphwalks-bfs->128k", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.4\",\"https://llm-stats.com/benchmarks/graphwalks-bfs-128k\",\"https://api.llm-stats.com/leaderboard/benchmarks/graphwalks-bfs->128k\"]", + "raw_score_id": "graphwalks-bfs->128k::gpt-5.4", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "graphwalks-parents-128k::graphwalks-parents-128k-gpt-5.4", + "evaluation_name": "llm_stats.graphwalks-parents-128k", + "source_data": { + "dataset_name": "Graphwalks parents <128k", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.4", + "https://llm-stats.com/benchmarks/graphwalks-parents-128k", + "https://api.llm-stats.com/leaderboard/benchmarks/graphwalks-parents-<128k" + ], + "additional_details": { + "raw_benchmark_id": "graphwalks-parents-<128k", + "raw_model_id": "gpt-5.4", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A graph reasoning benchmark that evaluates language models' ability to find parent nodes in graphs with context length under 128k tokens, requiring understanding of graph structure and edge relationships.", + "metric_id": "llm_stats.graphwalks-parents-128k.score", + "metric_name": "Graphwalks parents <128k score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "graphwalks-parents-<128k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Graphwalks parents <128k", + "raw_categories": "[\"reasoning\",\"spatial_reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.898, + "details": { + "raw_score": "0.898", + "raw_score_field": "score", + "raw_model_id": "gpt-5.4", + "raw_benchmark_id": "graphwalks-parents-<128k", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.4\",\"https://llm-stats.com/benchmarks/graphwalks-parents-128k\",\"https://api.llm-stats.com/leaderboard/benchmarks/graphwalks-parents-<128k\"]", + "raw_score_id": "graphwalks-parents-<128k::gpt-5.4", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "graphwalks-parents-128k::graphwalks-parents-128k-gpt-5.4", + "evaluation_name": "llm_stats.graphwalks-parents-128k", + "source_data": { + "dataset_name": "Graphwalks parents >128k", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.4", + "https://llm-stats.com/benchmarks/graphwalks-parents-128k", + "https://api.llm-stats.com/leaderboard/benchmarks/graphwalks-parents->128k" + ], + "additional_details": { + "raw_benchmark_id": "graphwalks-parents->128k", + "raw_model_id": "gpt-5.4", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A graph reasoning benchmark that evaluates language models' ability to find parent nodes in graphs with context length over 128k tokens, testing long-context reasoning and graph structure understanding.", + "metric_id": "llm_stats.graphwalks-parents-128k.score", + "metric_name": "Graphwalks parents >128k score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "graphwalks-parents->128k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Graphwalks parents >128k", + "raw_categories": "[\"long_context\",\"reasoning\",\"spatial_reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.324, + "details": { + "raw_score": "0.324", + "raw_score_field": "score", + "raw_model_id": "gpt-5.4", + "raw_benchmark_id": "graphwalks-parents->128k", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.4\",\"https://llm-stats.com/benchmarks/graphwalks-parents-128k\",\"https://api.llm-stats.com/leaderboard/benchmarks/graphwalks-parents->128k\"]", + "raw_score_id": "graphwalks-parents->128k::gpt-5.4", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mcp-atlas::mcp-atlas-gpt-5.4", + "evaluation_name": "llm_stats.mcp-atlas", + "source_data": { + "dataset_name": "MCP Atlas", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.4", + "https://llm-stats.com/benchmarks/mcp-atlas", + "https://api.llm-stats.com/leaderboard/benchmarks/mcp-atlas" + ], + "additional_details": { + "raw_benchmark_id": "mcp-atlas", + "raw_model_id": "gpt-5.4", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MCP Atlas is a benchmark for evaluating AI models on scaled tool use capabilities, measuring how well models can coordinate and utilize multiple tools across complex multi-step tasks.", + "metric_id": "llm_stats.mcp-atlas.score", + "metric_name": "MCP Atlas score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mcp-atlas", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MCP Atlas", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.672, + "details": { + "raw_score": "0.672", + "raw_score_field": "score", + "raw_model_id": "gpt-5.4", + "raw_benchmark_id": "mcp-atlas", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.4\",\"https://llm-stats.com/benchmarks/mcp-atlas\",\"https://api.llm-stats.com/leaderboard/benchmarks/mcp-atlas\"]", + "raw_score_id": "mcp-atlas::gpt-5.4", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmmu-pro::mmmu-pro-gpt-5.4", + "evaluation_name": "llm_stats.mmmu-pro", + "source_data": { + "dataset_name": "MMMU-Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.4", + "https://llm-stats.com/benchmarks/mmmu-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/mmmu-pro" + ], + "additional_details": { + "raw_benchmark_id": "mmmu-pro", + "raw_model_id": "gpt-5.4", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A more robust multi-discipline multimodal understanding benchmark that enhances MMMU through a three-step process: filtering text-only answerable questions, augmenting candidate options, and introducing vision-only input settings. Achieves significantly lower model performance (16.8-26.9%) compared to original MMMU, providing more rigorous evaluation that closely mimics real-world scenarios.", + "metric_id": "llm_stats.mmmu-pro.score", + "metric_name": "MMMU-Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmmu-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMMU-Pro", + "raw_categories": "[\"multimodal\",\"reasoning\",\"general\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "60" + } + }, + "score_details": { + "score": 0.812, + "details": { + "raw_score": "0.812", + "raw_score_field": "score", + "raw_model_id": "gpt-5.4", + "raw_benchmark_id": "mmmu-pro", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.4\",\"https://llm-stats.com/benchmarks/mmmu-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmmu-pro\"]", + "raw_score_id": "mmmu-pro::gpt-5.4", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "omnidocbench-1.5::omnidocbench-1.5-gpt-5.4", + "evaluation_name": "llm_stats.omnidocbench-1.5", + "source_data": { + "dataset_name": "OmniDocBench 1.5", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.4", + "https://llm-stats.com/benchmarks/omnidocbench-1.5", + "https://api.llm-stats.com/leaderboard/benchmarks/omnidocbench-1.5" + ], + "additional_details": { + "raw_benchmark_id": "omnidocbench-1.5", + "raw_model_id": "gpt-5.4", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OmniDocBench 1.5 is a comprehensive benchmark for evaluating multimodal large language models on document understanding tasks, including OCR, document parsing, information extraction, and visual question answering across diverse document types. Lower Overall Edit Distance scores are better.", + "metric_id": "llm_stats.omnidocbench-1.5.score", + "metric_name": "OmniDocBench 1.5 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "omnidocbench-1.5", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OmniDocBench 1.5", + "raw_categories": "[\"multimodal\",\"reasoning\",\"structured_output\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.891, + "details": { + "raw_score": "0.891", + "raw_score_field": "score", + "raw_model_id": "gpt-5.4", + "raw_benchmark_id": "omnidocbench-1.5", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.4\",\"https://llm-stats.com/benchmarks/omnidocbench-1.5\",\"https://api.llm-stats.com/leaderboard/benchmarks/omnidocbench-1.5\"]", + "raw_score_id": "omnidocbench-1.5::gpt-5.4", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "osworld-verified::osworld-verified-gpt-5.4", + "evaluation_name": "llm_stats.osworld-verified", + "source_data": { + "dataset_name": "OSWorld-Verified", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.4", + "https://llm-stats.com/benchmarks/osworld-verified", + "https://api.llm-stats.com/leaderboard/benchmarks/osworld-verified" + ], + "additional_details": { + "raw_benchmark_id": "osworld-verified", + "raw_model_id": "gpt-5.4", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OSWorld-Verified is a verified subset of OSWorld, a scalable real computer environment for multimodal agents supporting task setup, execution-based evaluation, and interactive learning across Ubuntu, Windows, and macOS.", + "metric_id": "llm_stats.osworld-verified.score", + "metric_name": "OSWorld-Verified score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "osworld-verified", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OSWorld-Verified", + "raw_categories": "[\"multimodal\",\"general\",\"agents\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.75, + "details": { + "raw_score": "0.75", + "raw_score_field": "score", + "raw_model_id": "gpt-5.4", + "raw_benchmark_id": "osworld-verified", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.4\",\"https://llm-stats.com/benchmarks/osworld-verified\",\"https://api.llm-stats.com/leaderboard/benchmarks/osworld-verified\"]", + "raw_score_id": "osworld-verified::gpt-5.4", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-pro::swe-bench-pro-gpt-5.4", + "evaluation_name": "llm_stats.swe-bench-pro", + "source_data": { + "dataset_name": "SWE-Bench Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.4", + "https://llm-stats.com/benchmarks/swe-bench-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-pro" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-pro", + "raw_model_id": "gpt-5.4", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SWE-Bench Pro is an advanced version of SWE-Bench that evaluates language models on complex, real-world software engineering tasks requiring extended reasoning and multi-step problem solving.", + "metric_id": "llm_stats.swe-bench-pro.score", + "metric_name": "SWE-Bench Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Bench Pro", + "raw_categories": "[\"reasoning\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "35" + } + }, + "score_details": { + "score": 0.577, + "details": { + "raw_score": "0.577", + "raw_score_field": "score", + "raw_model_id": "gpt-5.4", + "raw_benchmark_id": "swe-bench-pro", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.4\",\"https://llm-stats.com/benchmarks/swe-bench-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-pro\"]", + "raw_score_id": "swe-bench-pro::gpt-5.4", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "tau2-telecom::tau2-telecom-gpt-5.4", + "evaluation_name": "llm_stats.tau2-telecom", + "source_data": { + "dataset_name": "Tau2 Telecom", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.4", + "https://llm-stats.com/benchmarks/tau2-telecom", + "https://api.llm-stats.com/leaderboard/benchmarks/tau2-telecom" + ], + "additional_details": { + "raw_benchmark_id": "tau2-telecom", + "raw_model_id": "gpt-5.4", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "τ²-Bench telecom domain evaluates conversational agents in a dual-control environment modeled as a Dec-POMDP, where both agent and user use tools in shared telecommunications troubleshooting scenarios that test coordination and communication capabilities.", + "metric_id": "llm_stats.tau2-telecom.score", + "metric_name": "Tau2 Telecom score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "tau2-telecom", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Tau2 Telecom", + "raw_categories": "[\"reasoning\",\"communication\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "35" + } + }, + "score_details": { + "score": 0.989, + "details": { + "raw_score": "0.989", + "raw_score_field": "score", + "raw_model_id": "gpt-5.4", + "raw_benchmark_id": "tau2-telecom", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.4\",\"https://llm-stats.com/benchmarks/tau2-telecom\",\"https://api.llm-stats.com/leaderboard/benchmarks/tau2-telecom\"]", + "raw_score_id": "tau2-telecom::gpt-5.4", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2::terminal-bench-2-gpt-5.4", + "evaluation_name": "llm_stats.terminal-bench-2", + "source_data": { + "dataset_name": "Terminal-Bench 2.0", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.4", + "https://llm-stats.com/benchmarks/terminal-bench-2", + "https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2" + ], + "additional_details": { + "raw_benchmark_id": "terminal-bench-2", + "raw_model_id": "gpt-5.4", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Terminal-Bench 2.0 is an updated benchmark for testing AI agents' tool use ability to operate a computer via terminal. It evaluates how well models can handle real-world, end-to-end tasks autonomously, including compiling code, training models, setting up servers, system administration, security tasks, data science workflows, and cybersecurity vulnerabilities.", + "metric_id": "llm_stats.terminal-bench-2.score", + "metric_name": "Terminal-Bench 2.0 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "terminal-bench-2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Terminal-Bench 2.0", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "49" + } + }, + "score_details": { + "score": 0.751, + "details": { + "raw_score": "0.751", + "raw_score_field": "score", + "raw_model_id": "gpt-5.4", + "raw_benchmark_id": "terminal-bench-2", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.4\",\"https://llm-stats.com/benchmarks/terminal-bench-2\",\"https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2\"]", + "raw_score_id": "terminal-bench-2::gpt-5.4", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "toolathlon::toolathlon-gpt-5.4", + "evaluation_name": "llm_stats.toolathlon", + "source_data": { + "dataset_name": "Toolathlon", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.4", + "https://llm-stats.com/benchmarks/toolathlon", + "https://api.llm-stats.com/leaderboard/benchmarks/toolathlon" + ], + "additional_details": { + "raw_benchmark_id": "toolathlon", + "raw_model_id": "gpt-5.4", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Tool Decathlon is a comprehensive benchmark for evaluating AI agents' ability to use multiple tools across diverse task categories. It measures proficiency in tool selection, sequencing, and execution across ten different tool-use scenarios.", + "metric_id": "llm_stats.toolathlon.score", + "metric_name": "Toolathlon score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "toolathlon", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Toolathlon", + "raw_categories": "[\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "24" + } + }, + "score_details": { + "score": 0.546, + "details": { + "raw_score": "0.546", + "raw_score_field": "score", + "raw_model_id": "gpt-5.4", + "raw_benchmark_id": "toolathlon", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.4\",\"https://llm-stats.com/benchmarks/toolathlon\",\"https://api.llm-stats.com/leaderboard/benchmarks/toolathlon\"]", + "raw_score_id": "toolathlon::gpt-5.4", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/openai/gpt-5.5-pro/584d52ee-ffdc-4b45-b3ea-28a31d136d82.json b/data/llm_stats/openai/gpt-5.5-pro/584d52ee-ffdc-4b45-b3ea-28a31d136d82.json new file mode 100644 index 0000000000000000000000000000000000000000..a5667f00d91eae948347213741465bb9324397db --- /dev/null +++ b/data/llm_stats/openai/gpt-5.5-pro/584d52ee-ffdc-4b45-b3ea-28a31d136d82.json @@ -0,0 +1,306 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/openai_gpt-5.5-pro/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "GPT-5.5 Pro", + "id": "openai/gpt-5.5-pro", + "developer": "openai", + "additional_details": { + "raw_id": "gpt-5.5-pro", + "raw_name": "GPT-5.5 Pro", + "raw_model_id": "gpt-5.5-pro", + "raw_model_name": "GPT-5.5 Pro", + "raw_organization_id": "openai", + "raw_organization_name": "OpenAI", + "raw_release_date": "2026-04-23", + "raw_announcement_date": "2026-04-23", + "raw_multimodal": "true", + "raw_provider_slug": "openai", + "raw_provider_name": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "browsecomp::browsecomp-gpt-5.5-pro", + "evaluation_name": "llm_stats.browsecomp", + "source_data": { + "dataset_name": "BrowseComp", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.5-pro", + "https://llm-stats.com/benchmarks/browsecomp", + "https://api.llm-stats.com/leaderboard/benchmarks/browsecomp" + ], + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_model_id": "gpt-5.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BrowseComp is a benchmark comprising 1,266 questions that challenge AI agents to persistently navigate the internet in search of hard-to-find, entangled information. The benchmark measures agents' ability to exercise persistence in information gathering, demonstrate creativity in web navigation, and find concise, verifiable answers. Despite the difficulty of the questions, BrowseComp is simple and easy-to-use, as predicted answers are short and easily verifiable against reference answers.", + "metric_id": "llm_stats.browsecomp.score", + "metric_name": "BrowseComp score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BrowseComp", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "52" + } + }, + "score_details": { + "score": 0.901, + "details": { + "raw_score": "0.901", + "raw_score_field": "score", + "raw_model_id": "gpt-5.5-pro", + "raw_benchmark_id": "browsecomp", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.5-pro\",\"https://llm-stats.com/benchmarks/browsecomp\",\"https://api.llm-stats.com/leaderboard/benchmarks/browsecomp\"]", + "raw_score_id": "browsecomp::gpt-5.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "frontiermath::frontiermath-gpt-5.5-pro", + "evaluation_name": "llm_stats.frontiermath", + "source_data": { + "dataset_name": "FrontierMath", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.5-pro", + "https://llm-stats.com/benchmarks/frontiermath", + "https://api.llm-stats.com/leaderboard/benchmarks/frontiermath" + ], + "additional_details": { + "raw_benchmark_id": "frontiermath", + "raw_model_id": "gpt-5.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A benchmark of hundreds of original, exceptionally challenging mathematics problems crafted and vetted by expert mathematicians, covering most major branches of modern mathematics from number theory and real analysis to algebraic geometry and category theory.", + "metric_id": "llm_stats.frontiermath.score", + "metric_name": "FrontierMath score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "frontiermath", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FrontierMath", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "14" + } + }, + "score_details": { + "score": 0.396, + "details": { + "raw_score": "0.396", + "raw_score_field": "score", + "raw_model_id": "gpt-5.5-pro", + "raw_benchmark_id": "frontiermath", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.5-pro\",\"https://llm-stats.com/benchmarks/frontiermath\",\"https://api.llm-stats.com/leaderboard/benchmarks/frontiermath\"]", + "raw_score_id": "frontiermath::gpt-5.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gdpval-mm::gdpval-mm-gpt-5.5-pro", + "evaluation_name": "llm_stats.gdpval-mm", + "source_data": { + "dataset_name": "GDPval-MM", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.5-pro", + "https://llm-stats.com/benchmarks/gdpval-mm", + "https://api.llm-stats.com/leaderboard/benchmarks/gdpval-mm" + ], + "additional_details": { + "raw_benchmark_id": "gdpval-mm", + "raw_model_id": "gpt-5.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "GDPval-MM is the multimodal variant of the GDPval benchmark, evaluating AI model performance on real-world economically valuable tasks that require processing and generating multimodal content including documents, slides, diagrams, spreadsheets, images, and other professional deliverables across diverse industries.", + "metric_id": "llm_stats.gdpval-mm.score", + "metric_name": "GDPval-MM score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gdpval-mm", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GDPval-MM", + "raw_categories": "[\"multimodal\",\"reasoning\",\"finance\",\"general\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.823, + "details": { + "raw_score": "0.823", + "raw_score_field": "score", + "raw_model_id": "gpt-5.5-pro", + "raw_benchmark_id": "gdpval-mm", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.5-pro\",\"https://llm-stats.com/benchmarks/gdpval-mm\",\"https://api.llm-stats.com/leaderboard/benchmarks/gdpval-mm\"]", + "raw_score_id": "gdpval-mm::gpt-5.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "genebench::genebench-gpt-5.5-pro", + "evaluation_name": "llm_stats.genebench", + "source_data": { + "dataset_name": "GeneBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.5-pro", + "https://llm-stats.com/benchmarks/genebench", + "https://api.llm-stats.com/leaderboard/benchmarks/genebench" + ], + "additional_details": { + "raw_benchmark_id": "genebench", + "raw_model_id": "gpt-5.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "GeneBench is an evaluation focused on multi-stage scientific data analysis in genetics and quantitative biology. Tasks require reasoning about ambiguous or noisy data with minimal supervisory guidance, addressing realistic obstacles such as hidden confounders or QC failures, and correctly implementing and interpreting modern statistical methods.", + "metric_id": "llm_stats.genebench.score", + "metric_name": "GeneBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "genebench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GeneBench", + "raw_categories": "[\"reasoning\",\"science\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.332, + "details": { + "raw_score": "0.332", + "raw_score_field": "score", + "raw_model_id": "gpt-5.5-pro", + "raw_benchmark_id": "genebench", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.5-pro\",\"https://llm-stats.com/benchmarks/genebench\",\"https://api.llm-stats.com/leaderboard/benchmarks/genebench\"]", + "raw_score_id": "genebench::gpt-5.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "humanity-s-last-exam::humanity-s-last-exam-gpt-5.5-pro", + "evaluation_name": "llm_stats.humanity-s-last-exam", + "source_data": { + "dataset_name": "Humanity's Last Exam", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.5-pro", + "https://llm-stats.com/benchmarks/humanity-s-last-exam", + "https://api.llm-stats.com/leaderboard/benchmarks/humanity's-last-exam" + ], + "additional_details": { + "raw_benchmark_id": "humanity's-last-exam", + "raw_model_id": "gpt-5.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Humanity's Last Exam (HLE) is a multi-modal academic benchmark with 2,500 questions across mathematics, humanities, and natural sciences, designed to test LLM capabilities at the frontier of human knowledge with unambiguous, verifiable solutions", + "metric_id": "llm_stats.humanity-s-last-exam.score", + "metric_name": "Humanity's Last Exam score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "humanity's-last-exam", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Humanity's Last Exam", + "raw_categories": "[\"math\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "88" + } + }, + "score_details": { + "score": 0.572, + "details": { + "raw_score": "0.572", + "raw_score_field": "score", + "raw_model_id": "gpt-5.5-pro", + "raw_benchmark_id": "humanity's-last-exam", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.5-pro\",\"https://llm-stats.com/benchmarks/humanity-s-last-exam\",\"https://api.llm-stats.com/leaderboard/benchmarks/humanity's-last-exam\"]", + "raw_score_id": "humanity's-last-exam::gpt-5.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/openai/gpt-5.5/3d968798-7bfd-4d0f-8dd8-7cddaeb350cd.json b/data/llm_stats/openai/gpt-5.5/3d968798-7bfd-4d0f-8dd8-7cddaeb350cd.json new file mode 100644 index 0000000000000000000000000000000000000000..a6ae3244dc1dd90fc8671a2b417a0191728b2a36 --- /dev/null +++ b/data/llm_stats/openai/gpt-5.5/3d968798-7bfd-4d0f-8dd8-7cddaeb350cd.json @@ -0,0 +1,1192 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/openai_gpt-5.5/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.5", + "id": "openai/gpt-5.5", + "developer": "openai", + "additional_details": { + "raw_id": "gpt-5.5", + "raw_model_name": "gpt-5.5", + "raw_display_name": "GPT-5.5", + "raw_organization_id": "openai", + "raw_organization_name": "OpenAI", + "raw_context_length": "1050000", + "raw_input_modalities": "[\"image\",\"text\"]", + "raw_output_modalities": "[\"text\"]", + "raw_input_price": "5.0", + "raw_output_price": "30.0", + "raw_model_id": "gpt-5.5", + "raw_provider_slug": "openai", + "raw_provider_name": "ZeroEval" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "arc-agi-v2::arc-agi-v2-gpt-5.5", + "evaluation_name": "llm_stats.arc-agi-v2", + "source_data": { + "dataset_name": "ARC-AGI v2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.5", + "https://llm-stats.com/benchmarks/arc-agi-v2", + "https://api.llm-stats.com/leaderboard/benchmarks/arc-agi-v2" + ], + "additional_details": { + "raw_benchmark_id": "arc-agi-v2", + "raw_model_id": "gpt-5.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ARC-AGI-2 is an upgraded benchmark for measuring abstract reasoning and problem-solving abilities in AI systems through visual grid transformation tasks. It evaluates fluid intelligence via input-output grid pairs (1x1 to 30x30) using colored cells (0-9), requiring models to identify underlying transformation rules from demonstration examples and apply them to test cases. Designed to be easy for humans but challenging for AI, focusing on core cognitive abilities like spatial reasoning, pattern recognition, and compositional generalization.", + "metric_id": "llm_stats.arc-agi-v2.score", + "metric_name": "ARC-AGI v2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arc-agi-v2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ARC-AGI v2", + "raw_categories": "[\"reasoning\",\"spatial_reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.85, + "details": { + "raw_score": "0.85", + "raw_score_field": "score", + "raw_model_id": "gpt-5.5", + "raw_benchmark_id": "arc-agi-v2", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.5\",\"https://llm-stats.com/benchmarks/arc-agi-v2\",\"https://api.llm-stats.com/leaderboard/benchmarks/arc-agi-v2\"]", + "raw_score_id": "arc-agi-v2::gpt-5.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "arc-agi::arc-agi-gpt-5.5", + "evaluation_name": "llm_stats.arc-agi", + "source_data": { + "dataset_name": "ARC-AGI", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.5", + "https://llm-stats.com/benchmarks/arc-agi", + "https://api.llm-stats.com/leaderboard/benchmarks/arc-agi" + ], + "additional_details": { + "raw_benchmark_id": "arc-agi", + "raw_model_id": "gpt-5.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The Abstraction and Reasoning Corpus for Artificial General Intelligence (ARC-AGI) is a benchmark designed to test general intelligence and abstract reasoning capabilities through visual grid-based transformation tasks. Each task consists of 2-5 demonstration pairs showing input grids transformed into output grids according to underlying rules, with test-takers required to infer these rules and apply them to novel test inputs. The benchmark uses colored grids (up to 30x30) with 10 discrete colors/symbols, designed to measure human-like general fluid intelligence and skill-acquisition efficiency with minimal prior knowledge.", + "metric_id": "llm_stats.arc-agi.score", + "metric_name": "ARC-AGI score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arc-agi", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ARC-AGI", + "raw_categories": "[\"reasoning\",\"spatial_reasoning\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.95, + "details": { + "raw_score": "0.95", + "raw_score_field": "score", + "raw_model_id": "gpt-5.5", + "raw_benchmark_id": "arc-agi", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.5\",\"https://llm-stats.com/benchmarks/arc-agi\",\"https://api.llm-stats.com/leaderboard/benchmarks/arc-agi\"]", + "raw_score_id": "arc-agi::gpt-5.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bixbench::bixbench-gpt-5.5", + "evaluation_name": "llm_stats.bixbench", + "source_data": { + "dataset_name": "BixBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.5", + "https://llm-stats.com/benchmarks/bixbench", + "https://api.llm-stats.com/leaderboard/benchmarks/bixbench" + ], + "additional_details": { + "raw_benchmark_id": "bixbench", + "raw_model_id": "gpt-5.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BixBench is a benchmark for real-world bioinformatics and computational biology data analysis. It evaluates AI models on multi-step scientific workflows that require code execution, statistical reasoning, and biological domain knowledge to interpret experimental data.", + "metric_id": "llm_stats.bixbench.score", + "metric_name": "BixBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bixbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BixBench", + "raw_categories": "[\"reasoning\",\"science\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.805, + "details": { + "raw_score": "0.805", + "raw_score_field": "score", + "raw_model_id": "gpt-5.5", + "raw_benchmark_id": "bixbench", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.5\",\"https://llm-stats.com/benchmarks/bixbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/bixbench\"]", + "raw_score_id": "bixbench::gpt-5.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "browsecomp::browsecomp-gpt-5.5", + "evaluation_name": "llm_stats.browsecomp", + "source_data": { + "dataset_name": "BrowseComp", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.5", + "https://llm-stats.com/benchmarks/browsecomp", + "https://api.llm-stats.com/leaderboard/benchmarks/browsecomp" + ], + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_model_id": "gpt-5.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BrowseComp is a benchmark comprising 1,266 questions that challenge AI agents to persistently navigate the internet in search of hard-to-find, entangled information. The benchmark measures agents' ability to exercise persistence in information gathering, demonstrate creativity in web navigation, and find concise, verifiable answers. Despite the difficulty of the questions, BrowseComp is simple and easy-to-use, as predicted answers are short and easily verifiable against reference answers.", + "metric_id": "llm_stats.browsecomp.score", + "metric_name": "BrowseComp score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BrowseComp", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "52" + } + }, + "score_details": { + "score": 0.844, + "details": { + "raw_score": "0.844", + "raw_score_field": "score", + "raw_model_id": "gpt-5.5", + "raw_benchmark_id": "browsecomp", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.5\",\"https://llm-stats.com/benchmarks/browsecomp\",\"https://api.llm-stats.com/leaderboard/benchmarks/browsecomp\"]", + "raw_score_id": "browsecomp::gpt-5.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "cybergym::cybergym-gpt-5.5", + "evaluation_name": "llm_stats.cybergym", + "source_data": { + "dataset_name": "CyberGym", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.5", + "https://llm-stats.com/benchmarks/cybergym", + "https://api.llm-stats.com/leaderboard/benchmarks/cybergym" + ], + "additional_details": { + "raw_benchmark_id": "cybergym", + "raw_model_id": "gpt-5.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CyberGym is a benchmark for evaluating AI agents on cybersecurity tasks, testing their ability to identify vulnerabilities, perform security analysis, and complete security-related challenges in a controlled environment.", + "metric_id": "llm_stats.cybergym.score", + "metric_name": "CyberGym score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "cybergym", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CyberGym", + "raw_categories": "[\"safety\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.818, + "details": { + "raw_score": "0.818", + "raw_score_field": "score", + "raw_model_id": "gpt-5.5", + "raw_benchmark_id": "cybergym", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.5\",\"https://llm-stats.com/benchmarks/cybergym\",\"https://api.llm-stats.com/leaderboard/benchmarks/cybergym\"]", + "raw_score_id": "cybergym::gpt-5.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "finance-agent::finance-agent-gpt-5.5", + "evaluation_name": "llm_stats.finance-agent", + "source_data": { + "dataset_name": "Finance Agent", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.5", + "https://llm-stats.com/benchmarks/finance-agent", + "https://api.llm-stats.com/leaderboard/benchmarks/finance-agent" + ], + "additional_details": { + "raw_benchmark_id": "finance-agent", + "raw_model_id": "gpt-5.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Finance Agent is a benchmark for evaluating AI models on agentic financial analysis tasks, testing their ability to process financial data, perform calculations, and generate accurate analyses across various financial domains.", + "metric_id": "llm_stats.finance-agent.score", + "metric_name": "Finance Agent score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "finance-agent", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Finance Agent", + "raw_categories": "[\"reasoning\",\"finance\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.6, + "details": { + "raw_score": "0.6", + "raw_score_field": "score", + "raw_model_id": "gpt-5.5", + "raw_benchmark_id": "finance-agent", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.5\",\"https://llm-stats.com/benchmarks/finance-agent\",\"https://api.llm-stats.com/leaderboard/benchmarks/finance-agent\"]", + "raw_score_id": "finance-agent::gpt-5.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "frontiermath::frontiermath-gpt-5.5", + "evaluation_name": "llm_stats.frontiermath", + "source_data": { + "dataset_name": "FrontierMath", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.5", + "https://llm-stats.com/benchmarks/frontiermath", + "https://api.llm-stats.com/leaderboard/benchmarks/frontiermath" + ], + "additional_details": { + "raw_benchmark_id": "frontiermath", + "raw_model_id": "gpt-5.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A benchmark of hundreds of original, exceptionally challenging mathematics problems crafted and vetted by expert mathematicians, covering most major branches of modern mathematics from number theory and real analysis to algebraic geometry and category theory.", + "metric_id": "llm_stats.frontiermath.score", + "metric_name": "FrontierMath score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "frontiermath", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FrontierMath", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "14" + } + }, + "score_details": { + "score": 0.354, + "details": { + "raw_score": "0.354", + "raw_score_field": "score", + "raw_model_id": "gpt-5.5", + "raw_benchmark_id": "frontiermath", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.5\",\"https://llm-stats.com/benchmarks/frontiermath\",\"https://api.llm-stats.com/leaderboard/benchmarks/frontiermath\"]", + "raw_score_id": "frontiermath::gpt-5.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gdpval-mm::gdpval-mm-gpt-5.5", + "evaluation_name": "llm_stats.gdpval-mm", + "source_data": { + "dataset_name": "GDPval-MM", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.5", + "https://llm-stats.com/benchmarks/gdpval-mm", + "https://api.llm-stats.com/leaderboard/benchmarks/gdpval-mm" + ], + "additional_details": { + "raw_benchmark_id": "gdpval-mm", + "raw_model_id": "gpt-5.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "GDPval-MM is the multimodal variant of the GDPval benchmark, evaluating AI model performance on real-world economically valuable tasks that require processing and generating multimodal content including documents, slides, diagrams, spreadsheets, images, and other professional deliverables across diverse industries.", + "metric_id": "llm_stats.gdpval-mm.score", + "metric_name": "GDPval-MM score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gdpval-mm", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GDPval-MM", + "raw_categories": "[\"multimodal\",\"reasoning\",\"finance\",\"general\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.849, + "details": { + "raw_score": "0.849", + "raw_score_field": "score", + "raw_model_id": "gpt-5.5", + "raw_benchmark_id": "gdpval-mm", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.5\",\"https://llm-stats.com/benchmarks/gdpval-mm\",\"https://api.llm-stats.com/leaderboard/benchmarks/gdpval-mm\"]", + "raw_score_id": "gdpval-mm::gpt-5.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "genebench::genebench-gpt-5.5", + "evaluation_name": "llm_stats.genebench", + "source_data": { + "dataset_name": "GeneBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.5", + "https://llm-stats.com/benchmarks/genebench", + "https://api.llm-stats.com/leaderboard/benchmarks/genebench" + ], + "additional_details": { + "raw_benchmark_id": "genebench", + "raw_model_id": "gpt-5.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "GeneBench is an evaluation focused on multi-stage scientific data analysis in genetics and quantitative biology. Tasks require reasoning about ambiguous or noisy data with minimal supervisory guidance, addressing realistic obstacles such as hidden confounders or QC failures, and correctly implementing and interpreting modern statistical methods.", + "metric_id": "llm_stats.genebench.score", + "metric_name": "GeneBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "genebench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GeneBench", + "raw_categories": "[\"reasoning\",\"science\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.25, + "details": { + "raw_score": "0.25", + "raw_score_field": "score", + "raw_model_id": "gpt-5.5", + "raw_benchmark_id": "genebench", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.5\",\"https://llm-stats.com/benchmarks/genebench\",\"https://api.llm-stats.com/leaderboard/benchmarks/genebench\"]", + "raw_score_id": "genebench::gpt-5.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gpqa::gpqa-gpt-5.5", + "evaluation_name": "llm_stats.gpqa", + "source_data": { + "dataset_name": "GPQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.5", + "https://llm-stats.com/benchmarks/gpqa", + "https://api.llm-stats.com/leaderboard/benchmarks/gpqa" + ], + "additional_details": { + "raw_benchmark_id": "gpqa", + "raw_model_id": "gpt-5.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A challenging dataset of 448 multiple-choice questions written by domain experts in biology, physics, and chemistry. Questions are Google-proof and extremely difficult, with PhD experts reaching 65% accuracy.", + "metric_id": "llm_stats.gpqa.score", + "metric_name": "GPQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gpqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GPQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\",\"biology\",\"chemistry\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "226" + } + }, + "score_details": { + "score": 0.936, + "details": { + "raw_score": "0.936", + "raw_score_field": "score", + "raw_model_id": "gpt-5.5", + "raw_benchmark_id": "gpqa", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.5\",\"https://llm-stats.com/benchmarks/gpqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/gpqa\"]", + "raw_score_id": "gpqa::gpt-5.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "graphwalks-bfs-128k::graphwalks-bfs-128k-gpt-5.5", + "evaluation_name": "llm_stats.graphwalks-bfs-128k", + "source_data": { + "dataset_name": "Graphwalks BFS >128k", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.5", + "https://llm-stats.com/benchmarks/graphwalks-bfs-128k", + "https://api.llm-stats.com/leaderboard/benchmarks/graphwalks-bfs->128k" + ], + "additional_details": { + "raw_benchmark_id": "graphwalks-bfs->128k", + "raw_model_id": "gpt-5.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A graph reasoning benchmark that evaluates language models' ability to perform breadth-first search (BFS) operations on graphs with context length over 128k tokens, testing long-context reasoning capabilities.", + "metric_id": "llm_stats.graphwalks-bfs-128k.score", + "metric_name": "Graphwalks BFS >128k score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "graphwalks-bfs->128k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Graphwalks BFS >128k", + "raw_categories": "[\"long_context\",\"reasoning\",\"spatial_reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.454, + "details": { + "raw_score": "0.454", + "raw_score_field": "score", + "raw_model_id": "gpt-5.5", + "raw_benchmark_id": "graphwalks-bfs->128k", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.5\",\"https://llm-stats.com/benchmarks/graphwalks-bfs-128k\",\"https://api.llm-stats.com/leaderboard/benchmarks/graphwalks-bfs->128k\"]", + "raw_score_id": "graphwalks-bfs->128k::gpt-5.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "graphwalks-parents-128k::graphwalks-parents-128k-gpt-5.5", + "evaluation_name": "llm_stats.graphwalks-parents-128k", + "source_data": { + "dataset_name": "Graphwalks parents >128k", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.5", + "https://llm-stats.com/benchmarks/graphwalks-parents-128k", + "https://api.llm-stats.com/leaderboard/benchmarks/graphwalks-parents->128k" + ], + "additional_details": { + "raw_benchmark_id": "graphwalks-parents->128k", + "raw_model_id": "gpt-5.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A graph reasoning benchmark that evaluates language models' ability to find parent nodes in graphs with context length over 128k tokens, testing long-context reasoning and graph structure understanding.", + "metric_id": "llm_stats.graphwalks-parents-128k.score", + "metric_name": "Graphwalks parents >128k score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "graphwalks-parents->128k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Graphwalks parents >128k", + "raw_categories": "[\"long_context\",\"reasoning\",\"spatial_reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.585, + "details": { + "raw_score": "0.585", + "raw_score_field": "score", + "raw_model_id": "gpt-5.5", + "raw_benchmark_id": "graphwalks-parents->128k", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.5\",\"https://llm-stats.com/benchmarks/graphwalks-parents-128k\",\"https://api.llm-stats.com/leaderboard/benchmarks/graphwalks-parents->128k\"]", + "raw_score_id": "graphwalks-parents->128k::gpt-5.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "humanity-s-last-exam::humanity-s-last-exam-gpt-5.5", + "evaluation_name": "llm_stats.humanity-s-last-exam", + "source_data": { + "dataset_name": "Humanity's Last Exam", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.5", + "https://llm-stats.com/benchmarks/humanity-s-last-exam", + "https://api.llm-stats.com/leaderboard/benchmarks/humanity's-last-exam" + ], + "additional_details": { + "raw_benchmark_id": "humanity's-last-exam", + "raw_model_id": "gpt-5.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Humanity's Last Exam (HLE) is a multi-modal academic benchmark with 2,500 questions across mathematics, humanities, and natural sciences, designed to test LLM capabilities at the frontier of human knowledge with unambiguous, verifiable solutions", + "metric_id": "llm_stats.humanity-s-last-exam.score", + "metric_name": "Humanity's Last Exam score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "humanity's-last-exam", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Humanity's Last Exam", + "raw_categories": "[\"math\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "88" + } + }, + "score_details": { + "score": 0.522, + "details": { + "raw_score": "0.522", + "raw_score_field": "score", + "raw_model_id": "gpt-5.5", + "raw_benchmark_id": "humanity's-last-exam", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.5\",\"https://llm-stats.com/benchmarks/humanity-s-last-exam\",\"https://api.llm-stats.com/leaderboard/benchmarks/humanity's-last-exam\"]", + "raw_score_id": "humanity's-last-exam::gpt-5.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mcp-atlas::mcp-atlas-gpt-5.5", + "evaluation_name": "llm_stats.mcp-atlas", + "source_data": { + "dataset_name": "MCP Atlas", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.5", + "https://llm-stats.com/benchmarks/mcp-atlas", + "https://api.llm-stats.com/leaderboard/benchmarks/mcp-atlas" + ], + "additional_details": { + "raw_benchmark_id": "mcp-atlas", + "raw_model_id": "gpt-5.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MCP Atlas is a benchmark for evaluating AI models on scaled tool use capabilities, measuring how well models can coordinate and utilize multiple tools across complex multi-step tasks.", + "metric_id": "llm_stats.mcp-atlas.score", + "metric_name": "MCP Atlas score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mcp-atlas", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MCP Atlas", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.753, + "details": { + "raw_score": "0.753", + "raw_score_field": "score", + "raw_model_id": "gpt-5.5", + "raw_benchmark_id": "mcp-atlas", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.5\",\"https://llm-stats.com/benchmarks/mcp-atlas\",\"https://api.llm-stats.com/leaderboard/benchmarks/mcp-atlas\"]", + "raw_score_id": "mcp-atlas::gpt-5.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmmu-pro::mmmu-pro-gpt-5.5", + "evaluation_name": "llm_stats.mmmu-pro", + "source_data": { + "dataset_name": "MMMU-Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.5", + "https://llm-stats.com/benchmarks/mmmu-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/mmmu-pro" + ], + "additional_details": { + "raw_benchmark_id": "mmmu-pro", + "raw_model_id": "gpt-5.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A more robust multi-discipline multimodal understanding benchmark that enhances MMMU through a three-step process: filtering text-only answerable questions, augmenting candidate options, and introducing vision-only input settings. Achieves significantly lower model performance (16.8-26.9%) compared to original MMMU, providing more rigorous evaluation that closely mimics real-world scenarios.", + "metric_id": "llm_stats.mmmu-pro.score", + "metric_name": "MMMU-Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmmu-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMMU-Pro", + "raw_categories": "[\"multimodal\",\"reasoning\",\"general\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "60" + } + }, + "score_details": { + "score": 0.832, + "details": { + "raw_score": "0.832", + "raw_score_field": "score", + "raw_model_id": "gpt-5.5", + "raw_benchmark_id": "mmmu-pro", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.5\",\"https://llm-stats.com/benchmarks/mmmu-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmmu-pro\"]", + "raw_score_id": "mmmu-pro::gpt-5.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mrcr-v2-8-needle::mrcr-v2-8-needle-gpt-5.5", + "evaluation_name": "llm_stats.mrcr-v2-8-needle", + "source_data": { + "dataset_name": "MRCR v2 (8-needle)", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.5", + "https://llm-stats.com/benchmarks/mrcr-v2-8-needle", + "https://api.llm-stats.com/leaderboard/benchmarks/mrcr-v2-(8-needle)" + ], + "additional_details": { + "raw_benchmark_id": "mrcr-v2-(8-needle)", + "raw_model_id": "gpt-5.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MRCR v2 (8-needle) is a variant of the Multi-Round Coreference Resolution benchmark that includes 8 needle items to retrieve from long contexts. This tests models' ability to simultaneously track and reason about multiple pieces of information across extended conversations.", + "metric_id": "llm_stats.mrcr-v2-8-needle.score", + "metric_name": "MRCR v2 (8-needle) score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mrcr-v2-(8-needle)", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MRCR v2 (8-needle)", + "raw_categories": "[\"long_context\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.74, + "details": { + "raw_score": "0.74", + "raw_score_field": "score", + "raw_model_id": "gpt-5.5", + "raw_benchmark_id": "mrcr-v2-(8-needle)", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.5\",\"https://llm-stats.com/benchmarks/mrcr-v2-8-needle\",\"https://api.llm-stats.com/leaderboard/benchmarks/mrcr-v2-(8-needle)\"]", + "raw_score_id": "mrcr-v2-(8-needle)::gpt-5.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "officeqa-pro::officeqa-pro-gpt-5.5", + "evaluation_name": "llm_stats.officeqa-pro", + "source_data": { + "dataset_name": "OfficeQA Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.5", + "https://llm-stats.com/benchmarks/officeqa-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/officeqa-pro" + ], + "additional_details": { + "raw_benchmark_id": "officeqa-pro", + "raw_model_id": "gpt-5.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OfficeQA Pro evaluates AI models on professional knowledge-work questions and tasks drawn from real office workflows, including document analysis, spreadsheet reasoning, and information synthesis across business domains.", + "metric_id": "llm_stats.officeqa-pro.score", + "metric_name": "OfficeQA Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "officeqa-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OfficeQA Pro", + "raw_categories": "[\"reasoning\",\"general\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.541, + "details": { + "raw_score": "0.541", + "raw_score_field": "score", + "raw_model_id": "gpt-5.5", + "raw_benchmark_id": "officeqa-pro", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.5\",\"https://llm-stats.com/benchmarks/officeqa-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/officeqa-pro\"]", + "raw_score_id": "officeqa-pro::gpt-5.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "osworld-verified::osworld-verified-gpt-5.5", + "evaluation_name": "llm_stats.osworld-verified", + "source_data": { + "dataset_name": "OSWorld-Verified", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.5", + "https://llm-stats.com/benchmarks/osworld-verified", + "https://api.llm-stats.com/leaderboard/benchmarks/osworld-verified" + ], + "additional_details": { + "raw_benchmark_id": "osworld-verified", + "raw_model_id": "gpt-5.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OSWorld-Verified is a verified subset of OSWorld, a scalable real computer environment for multimodal agents supporting task setup, execution-based evaluation, and interactive learning across Ubuntu, Windows, and macOS.", + "metric_id": "llm_stats.osworld-verified.score", + "metric_name": "OSWorld-Verified score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "osworld-verified", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OSWorld-Verified", + "raw_categories": "[\"multimodal\",\"general\",\"agents\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.787, + "details": { + "raw_score": "0.787", + "raw_score_field": "score", + "raw_model_id": "gpt-5.5", + "raw_benchmark_id": "osworld-verified", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.5\",\"https://llm-stats.com/benchmarks/osworld-verified\",\"https://api.llm-stats.com/leaderboard/benchmarks/osworld-verified\"]", + "raw_score_id": "osworld-verified::gpt-5.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-pro::swe-bench-pro-gpt-5.5", + "evaluation_name": "llm_stats.swe-bench-pro", + "source_data": { + "dataset_name": "SWE-Bench Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.5", + "https://llm-stats.com/benchmarks/swe-bench-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-pro" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-pro", + "raw_model_id": "gpt-5.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SWE-Bench Pro is an advanced version of SWE-Bench that evaluates language models on complex, real-world software engineering tasks requiring extended reasoning and multi-step problem solving.", + "metric_id": "llm_stats.swe-bench-pro.score", + "metric_name": "SWE-Bench Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Bench Pro", + "raw_categories": "[\"reasoning\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "35" + } + }, + "score_details": { + "score": 0.586, + "details": { + "raw_score": "0.586", + "raw_score_field": "score", + "raw_model_id": "gpt-5.5", + "raw_benchmark_id": "swe-bench-pro", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.5\",\"https://llm-stats.com/benchmarks/swe-bench-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-pro\"]", + "raw_score_id": "swe-bench-pro::gpt-5.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "tau2-telecom::tau2-telecom-gpt-5.5", + "evaluation_name": "llm_stats.tau2-telecom", + "source_data": { + "dataset_name": "Tau2 Telecom", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.5", + "https://llm-stats.com/benchmarks/tau2-telecom", + "https://api.llm-stats.com/leaderboard/benchmarks/tau2-telecom" + ], + "additional_details": { + "raw_benchmark_id": "tau2-telecom", + "raw_model_id": "gpt-5.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "τ²-Bench telecom domain evaluates conversational agents in a dual-control environment modeled as a Dec-POMDP, where both agent and user use tools in shared telecommunications troubleshooting scenarios that test coordination and communication capabilities.", + "metric_id": "llm_stats.tau2-telecom.score", + "metric_name": "Tau2 Telecom score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "tau2-telecom", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Tau2 Telecom", + "raw_categories": "[\"reasoning\",\"communication\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "35" + } + }, + "score_details": { + "score": 0.98, + "details": { + "raw_score": "0.98", + "raw_score_field": "score", + "raw_model_id": "gpt-5.5", + "raw_benchmark_id": "tau2-telecom", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.5\",\"https://llm-stats.com/benchmarks/tau2-telecom\",\"https://api.llm-stats.com/leaderboard/benchmarks/tau2-telecom\"]", + "raw_score_id": "tau2-telecom::gpt-5.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2::terminal-bench-2-gpt-5.5", + "evaluation_name": "llm_stats.terminal-bench-2", + "source_data": { + "dataset_name": "Terminal-Bench 2.0", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.5", + "https://llm-stats.com/benchmarks/terminal-bench-2", + "https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2" + ], + "additional_details": { + "raw_benchmark_id": "terminal-bench-2", + "raw_model_id": "gpt-5.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Terminal-Bench 2.0 is an updated benchmark for testing AI agents' tool use ability to operate a computer via terminal. It evaluates how well models can handle real-world, end-to-end tasks autonomously, including compiling code, training models, setting up servers, system administration, security tasks, data science workflows, and cybersecurity vulnerabilities.", + "metric_id": "llm_stats.terminal-bench-2.score", + "metric_name": "Terminal-Bench 2.0 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "terminal-bench-2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Terminal-Bench 2.0", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "49" + } + }, + "score_details": { + "score": 0.827, + "details": { + "raw_score": "0.827", + "raw_score_field": "score", + "raw_model_id": "gpt-5.5", + "raw_benchmark_id": "terminal-bench-2", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.5\",\"https://llm-stats.com/benchmarks/terminal-bench-2\",\"https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2\"]", + "raw_score_id": "terminal-bench-2::gpt-5.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "toolathlon::toolathlon-gpt-5.5", + "evaluation_name": "llm_stats.toolathlon", + "source_data": { + "dataset_name": "Toolathlon", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/gpt-5.5", + "https://llm-stats.com/benchmarks/toolathlon", + "https://api.llm-stats.com/leaderboard/benchmarks/toolathlon" + ], + "additional_details": { + "raw_benchmark_id": "toolathlon", + "raw_model_id": "gpt-5.5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Tool Decathlon is a comprehensive benchmark for evaluating AI agents' ability to use multiple tools across diverse task categories. It measures proficiency in tool selection, sequencing, and execution across ten different tool-use scenarios.", + "metric_id": "llm_stats.toolathlon.score", + "metric_name": "Toolathlon score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "toolathlon", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Toolathlon", + "raw_categories": "[\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "24" + } + }, + "score_details": { + "score": 0.556, + "details": { + "raw_score": "0.556", + "raw_score_field": "score", + "raw_model_id": "gpt-5.5", + "raw_benchmark_id": "toolathlon", + "source_urls_json": "[\"https://llm-stats.com/models/gpt-5.5\",\"https://llm-stats.com/benchmarks/toolathlon\",\"https://api.llm-stats.com/leaderboard/benchmarks/toolathlon\"]", + "raw_score_id": "toolathlon::gpt-5.5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/openbmb/minicpm-sala/81e3313e-3b9d-44d4-b7ba-5614ac4f99d9.json b/data/llm_stats/openbmb/minicpm-sala/81e3313e-3b9d-44d4-b7ba-5614ac4f99d9.json new file mode 100644 index 0000000000000000000000000000000000000000..d110a0df6e0557861086de37ebab333ed99296c2 --- /dev/null +++ b/data/llm_stats/openbmb/minicpm-sala/81e3313e-3b9d-44d4-b7ba-5614ac4f99d9.json @@ -0,0 +1,1087 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/openbmb_minicpm-sala/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "MiniCPM-SALA", + "id": "openbmb/minicpm-sala", + "developer": "openbmb", + "additional_details": { + "raw_id": "minicpm-sala", + "raw_name": "MiniCPM-SALA", + "raw_model_id": "minicpm-sala", + "raw_model_name": "MiniCPM-SALA", + "raw_organization_id": "openbmb", + "raw_organization_name": "OpenBMB", + "raw_release_date": "2026-02-11", + "raw_announcement_date": "2026-02-11", + "raw_multimodal": "false", + "raw_param_count": "9477203968", + "raw_provider_slug": "openbmb", + "raw_provider_name": "OpenBMB" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aime-2024::aime-2024-minicpm-sala", + "evaluation_name": "llm_stats.aime-2024", + "source_data": { + "dataset_name": "AIME 2024", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minicpm-sala", + "https://llm-stats.com/benchmarks/aime-2024", + "https://api.llm-stats.com/leaderboard/benchmarks/aime-2024" + ], + "additional_details": { + "raw_benchmark_id": "aime-2024", + "raw_model_id": "minicpm-sala", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "American Invitational Mathematics Examination 2024, consisting of 30 challenging mathematical reasoning problems from AIME I and AIME II competitions. Each problem requires an integer answer between 0-999 and tests advanced mathematical reasoning across algebra, geometry, combinatorics, and number theory. Used as a benchmark for evaluating mathematical reasoning capabilities in large language models at Olympiad-level difficulty.", + "metric_id": "llm_stats.aime-2024.score", + "metric_name": "AIME 2024 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aime-2024", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AIME 2024", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "53" + } + }, + "score_details": { + "score": 0.8375, + "details": { + "raw_score": "0.8375", + "raw_score_field": "score", + "raw_model_id": "minicpm-sala", + "raw_benchmark_id": "aime-2024", + "source_urls_json": "[\"https://llm-stats.com/models/minicpm-sala\",\"https://llm-stats.com/benchmarks/aime-2024\",\"https://api.llm-stats.com/leaderboard/benchmarks/aime-2024\"]", + "raw_score_id": "aime-2024::minicpm-sala", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bbh::bbh-minicpm-sala", + "evaluation_name": "llm_stats.bbh", + "source_data": { + "dataset_name": "BBH", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minicpm-sala", + "https://llm-stats.com/benchmarks/bbh", + "https://api.llm-stats.com/leaderboard/benchmarks/bbh" + ], + "additional_details": { + "raw_benchmark_id": "bbh", + "raw_model_id": "minicpm-sala", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Big-Bench Hard (BBH) is a suite of 23 challenging tasks selected from BIG-Bench for which prior language model evaluations did not outperform the average human-rater. These tasks require multi-step reasoning across diverse domains including arithmetic, logical reasoning, reading comprehension, and commonsense reasoning. The benchmark was designed to test capabilities believed to be beyond current language models and focuses on evaluating complex reasoning skills including temporal understanding, spatial reasoning, causal understanding, and deductive logical reasoning.", + "metric_id": "llm_stats.bbh.score", + "metric_name": "BBH score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bbh", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BBH", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.8155, + "details": { + "raw_score": "0.8155", + "raw_score_field": "score", + "raw_model_id": "minicpm-sala", + "raw_benchmark_id": "bbh", + "source_urls_json": "[\"https://llm-stats.com/models/minicpm-sala\",\"https://llm-stats.com/benchmarks/bbh\",\"https://api.llm-stats.com/leaderboard/benchmarks/bbh\"]", + "raw_score_id": "bbh::minicpm-sala", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "cmmlu::cmmlu-minicpm-sala", + "evaluation_name": "llm_stats.cmmlu", + "source_data": { + "dataset_name": "CMMLU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minicpm-sala", + "https://llm-stats.com/benchmarks/cmmlu", + "https://api.llm-stats.com/leaderboard/benchmarks/cmmlu" + ], + "additional_details": { + "raw_benchmark_id": "cmmlu", + "raw_model_id": "minicpm-sala", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CMMLU (Chinese Massive Multitask Language Understanding) is a comprehensive Chinese benchmark that evaluates the knowledge and reasoning capabilities of large language models across 67 different subject topics. The benchmark covers natural sciences, social sciences, engineering, and humanities with multiple-choice questions ranging from basic to advanced professional levels.", + "metric_id": "llm_stats.cmmlu.score", + "metric_name": "CMMLU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "cmmlu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CMMLU", + "raw_categories": "[\"language\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.8155, + "details": { + "raw_score": "0.8155", + "raw_score_field": "score", + "raw_model_id": "minicpm-sala", + "raw_benchmark_id": "cmmlu", + "source_urls_json": "[\"https://llm-stats.com/models/minicpm-sala\",\"https://llm-stats.com/benchmarks/cmmlu\",\"https://api.llm-stats.com/leaderboard/benchmarks/cmmlu\"]", + "raw_score_id": "cmmlu::minicpm-sala", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "humaneval::humaneval-minicpm-sala", + "evaluation_name": "llm_stats.humaneval", + "source_data": { + "dataset_name": "HumanEval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minicpm-sala", + "https://llm-stats.com/benchmarks/humaneval", + "https://api.llm-stats.com/leaderboard/benchmarks/humaneval" + ], + "additional_details": { + "raw_benchmark_id": "humaneval", + "raw_model_id": "minicpm-sala", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A benchmark that measures functional correctness for synthesizing programs from docstrings, consisting of 164 original programming problems assessing language comprehension, algorithms, and simple mathematics", + "metric_id": "llm_stats.humaneval.score", + "metric_name": "HumanEval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "humaneval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HumanEval", + "raw_categories": "[\"reasoning\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "66" + } + }, + "score_details": { + "score": 0.9512, + "details": { + "raw_score": "0.9512", + "raw_score_field": "score", + "raw_model_id": "minicpm-sala", + "raw_benchmark_id": "humaneval", + "source_urls_json": "[\"https://llm-stats.com/models/minicpm-sala\",\"https://llm-stats.com/benchmarks/humaneval\",\"https://api.llm-stats.com/leaderboard/benchmarks/humaneval\"]", + "raw_score_id": "humaneval::minicpm-sala", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "livecodebench-v5::livecodebench-v5-minicpm-sala", + "evaluation_name": "llm_stats.livecodebench-v5", + "source_data": { + "dataset_name": "LiveCodeBench v5", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minicpm-sala", + "https://llm-stats.com/benchmarks/livecodebench-v5", + "https://api.llm-stats.com/leaderboard/benchmarks/livecodebench-v5" + ], + "additional_details": { + "raw_benchmark_id": "livecodebench-v5", + "raw_model_id": "minicpm-sala", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LiveCodeBench is a holistic and contamination-free evaluation benchmark for large language models for code. It continuously collects new problems from programming contests (LeetCode, AtCoder, CodeForces) and evaluates four different scenarios: code generation, self-repair, code execution, and test output prediction. Problems are annotated with release dates to enable evaluation on unseen problems released after a model's training cutoff.", + "metric_id": "llm_stats.livecodebench-v5.score", + "metric_name": "LiveCodeBench v5 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "livecodebench-v5", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LiveCodeBench v5", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.6048, + "details": { + "raw_score": "0.6048", + "raw_score_field": "score", + "raw_model_id": "minicpm-sala", + "raw_benchmark_id": "livecodebench-v5", + "source_urls_json": "[\"https://llm-stats.com/models/minicpm-sala\",\"https://llm-stats.com/benchmarks/livecodebench-v5\",\"https://api.llm-stats.com/leaderboard/benchmarks/livecodebench-v5\"]", + "raw_score_id": "livecodebench-v5::minicpm-sala", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mbpp::mbpp-minicpm-sala", + "evaluation_name": "llm_stats.mbpp", + "source_data": { + "dataset_name": "MBPP", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minicpm-sala", + "https://llm-stats.com/benchmarks/mbpp", + "https://api.llm-stats.com/leaderboard/benchmarks/mbpp" + ], + "additional_details": { + "raw_benchmark_id": "mbpp", + "raw_model_id": "minicpm-sala", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MBPP (Mostly Basic Python Problems) is a benchmark of 974 crowd-sourced Python programming problems designed to be solvable by entry-level programmers. Each problem consists of a task description, code solution, and 3 automated test cases covering programming fundamentals and standard library functionality.", + "metric_id": "llm_stats.mbpp.score", + "metric_name": "MBPP score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mbpp", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MBPP", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "33" + } + }, + "score_details": { + "score": 0.8911, + "details": { + "raw_score": "0.8911", + "raw_score_field": "score", + "raw_model_id": "minicpm-sala", + "raw_benchmark_id": "mbpp", + "source_urls_json": "[\"https://llm-stats.com/models/minicpm-sala\",\"https://llm-stats.com/benchmarks/mbpp\",\"https://api.llm-stats.com/leaderboard/benchmarks/mbpp\"]", + "raw_score_id": "mbpp::minicpm-sala", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mrcr-128k-2-needle::mrcr-128k-2-needle-minicpm-sala", + "evaluation_name": "llm_stats.mrcr-128k-2-needle", + "source_data": { + "dataset_name": "MRCR 128K (2-needle)", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minicpm-sala", + "https://llm-stats.com/benchmarks/mrcr-128k-2-needle", + "https://api.llm-stats.com/leaderboard/benchmarks/mrcr-128k-(2-needle)" + ], + "additional_details": { + "raw_benchmark_id": "mrcr-128k-(2-needle)", + "raw_model_id": "minicpm-sala", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MRCR (Multi-Round Coreference Resolution) at 128K context length with 2 needles. Models must navigate long conversations to reproduce specific model outputs, testing attention and reasoning across 128K-token contexts with 2 items to retrieve.", + "metric_id": "llm_stats.mrcr-128k-2-needle.score", + "metric_name": "MRCR 128K (2-needle) score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mrcr-128k-(2-needle)", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MRCR 128K (2-needle)", + "raw_categories": "[\"long_context\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.2862, + "details": { + "raw_score": "0.2862", + "raw_score_field": "score", + "raw_model_id": "minicpm-sala", + "raw_benchmark_id": "mrcr-128k-(2-needle)", + "source_urls_json": "[\"https://llm-stats.com/models/minicpm-sala\",\"https://llm-stats.com/benchmarks/mrcr-128k-2-needle\",\"https://api.llm-stats.com/leaderboard/benchmarks/mrcr-128k-(2-needle)\"]", + "raw_score_id": "mrcr-128k-(2-needle)::minicpm-sala", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mrcr-128k-4-needle::mrcr-128k-4-needle-minicpm-sala", + "evaluation_name": "llm_stats.mrcr-128k-4-needle", + "source_data": { + "dataset_name": "MRCR 128K (4-needle)", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minicpm-sala", + "https://llm-stats.com/benchmarks/mrcr-128k-4-needle", + "https://api.llm-stats.com/leaderboard/benchmarks/mrcr-128k-(4-needle)" + ], + "additional_details": { + "raw_benchmark_id": "mrcr-128k-(4-needle)", + "raw_model_id": "minicpm-sala", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MRCR (Multi-Round Coreference Resolution) at 128K context length with 4 needles. Models must navigate long conversations to reproduce specific model outputs, testing attention and reasoning across 128K-token contexts with 4 items to retrieve.", + "metric_id": "llm_stats.mrcr-128k-4-needle.score", + "metric_name": "MRCR 128K (4-needle) score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mrcr-128k-(4-needle)", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MRCR 128K (4-needle)", + "raw_categories": "[\"long_context\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.1962, + "details": { + "raw_score": "0.1962", + "raw_score_field": "score", + "raw_model_id": "minicpm-sala", + "raw_benchmark_id": "mrcr-128k-(4-needle)", + "source_urls_json": "[\"https://llm-stats.com/models/minicpm-sala\",\"https://llm-stats.com/benchmarks/mrcr-128k-4-needle\",\"https://api.llm-stats.com/leaderboard/benchmarks/mrcr-128k-(4-needle)\"]", + "raw_score_id": "mrcr-128k-(4-needle)::minicpm-sala", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mrcr-128k-8-needle::mrcr-128k-8-needle-minicpm-sala", + "evaluation_name": "llm_stats.mrcr-128k-8-needle", + "source_data": { + "dataset_name": "MRCR 128K (8-needle)", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minicpm-sala", + "https://llm-stats.com/benchmarks/mrcr-128k-8-needle", + "https://api.llm-stats.com/leaderboard/benchmarks/mrcr-128k-(8-needle)" + ], + "additional_details": { + "raw_benchmark_id": "mrcr-128k-(8-needle)", + "raw_model_id": "minicpm-sala", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MRCR (Multi-Round Coreference Resolution) at 128K context length with 8 needles. Models must navigate long conversations to reproduce specific model outputs, testing attention and reasoning across 128K-token contexts with 8 items to retrieve.", + "metric_id": "llm_stats.mrcr-128k-8-needle.score", + "metric_name": "MRCR 128K (8-needle) score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mrcr-128k-(8-needle)", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MRCR 128K (8-needle)", + "raw_categories": "[\"long_context\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.1012, + "details": { + "raw_score": "0.1012", + "raw_score_field": "score", + "raw_model_id": "minicpm-sala", + "raw_benchmark_id": "mrcr-128k-(8-needle)", + "source_urls_json": "[\"https://llm-stats.com/models/minicpm-sala\",\"https://llm-stats.com/benchmarks/mrcr-128k-8-needle\",\"https://api.llm-stats.com/leaderboard/benchmarks/mrcr-128k-(8-needle)\"]", + "raw_score_id": "mrcr-128k-(8-needle)::minicpm-sala", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mrcr-64k-2-needle::mrcr-64k-2-needle-minicpm-sala", + "evaluation_name": "llm_stats.mrcr-64k-2-needle", + "source_data": { + "dataset_name": "MRCR 64K (2-needle)", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minicpm-sala", + "https://llm-stats.com/benchmarks/mrcr-64k-2-needle", + "https://api.llm-stats.com/leaderboard/benchmarks/mrcr-64k-(2-needle)" + ], + "additional_details": { + "raw_benchmark_id": "mrcr-64k-(2-needle)", + "raw_model_id": "minicpm-sala", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MRCR (Multi-Round Coreference Resolution) at 64K context length with 2 needles. Models must navigate long conversations to reproduce specific model outputs, testing attention and reasoning across 64K-token contexts with 2 items to retrieve.", + "metric_id": "llm_stats.mrcr-64k-2-needle.score", + "metric_name": "MRCR 64K (2-needle) score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mrcr-64k-(2-needle)", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MRCR 64K (2-needle)", + "raw_categories": "[\"long_context\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.2977, + "details": { + "raw_score": "0.2977", + "raw_score_field": "score", + "raw_model_id": "minicpm-sala", + "raw_benchmark_id": "mrcr-64k-(2-needle)", + "source_urls_json": "[\"https://llm-stats.com/models/minicpm-sala\",\"https://llm-stats.com/benchmarks/mrcr-64k-2-needle\",\"https://api.llm-stats.com/leaderboard/benchmarks/mrcr-64k-(2-needle)\"]", + "raw_score_id": "mrcr-64k-(2-needle)::minicpm-sala", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mrcr-64k-4-needle::mrcr-64k-4-needle-minicpm-sala", + "evaluation_name": "llm_stats.mrcr-64k-4-needle", + "source_data": { + "dataset_name": "MRCR 64K (4-needle)", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minicpm-sala", + "https://llm-stats.com/benchmarks/mrcr-64k-4-needle", + "https://api.llm-stats.com/leaderboard/benchmarks/mrcr-64k-(4-needle)" + ], + "additional_details": { + "raw_benchmark_id": "mrcr-64k-(4-needle)", + "raw_model_id": "minicpm-sala", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MRCR (Multi-Round Coreference Resolution) at 64K context length with 4 needles. Models must navigate long conversations to reproduce specific model outputs, testing attention and reasoning across 64K-token contexts with 4 items to retrieve.", + "metric_id": "llm_stats.mrcr-64k-4-needle.score", + "metric_name": "MRCR 64K (4-needle) score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mrcr-64k-(4-needle)", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MRCR 64K (4-needle)", + "raw_categories": "[\"long_context\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.2057, + "details": { + "raw_score": "0.2057", + "raw_score_field": "score", + "raw_model_id": "minicpm-sala", + "raw_benchmark_id": "mrcr-64k-(4-needle)", + "source_urls_json": "[\"https://llm-stats.com/models/minicpm-sala\",\"https://llm-stats.com/benchmarks/mrcr-64k-4-needle\",\"https://api.llm-stats.com/leaderboard/benchmarks/mrcr-64k-(4-needle)\"]", + "raw_score_id": "mrcr-64k-(4-needle)::minicpm-sala", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mrcr-64k-8-needle::mrcr-64k-8-needle-minicpm-sala", + "evaluation_name": "llm_stats.mrcr-64k-8-needle", + "source_data": { + "dataset_name": "MRCR 64K (8-needle)", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minicpm-sala", + "https://llm-stats.com/benchmarks/mrcr-64k-8-needle", + "https://api.llm-stats.com/leaderboard/benchmarks/mrcr-64k-(8-needle)" + ], + "additional_details": { + "raw_benchmark_id": "mrcr-64k-(8-needle)", + "raw_model_id": "minicpm-sala", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MRCR (Multi-Round Coreference Resolution) at 64K context length with 8 needles. Models must navigate long conversations to reproduce specific model outputs, testing attention and reasoning across 64K-token contexts with 8 items to retrieve.", + "metric_id": "llm_stats.mrcr-64k-8-needle.score", + "metric_name": "MRCR 64K (8-needle) score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mrcr-64k-(8-needle)", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MRCR 64K (8-needle)", + "raw_categories": "[\"long_context\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.1656, + "details": { + "raw_score": "0.1656", + "raw_score_field": "score", + "raw_model_id": "minicpm-sala", + "raw_benchmark_id": "mrcr-64k-(8-needle)", + "source_urls_json": "[\"https://llm-stats.com/models/minicpm-sala\",\"https://llm-stats.com/benchmarks/mrcr-64k-8-needle\",\"https://api.llm-stats.com/leaderboard/benchmarks/mrcr-64k-(8-needle)\"]", + "raw_score_id": "mrcr-64k-(8-needle)::minicpm-sala", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "nolima-128k::nolima-128k-minicpm-sala", + "evaluation_name": "llm_stats.nolima-128k", + "source_data": { + "dataset_name": "NoLiMa 128K", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minicpm-sala", + "https://llm-stats.com/benchmarks/nolima-128k", + "https://api.llm-stats.com/leaderboard/benchmarks/nolima-128k" + ], + "additional_details": { + "raw_benchmark_id": "nolima-128k", + "raw_model_id": "minicpm-sala", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "NoLiMa evaluated at a 131072-token context length. Tests latent associative reasoning in long contexts with minimal lexical overlap between questions and needles.", + "metric_id": "llm_stats.nolima-128k.score", + "metric_name": "NoLiMa 128K score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "nolima-128k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "NoLiMa 128K", + "raw_categories": "[\"long_context\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.2386, + "details": { + "raw_score": "0.2386", + "raw_score_field": "score", + "raw_model_id": "minicpm-sala", + "raw_benchmark_id": "nolima-128k", + "source_urls_json": "[\"https://llm-stats.com/models/minicpm-sala\",\"https://llm-stats.com/benchmarks/nolima-128k\",\"https://api.llm-stats.com/leaderboard/benchmarks/nolima-128k\"]", + "raw_score_id": "nolima-128k::minicpm-sala", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "nolima-32k::nolima-32k-minicpm-sala", + "evaluation_name": "llm_stats.nolima-32k", + "source_data": { + "dataset_name": "NoLiMa 32K", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minicpm-sala", + "https://llm-stats.com/benchmarks/nolima-32k", + "https://api.llm-stats.com/leaderboard/benchmarks/nolima-32k" + ], + "additional_details": { + "raw_benchmark_id": "nolima-32k", + "raw_model_id": "minicpm-sala", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "NoLiMa evaluated at a 32768-token context length. Tests latent associative reasoning in long contexts with minimal lexical overlap between questions and needles.", + "metric_id": "llm_stats.nolima-32k.score", + "metric_name": "NoLiMa 32K score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "nolima-32k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "NoLiMa 32K", + "raw_categories": "[\"long_context\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.5454, + "details": { + "raw_score": "0.5454", + "raw_score_field": "score", + "raw_model_id": "minicpm-sala", + "raw_benchmark_id": "nolima-32k", + "source_urls_json": "[\"https://llm-stats.com/models/minicpm-sala\",\"https://llm-stats.com/benchmarks/nolima-32k\",\"https://api.llm-stats.com/leaderboard/benchmarks/nolima-32k\"]", + "raw_score_id": "nolima-32k::minicpm-sala", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "nolima-64k::nolima-64k-minicpm-sala", + "evaluation_name": "llm_stats.nolima-64k", + "source_data": { + "dataset_name": "NoLiMa 64K", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minicpm-sala", + "https://llm-stats.com/benchmarks/nolima-64k", + "https://api.llm-stats.com/leaderboard/benchmarks/nolima-64k" + ], + "additional_details": { + "raw_benchmark_id": "nolima-64k", + "raw_model_id": "minicpm-sala", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "NoLiMa evaluated at a 65536-token context length. Tests latent associative reasoning in long contexts with minimal lexical overlap between questions and needles.", + "metric_id": "llm_stats.nolima-64k.score", + "metric_name": "NoLiMa 64K score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "nolima-64k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "NoLiMa 64K", + "raw_categories": "[\"long_context\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.4295, + "details": { + "raw_score": "0.4295", + "raw_score_field": "score", + "raw_model_id": "minicpm-sala", + "raw_benchmark_id": "nolima-64k", + "source_urls_json": "[\"https://llm-stats.com/models/minicpm-sala\",\"https://llm-stats.com/benchmarks/nolima-64k\",\"https://api.llm-stats.com/leaderboard/benchmarks/nolima-64k\"]", + "raw_score_id": "nolima-64k::minicpm-sala", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ruler-1000k::ruler-1000k-minicpm-sala", + "evaluation_name": "llm_stats.ruler-1000k", + "source_data": { + "dataset_name": "RULER 1000K", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minicpm-sala", + "https://llm-stats.com/benchmarks/ruler-1000k", + "https://api.llm-stats.com/leaderboard/benchmarks/ruler-1000k" + ], + "additional_details": { + "raw_benchmark_id": "ruler-1000k", + "raw_model_id": "minicpm-sala", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "RULER 1000K evaluates the official 13-task RULER v1 suite at a 1048576-token (1M) context budget.", + "metric_id": "llm_stats.ruler-1000k.score", + "metric_name": "RULER 1000K score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ruler-1000k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "RULER 1000K", + "raw_categories": "[\"long_context\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.863, + "details": { + "raw_score": "0.863", + "raw_score_field": "score", + "raw_model_id": "minicpm-sala", + "raw_benchmark_id": "ruler-1000k", + "source_urls_json": "[\"https://llm-stats.com/models/minicpm-sala\",\"https://llm-stats.com/benchmarks/ruler-1000k\",\"https://api.llm-stats.com/leaderboard/benchmarks/ruler-1000k\"]", + "raw_score_id": "ruler-1000k::minicpm-sala", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ruler-128k::ruler-128k-minicpm-sala", + "evaluation_name": "llm_stats.ruler-128k", + "source_data": { + "dataset_name": "RULER 128k", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minicpm-sala", + "https://llm-stats.com/benchmarks/ruler-128k", + "https://api.llm-stats.com/leaderboard/benchmarks/ruler-128k" + ], + "additional_details": { + "raw_benchmark_id": "ruler-128k", + "raw_model_id": "minicpm-sala", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "RULER 128k evaluates the official 13-task RULER v1 suite at a 131072-token context budget.", + "metric_id": "llm_stats.ruler-128k.score", + "metric_name": "RULER 128k score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ruler-128k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "RULER 128k", + "raw_categories": "[\"long_context\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.8937, + "details": { + "raw_score": "0.8937", + "raw_score_field": "score", + "raw_model_id": "minicpm-sala", + "raw_benchmark_id": "ruler-128k", + "source_urls_json": "[\"https://llm-stats.com/models/minicpm-sala\",\"https://llm-stats.com/benchmarks/ruler-128k\",\"https://api.llm-stats.com/leaderboard/benchmarks/ruler-128k\"]", + "raw_score_id": "ruler-128k::minicpm-sala", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ruler-2048k::ruler-2048k-minicpm-sala", + "evaluation_name": "llm_stats.ruler-2048k", + "source_data": { + "dataset_name": "RULER 2048K", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minicpm-sala", + "https://llm-stats.com/benchmarks/ruler-2048k", + "https://api.llm-stats.com/leaderboard/benchmarks/ruler-2048k" + ], + "additional_details": { + "raw_benchmark_id": "ruler-2048k", + "raw_model_id": "minicpm-sala", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "RULER 2048K evaluates the official 13-task RULER v1 suite at a 2097152-token (2M) context budget.", + "metric_id": "llm_stats.ruler-2048k.score", + "metric_name": "RULER 2048K score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ruler-2048k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "RULER 2048K", + "raw_categories": "[\"long_context\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.816, + "details": { + "raw_score": "0.816", + "raw_score_field": "score", + "raw_model_id": "minicpm-sala", + "raw_benchmark_id": "ruler-2048k", + "source_urls_json": "[\"https://llm-stats.com/models/minicpm-sala\",\"https://llm-stats.com/benchmarks/ruler-2048k\",\"https://api.llm-stats.com/leaderboard/benchmarks/ruler-2048k\"]", + "raw_score_id": "ruler-2048k::minicpm-sala", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ruler-512k::ruler-512k-minicpm-sala", + "evaluation_name": "llm_stats.ruler-512k", + "source_data": { + "dataset_name": "RULER 512K", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minicpm-sala", + "https://llm-stats.com/benchmarks/ruler-512k", + "https://api.llm-stats.com/leaderboard/benchmarks/ruler-512k" + ], + "additional_details": { + "raw_benchmark_id": "ruler-512k", + "raw_model_id": "minicpm-sala", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "RULER 512K evaluates the official 13-task RULER v1 suite at a 524288-token context budget.", + "metric_id": "llm_stats.ruler-512k.score", + "metric_name": "RULER 512K score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ruler-512k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "RULER 512K", + "raw_categories": "[\"long_context\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.871, + "details": { + "raw_score": "0.871", + "raw_score_field": "score", + "raw_model_id": "minicpm-sala", + "raw_benchmark_id": "ruler-512k", + "source_urls_json": "[\"https://llm-stats.com/models/minicpm-sala\",\"https://llm-stats.com/benchmarks/ruler-512k\",\"https://api.llm-stats.com/leaderboard/benchmarks/ruler-512k\"]", + "raw_score_id": "ruler-512k::minicpm-sala", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ruler-64k::ruler-64k-minicpm-sala", + "evaluation_name": "llm_stats.ruler-64k", + "source_data": { + "dataset_name": "RULER 64k", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/minicpm-sala", + "https://llm-stats.com/benchmarks/ruler-64k", + "https://api.llm-stats.com/leaderboard/benchmarks/ruler-64k" + ], + "additional_details": { + "raw_benchmark_id": "ruler-64k", + "raw_model_id": "minicpm-sala", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "RULER 64k evaluates the official 13-task RULER v1 suite at a 65536-token context budget.", + "metric_id": "llm_stats.ruler-64k.score", + "metric_name": "RULER 64k score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ruler-64k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "RULER 64k", + "raw_categories": "[\"long_context\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.9265, + "details": { + "raw_score": "0.9265", + "raw_score_field": "score", + "raw_model_id": "minicpm-sala", + "raw_benchmark_id": "ruler-64k", + "source_urls_json": "[\"https://llm-stats.com/models/minicpm-sala\",\"https://llm-stats.com/benchmarks/ruler-64k\",\"https://api.llm-stats.com/leaderboard/benchmarks/ruler-64k\"]", + "raw_score_id": "ruler-64k::minicpm-sala", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/qwen/qwen-2.5-14b-instruct/6c2b628d-96d0-4737-a1e5-8ded0e240a0a.json b/data/llm_stats/qwen/qwen-2.5-14b-instruct/6c2b628d-96d0-4737-a1e5-8ded0e240a0a.json new file mode 100644 index 0000000000000000000000000000000000000000..6ce6186bf2d722a9143598b02edbb15cc9bf2b60 --- /dev/null +++ b/data/llm_stats/qwen/qwen-2.5-14b-instruct/6c2b628d-96d0-4737-a1e5-8ded0e240a0a.json @@ -0,0 +1,567 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/qwen_qwen-2.5-14b-instruct/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Qwen2.5 14B Instruct", + "id": "qwen/qwen-2.5-14b-instruct", + "developer": "qwen", + "additional_details": { + "raw_id": "qwen-2.5-14b-instruct", + "raw_name": "Qwen2.5 14B Instruct", + "raw_model_id": "qwen-2.5-14b-instruct", + "raw_model_name": "Qwen2.5 14B Instruct", + "raw_organization_id": "qwen", + "raw_organization_name": "Alibaba Cloud / Qwen Team", + "raw_release_date": "2024-09-19", + "raw_announcement_date": "2024-09-19", + "raw_multimodal": "false", + "raw_param_count": "14700000000", + "raw_provider_slug": "qwen", + "raw_provider_name": "Alibaba Cloud / Qwen Team" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "bbh::bbh-qwen-2.5-14b-instruct", + "evaluation_name": "llm_stats.bbh", + "source_data": { + "dataset_name": "BBH", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-14b-instruct", + "https://llm-stats.com/benchmarks/bbh", + "https://api.llm-stats.com/leaderboard/benchmarks/bbh" + ], + "additional_details": { + "raw_benchmark_id": "bbh", + "raw_model_id": "qwen-2.5-14b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Big-Bench Hard (BBH) is a suite of 23 challenging tasks selected from BIG-Bench for which prior language model evaluations did not outperform the average human-rater. These tasks require multi-step reasoning across diverse domains including arithmetic, logical reasoning, reading comprehension, and commonsense reasoning. The benchmark was designed to test capabilities believed to be beyond current language models and focuses on evaluating complex reasoning skills including temporal understanding, spatial reasoning, causal understanding, and deductive logical reasoning.", + "metric_id": "llm_stats.bbh.score", + "metric_name": "BBH score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bbh", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BBH", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.782, + "details": { + "raw_score": "0.782", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-14b-instruct", + "raw_benchmark_id": "bbh", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-14b-instruct\",\"https://llm-stats.com/benchmarks/bbh\",\"https://api.llm-stats.com/leaderboard/benchmarks/bbh\"]", + "raw_score_id": "bbh::qwen-2.5-14b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gsm8k::gsm8k-qwen-2.5-14b-instruct", + "evaluation_name": "llm_stats.gsm8k", + "source_data": { + "dataset_name": "GSM8k", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-14b-instruct", + "https://llm-stats.com/benchmarks/gsm8k", + "https://api.llm-stats.com/leaderboard/benchmarks/gsm8k" + ], + "additional_details": { + "raw_benchmark_id": "gsm8k", + "raw_model_id": "qwen-2.5-14b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Grade School Math 8K, a dataset of 8.5K high-quality linguistically diverse grade school math word problems requiring multi-step reasoning and elementary arithmetic operations.", + "metric_id": "llm_stats.gsm8k.score", + "metric_name": "GSM8k score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gsm8k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GSM8k", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "48" + } + }, + "score_details": { + "score": 0.948, + "details": { + "raw_score": "0.948", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-14b-instruct", + "raw_benchmark_id": "gsm8k", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-14b-instruct\",\"https://llm-stats.com/benchmarks/gsm8k\",\"https://api.llm-stats.com/leaderboard/benchmarks/gsm8k\"]", + "raw_score_id": "gsm8k::qwen-2.5-14b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "humaneval::humaneval-qwen-2.5-14b-instruct", + "evaluation_name": "llm_stats.humaneval", + "source_data": { + "dataset_name": "HumanEval+", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-14b-instruct", + "https://llm-stats.com/benchmarks/humaneval", + "https://api.llm-stats.com/leaderboard/benchmarks/humaneval+" + ], + "additional_details": { + "raw_benchmark_id": "humaneval+", + "raw_model_id": "qwen-2.5-14b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Enhanced version of HumanEval that extends the original test cases by 80x using EvalPlus framework for rigorous evaluation of LLM-synthesized code functional correctness, detecting previously undetected wrong code", + "metric_id": "llm_stats.humaneval.score", + "metric_name": "HumanEval+ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "humaneval+", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HumanEval+", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.512, + "details": { + "raw_score": "0.512", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-14b-instruct", + "raw_benchmark_id": "humaneval+", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-14b-instruct\",\"https://llm-stats.com/benchmarks/humaneval\",\"https://api.llm-stats.com/leaderboard/benchmarks/humaneval+\"]", + "raw_score_id": "humaneval+::qwen-2.5-14b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "math::math-qwen-2.5-14b-instruct", + "evaluation_name": "llm_stats.math", + "source_data": { + "dataset_name": "MATH", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-14b-instruct", + "https://llm-stats.com/benchmarks/math", + "https://api.llm-stats.com/leaderboard/benchmarks/math" + ], + "additional_details": { + "raw_benchmark_id": "math", + "raw_model_id": "qwen-2.5-14b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MATH dataset contains 12,500 challenging competition mathematics problems from AMC 10, AMC 12, AIME, and other mathematics competitions. Each problem includes full step-by-step solutions and spans multiple difficulty levels (1-5) across seven mathematical subjects including Prealgebra, Algebra, Number Theory, Counting and Probability, Geometry, Intermediate Algebra, and Precalculus.", + "metric_id": "llm_stats.math.score", + "metric_name": "MATH score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "math", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MATH", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "71" + } + }, + "score_details": { + "score": 0.8, + "details": { + "raw_score": "0.8", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-14b-instruct", + "raw_benchmark_id": "math", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-14b-instruct\",\"https://llm-stats.com/benchmarks/math\",\"https://api.llm-stats.com/leaderboard/benchmarks/math\"]", + "raw_score_id": "math::qwen-2.5-14b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mbpp::mbpp-qwen-2.5-14b-instruct", + "evaluation_name": "llm_stats.mbpp", + "source_data": { + "dataset_name": "MBPP", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-14b-instruct", + "https://llm-stats.com/benchmarks/mbpp", + "https://api.llm-stats.com/leaderboard/benchmarks/mbpp" + ], + "additional_details": { + "raw_benchmark_id": "mbpp", + "raw_model_id": "qwen-2.5-14b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MBPP (Mostly Basic Python Problems) is a benchmark of 974 crowd-sourced Python programming problems designed to be solvable by entry-level programmers. Each problem consists of a task description, code solution, and 3 automated test cases covering programming fundamentals and standard library functionality.", + "metric_id": "llm_stats.mbpp.score", + "metric_name": "MBPP score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mbpp", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MBPP", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "33" + } + }, + "score_details": { + "score": 0.82, + "details": { + "raw_score": "0.82", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-14b-instruct", + "raw_benchmark_id": "mbpp", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-14b-instruct\",\"https://llm-stats.com/benchmarks/mbpp\",\"https://api.llm-stats.com/leaderboard/benchmarks/mbpp\"]", + "raw_score_id": "mbpp::qwen-2.5-14b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mbpp::mbpp-qwen-2.5-14b-instruct", + "evaluation_name": "llm_stats.mbpp", + "source_data": { + "dataset_name": "MBPP+", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-14b-instruct", + "https://llm-stats.com/benchmarks/mbpp", + "https://api.llm-stats.com/leaderboard/benchmarks/mbpp+" + ], + "additional_details": { + "raw_benchmark_id": "mbpp+", + "raw_model_id": "qwen-2.5-14b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MBPP+ is an enhanced version of MBPP (Mostly Basic Python Problems) with significantly more test cases (35x) for more rigorous evaluation. MBPP is a benchmark of 974 crowd-sourced Python programming problems designed to be solvable by entry-level programmers, covering programming fundamentals and standard library functionality.", + "metric_id": "llm_stats.mbpp.score", + "metric_name": "MBPP+ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mbpp+", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MBPP+", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.632, + "details": { + "raw_score": "0.632", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-14b-instruct", + "raw_benchmark_id": "mbpp+", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-14b-instruct\",\"https://llm-stats.com/benchmarks/mbpp\",\"https://api.llm-stats.com/leaderboard/benchmarks/mbpp+\"]", + "raw_score_id": "mbpp+::qwen-2.5-14b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu-stem::mmlu-stem-qwen-2.5-14b-instruct", + "evaluation_name": "llm_stats.mmlu-stem", + "source_data": { + "dataset_name": "MMLU-STEM", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-14b-instruct", + "https://llm-stats.com/benchmarks/mmlu-stem", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu-stem" + ], + "additional_details": { + "raw_benchmark_id": "mmlu-stem", + "raw_model_id": "qwen-2.5-14b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "STEM-focused subset of the Massive Multitask Language Understanding benchmark, evaluating language models on science, technology, engineering, and mathematics topics including physics, chemistry, mathematics, and other technical subjects.", + "metric_id": "llm_stats.mmlu-stem.score", + "metric_name": "MMLU-STEM score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu-stem", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU-STEM", + "raw_categories": "[\"math\",\"physics\",\"reasoning\",\"chemistry\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.764, + "details": { + "raw_score": "0.764", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-14b-instruct", + "raw_benchmark_id": "mmlu-stem", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-14b-instruct\",\"https://llm-stats.com/benchmarks/mmlu-stem\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu-stem\"]", + "raw_score_id": "mmlu-stem::qwen-2.5-14b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "multipl-e::multipl-e-qwen-2.5-14b-instruct", + "evaluation_name": "llm_stats.multipl-e", + "source_data": { + "dataset_name": "MultiPL-E", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-14b-instruct", + "https://llm-stats.com/benchmarks/multipl-e", + "https://api.llm-stats.com/leaderboard/benchmarks/multipl-e" + ], + "additional_details": { + "raw_benchmark_id": "multipl-e", + "raw_model_id": "qwen-2.5-14b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MultiPL-E is a scalable and extensible system for translating unit test-driven code generation benchmarks to multiple programming languages. It extends HumanEval and MBPP Python benchmarks to 18 additional programming languages, enabling evaluation of neural code generation models across diverse programming paradigms and language features.", + "metric_id": "llm_stats.multipl-e.score", + "metric_name": "MultiPL-E score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "multipl-e", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MultiPL-E", + "raw_categories": "[\"language\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.728, + "details": { + "raw_score": "0.728", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-14b-instruct", + "raw_benchmark_id": "multipl-e", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-14b-instruct\",\"https://llm-stats.com/benchmarks/multipl-e\",\"https://api.llm-stats.com/leaderboard/benchmarks/multipl-e\"]", + "raw_score_id": "multipl-e::qwen-2.5-14b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "theoremqa::theoremqa-qwen-2.5-14b-instruct", + "evaluation_name": "llm_stats.theoremqa", + "source_data": { + "dataset_name": "TheoremQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-14b-instruct", + "https://llm-stats.com/benchmarks/theoremqa", + "https://api.llm-stats.com/leaderboard/benchmarks/theoremqa" + ], + "additional_details": { + "raw_benchmark_id": "theoremqa", + "raw_model_id": "qwen-2.5-14b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A theorem-driven question answering dataset containing 800 high-quality questions covering 350+ theorems from Math, Physics, EE&CS, and Finance. Designed to evaluate AI models' capabilities to apply theorems to solve challenging university-level science problems.", + "metric_id": "llm_stats.theoremqa.score", + "metric_name": "TheoremQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "theoremqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TheoremQA", + "raw_categories": "[\"math\",\"physics\",\"reasoning\",\"finance\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.43, + "details": { + "raw_score": "0.43", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-14b-instruct", + "raw_benchmark_id": "theoremqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-14b-instruct\",\"https://llm-stats.com/benchmarks/theoremqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/theoremqa\"]", + "raw_score_id": "theoremqa::qwen-2.5-14b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "truthfulqa::truthfulqa-qwen-2.5-14b-instruct", + "evaluation_name": "llm_stats.truthfulqa", + "source_data": { + "dataset_name": "TruthfulQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-14b-instruct", + "https://llm-stats.com/benchmarks/truthfulqa", + "https://api.llm-stats.com/leaderboard/benchmarks/truthfulqa" + ], + "additional_details": { + "raw_benchmark_id": "truthfulqa", + "raw_model_id": "qwen-2.5-14b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "TruthfulQA is a benchmark to measure whether language models are truthful in generating answers to questions. It comprises 817 questions that span 38 categories, including health, law, finance and politics. The questions are crafted such that some humans would answer falsely due to a false belief or misconception, testing models' ability to avoid generating false answers learned from human texts.", + "metric_id": "llm_stats.truthfulqa.score", + "metric_name": "TruthfulQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "truthfulqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TruthfulQA", + "raw_categories": "[\"legal\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.584, + "details": { + "raw_score": "0.584", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-14b-instruct", + "raw_benchmark_id": "truthfulqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-14b-instruct\",\"https://llm-stats.com/benchmarks/truthfulqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/truthfulqa\"]", + "raw_score_id": "truthfulqa::qwen-2.5-14b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/qwen/qwen-2.5-32b-instruct/736933e5-f503-4206-9ef4-3206245d6e9d.json b/data/llm_stats/qwen/qwen-2.5-32b-instruct/736933e5-f503-4206-9ef4-3206245d6e9d.json new file mode 100644 index 0000000000000000000000000000000000000000..5938f0a9ff9422ea4dce2c2cf43dd6ca8eba44cb --- /dev/null +++ b/data/llm_stats/qwen/qwen-2.5-32b-instruct/736933e5-f503-4206-9ef4-3206245d6e9d.json @@ -0,0 +1,671 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/qwen_qwen-2.5-32b-instruct/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Qwen2.5 32B Instruct", + "id": "qwen/qwen-2.5-32b-instruct", + "developer": "qwen", + "additional_details": { + "raw_id": "qwen-2.5-32b-instruct", + "raw_name": "Qwen2.5 32B Instruct", + "raw_model_id": "qwen-2.5-32b-instruct", + "raw_model_name": "Qwen2.5 32B Instruct", + "raw_organization_id": "qwen", + "raw_organization_name": "Alibaba Cloud / Qwen Team", + "raw_release_date": "2024-09-19", + "raw_announcement_date": "2024-09-19", + "raw_multimodal": "false", + "raw_param_count": "32500000000", + "raw_provider_slug": "qwen", + "raw_provider_name": "Alibaba Cloud / Qwen Team" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "bbh::bbh-qwen-2.5-32b-instruct", + "evaluation_name": "llm_stats.bbh", + "source_data": { + "dataset_name": "BBH", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-32b-instruct", + "https://llm-stats.com/benchmarks/bbh", + "https://api.llm-stats.com/leaderboard/benchmarks/bbh" + ], + "additional_details": { + "raw_benchmark_id": "bbh", + "raw_model_id": "qwen-2.5-32b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Big-Bench Hard (BBH) is a suite of 23 challenging tasks selected from BIG-Bench for which prior language model evaluations did not outperform the average human-rater. These tasks require multi-step reasoning across diverse domains including arithmetic, logical reasoning, reading comprehension, and commonsense reasoning. The benchmark was designed to test capabilities believed to be beyond current language models and focuses on evaluating complex reasoning skills including temporal understanding, spatial reasoning, causal understanding, and deductive logical reasoning.", + "metric_id": "llm_stats.bbh.score", + "metric_name": "BBH score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bbh", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BBH", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.845, + "details": { + "raw_score": "0.845", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-32b-instruct", + "raw_benchmark_id": "bbh", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-32b-instruct\",\"https://llm-stats.com/benchmarks/bbh\",\"https://api.llm-stats.com/leaderboard/benchmarks/bbh\"]", + "raw_score_id": "bbh::qwen-2.5-32b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gsm8k::gsm8k-qwen-2.5-32b-instruct", + "evaluation_name": "llm_stats.gsm8k", + "source_data": { + "dataset_name": "GSM8k", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-32b-instruct", + "https://llm-stats.com/benchmarks/gsm8k", + "https://api.llm-stats.com/leaderboard/benchmarks/gsm8k" + ], + "additional_details": { + "raw_benchmark_id": "gsm8k", + "raw_model_id": "qwen-2.5-32b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Grade School Math 8K, a dataset of 8.5K high-quality linguistically diverse grade school math word problems requiring multi-step reasoning and elementary arithmetic operations.", + "metric_id": "llm_stats.gsm8k.score", + "metric_name": "GSM8k score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gsm8k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GSM8k", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "48" + } + }, + "score_details": { + "score": 0.959, + "details": { + "raw_score": "0.959", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-32b-instruct", + "raw_benchmark_id": "gsm8k", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-32b-instruct\",\"https://llm-stats.com/benchmarks/gsm8k\",\"https://api.llm-stats.com/leaderboard/benchmarks/gsm8k\"]", + "raw_score_id": "gsm8k::qwen-2.5-32b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hellaswag::hellaswag-qwen-2.5-32b-instruct", + "evaluation_name": "llm_stats.hellaswag", + "source_data": { + "dataset_name": "HellaSwag", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-32b-instruct", + "https://llm-stats.com/benchmarks/hellaswag", + "https://api.llm-stats.com/leaderboard/benchmarks/hellaswag" + ], + "additional_details": { + "raw_benchmark_id": "hellaswag", + "raw_model_id": "qwen-2.5-32b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A challenging commonsense natural language inference dataset that uses Adversarial Filtering to create questions trivial for humans (>95% accuracy) but difficult for state-of-the-art models, requiring completion of sentence endings based on physical situations and everyday activities", + "metric_id": "llm_stats.hellaswag.score", + "metric_name": "HellaSwag score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hellaswag", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HellaSwag", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.852, + "details": { + "raw_score": "0.852", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-32b-instruct", + "raw_benchmark_id": "hellaswag", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-32b-instruct\",\"https://llm-stats.com/benchmarks/hellaswag\",\"https://api.llm-stats.com/leaderboard/benchmarks/hellaswag\"]", + "raw_score_id": "hellaswag::qwen-2.5-32b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "humaneval::humaneval-qwen-2.5-32b-instruct", + "evaluation_name": "llm_stats.humaneval", + "source_data": { + "dataset_name": "HumanEval+", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-32b-instruct", + "https://llm-stats.com/benchmarks/humaneval", + "https://api.llm-stats.com/leaderboard/benchmarks/humaneval+" + ], + "additional_details": { + "raw_benchmark_id": "humaneval+", + "raw_model_id": "qwen-2.5-32b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Enhanced version of HumanEval that extends the original test cases by 80x using EvalPlus framework for rigorous evaluation of LLM-synthesized code functional correctness, detecting previously undetected wrong code", + "metric_id": "llm_stats.humaneval.score", + "metric_name": "HumanEval+ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "humaneval+", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HumanEval+", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.524, + "details": { + "raw_score": "0.524", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-32b-instruct", + "raw_benchmark_id": "humaneval+", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-32b-instruct\",\"https://llm-stats.com/benchmarks/humaneval\",\"https://api.llm-stats.com/leaderboard/benchmarks/humaneval+\"]", + "raw_score_id": "humaneval+::qwen-2.5-32b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "math::math-qwen-2.5-32b-instruct", + "evaluation_name": "llm_stats.math", + "source_data": { + "dataset_name": "MATH", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-32b-instruct", + "https://llm-stats.com/benchmarks/math", + "https://api.llm-stats.com/leaderboard/benchmarks/math" + ], + "additional_details": { + "raw_benchmark_id": "math", + "raw_model_id": "qwen-2.5-32b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MATH dataset contains 12,500 challenging competition mathematics problems from AMC 10, AMC 12, AIME, and other mathematics competitions. Each problem includes full step-by-step solutions and spans multiple difficulty levels (1-5) across seven mathematical subjects including Prealgebra, Algebra, Number Theory, Counting and Probability, Geometry, Intermediate Algebra, and Precalculus.", + "metric_id": "llm_stats.math.score", + "metric_name": "MATH score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "math", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MATH", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "71" + } + }, + "score_details": { + "score": 0.831, + "details": { + "raw_score": "0.831", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-32b-instruct", + "raw_benchmark_id": "math", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-32b-instruct\",\"https://llm-stats.com/benchmarks/math\",\"https://api.llm-stats.com/leaderboard/benchmarks/math\"]", + "raw_score_id": "math::qwen-2.5-32b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mbpp::mbpp-qwen-2.5-32b-instruct", + "evaluation_name": "llm_stats.mbpp", + "source_data": { + "dataset_name": "MBPP", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-32b-instruct", + "https://llm-stats.com/benchmarks/mbpp", + "https://api.llm-stats.com/leaderboard/benchmarks/mbpp" + ], + "additional_details": { + "raw_benchmark_id": "mbpp", + "raw_model_id": "qwen-2.5-32b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MBPP (Mostly Basic Python Problems) is a benchmark of 974 crowd-sourced Python programming problems designed to be solvable by entry-level programmers. Each problem consists of a task description, code solution, and 3 automated test cases covering programming fundamentals and standard library functionality.", + "metric_id": "llm_stats.mbpp.score", + "metric_name": "MBPP score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mbpp", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MBPP", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "33" + } + }, + "score_details": { + "score": 0.84, + "details": { + "raw_score": "0.84", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-32b-instruct", + "raw_benchmark_id": "mbpp", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-32b-instruct\",\"https://llm-stats.com/benchmarks/mbpp\",\"https://api.llm-stats.com/leaderboard/benchmarks/mbpp\"]", + "raw_score_id": "mbpp::qwen-2.5-32b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mbpp::mbpp-qwen-2.5-32b-instruct", + "evaluation_name": "llm_stats.mbpp", + "source_data": { + "dataset_name": "MBPP+", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-32b-instruct", + "https://llm-stats.com/benchmarks/mbpp", + "https://api.llm-stats.com/leaderboard/benchmarks/mbpp+" + ], + "additional_details": { + "raw_benchmark_id": "mbpp+", + "raw_model_id": "qwen-2.5-32b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MBPP+ is an enhanced version of MBPP (Mostly Basic Python Problems) with significantly more test cases (35x) for more rigorous evaluation. MBPP is a benchmark of 974 crowd-sourced Python programming problems designed to be solvable by entry-level programmers, covering programming fundamentals and standard library functionality.", + "metric_id": "llm_stats.mbpp.score", + "metric_name": "MBPP+ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mbpp+", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MBPP+", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.672, + "details": { + "raw_score": "0.672", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-32b-instruct", + "raw_benchmark_id": "mbpp+", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-32b-instruct\",\"https://llm-stats.com/benchmarks/mbpp\",\"https://api.llm-stats.com/leaderboard/benchmarks/mbpp+\"]", + "raw_score_id": "mbpp+::qwen-2.5-32b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu-stem::mmlu-stem-qwen-2.5-32b-instruct", + "evaluation_name": "llm_stats.mmlu-stem", + "source_data": { + "dataset_name": "MMLU-STEM", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-32b-instruct", + "https://llm-stats.com/benchmarks/mmlu-stem", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu-stem" + ], + "additional_details": { + "raw_benchmark_id": "mmlu-stem", + "raw_model_id": "qwen-2.5-32b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "STEM-focused subset of the Massive Multitask Language Understanding benchmark, evaluating language models on science, technology, engineering, and mathematics topics including physics, chemistry, mathematics, and other technical subjects.", + "metric_id": "llm_stats.mmlu-stem.score", + "metric_name": "MMLU-STEM score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu-stem", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU-STEM", + "raw_categories": "[\"math\",\"physics\",\"reasoning\",\"chemistry\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.809, + "details": { + "raw_score": "0.809", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-32b-instruct", + "raw_benchmark_id": "mmlu-stem", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-32b-instruct\",\"https://llm-stats.com/benchmarks/mmlu-stem\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu-stem\"]", + "raw_score_id": "mmlu-stem::qwen-2.5-32b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "multipl-e::multipl-e-qwen-2.5-32b-instruct", + "evaluation_name": "llm_stats.multipl-e", + "source_data": { + "dataset_name": "MultiPL-E", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-32b-instruct", + "https://llm-stats.com/benchmarks/multipl-e", + "https://api.llm-stats.com/leaderboard/benchmarks/multipl-e" + ], + "additional_details": { + "raw_benchmark_id": "multipl-e", + "raw_model_id": "qwen-2.5-32b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MultiPL-E is a scalable and extensible system for translating unit test-driven code generation benchmarks to multiple programming languages. It extends HumanEval and MBPP Python benchmarks to 18 additional programming languages, enabling evaluation of neural code generation models across diverse programming paradigms and language features.", + "metric_id": "llm_stats.multipl-e.score", + "metric_name": "MultiPL-E score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "multipl-e", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MultiPL-E", + "raw_categories": "[\"language\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.754, + "details": { + "raw_score": "0.754", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-32b-instruct", + "raw_benchmark_id": "multipl-e", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-32b-instruct\",\"https://llm-stats.com/benchmarks/multipl-e\",\"https://api.llm-stats.com/leaderboard/benchmarks/multipl-e\"]", + "raw_score_id": "multipl-e::qwen-2.5-32b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "theoremqa::theoremqa-qwen-2.5-32b-instruct", + "evaluation_name": "llm_stats.theoremqa", + "source_data": { + "dataset_name": "TheoremQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-32b-instruct", + "https://llm-stats.com/benchmarks/theoremqa", + "https://api.llm-stats.com/leaderboard/benchmarks/theoremqa" + ], + "additional_details": { + "raw_benchmark_id": "theoremqa", + "raw_model_id": "qwen-2.5-32b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A theorem-driven question answering dataset containing 800 high-quality questions covering 350+ theorems from Math, Physics, EE&CS, and Finance. Designed to evaluate AI models' capabilities to apply theorems to solve challenging university-level science problems.", + "metric_id": "llm_stats.theoremqa.score", + "metric_name": "TheoremQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "theoremqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TheoremQA", + "raw_categories": "[\"math\",\"physics\",\"reasoning\",\"finance\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.441, + "details": { + "raw_score": "0.441", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-32b-instruct", + "raw_benchmark_id": "theoremqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-32b-instruct\",\"https://llm-stats.com/benchmarks/theoremqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/theoremqa\"]", + "raw_score_id": "theoremqa::qwen-2.5-32b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "truthfulqa::truthfulqa-qwen-2.5-32b-instruct", + "evaluation_name": "llm_stats.truthfulqa", + "source_data": { + "dataset_name": "TruthfulQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-32b-instruct", + "https://llm-stats.com/benchmarks/truthfulqa", + "https://api.llm-stats.com/leaderboard/benchmarks/truthfulqa" + ], + "additional_details": { + "raw_benchmark_id": "truthfulqa", + "raw_model_id": "qwen-2.5-32b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "TruthfulQA is a benchmark to measure whether language models are truthful in generating answers to questions. It comprises 817 questions that span 38 categories, including health, law, finance and politics. The questions are crafted such that some humans would answer falsely due to a false belief or misconception, testing models' ability to avoid generating false answers learned from human texts.", + "metric_id": "llm_stats.truthfulqa.score", + "metric_name": "TruthfulQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "truthfulqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TruthfulQA", + "raw_categories": "[\"legal\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.578, + "details": { + "raw_score": "0.578", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-32b-instruct", + "raw_benchmark_id": "truthfulqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-32b-instruct\",\"https://llm-stats.com/benchmarks/truthfulqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/truthfulqa\"]", + "raw_score_id": "truthfulqa::qwen-2.5-32b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "winogrande::winogrande-qwen-2.5-32b-instruct", + "evaluation_name": "llm_stats.winogrande", + "source_data": { + "dataset_name": "Winogrande", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-32b-instruct", + "https://llm-stats.com/benchmarks/winogrande", + "https://api.llm-stats.com/leaderboard/benchmarks/winogrande" + ], + "additional_details": { + "raw_benchmark_id": "winogrande", + "raw_model_id": "qwen-2.5-32b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WinoGrande: An Adversarial Winograd Schema Challenge at Scale. A large-scale dataset of 44,000 pronoun resolution problems designed to test machine commonsense reasoning. Uses adversarial filtering to reduce spurious biases and provides a more robust evaluation of whether AI systems truly understand commonsense or exploit statistical shortcuts. Current best AI methods achieve 59.4-79.1% accuracy, significantly below human performance of 94.0%.", + "metric_id": "llm_stats.winogrande.score", + "metric_name": "Winogrande score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "winogrande", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Winogrande", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "22" + } + }, + "score_details": { + "score": 0.82, + "details": { + "raw_score": "0.82", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-32b-instruct", + "raw_benchmark_id": "winogrande", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-32b-instruct\",\"https://llm-stats.com/benchmarks/winogrande\",\"https://api.llm-stats.com/leaderboard/benchmarks/winogrande\"]", + "raw_score_id": "winogrande::qwen-2.5-32b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/qwen/qwen-2.5-72b-instruct/bddc4d45-0c75-49f4-8290-47f46835d855.json b/data/llm_stats/qwen/qwen-2.5-72b-instruct/bddc4d45-0c75-49f4-8290-47f46835d855.json new file mode 100644 index 0000000000000000000000000000000000000000..06dd5a5e1fa7d2f4ef5e5dcce1cb02cc19fb9a9f --- /dev/null +++ b/data/llm_stats/qwen/qwen-2.5-72b-instruct/bddc4d45-0c75-49f4-8290-47f46835d855.json @@ -0,0 +1,411 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/qwen_qwen-2.5-72b-instruct/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Qwen2.5 72B Instruct", + "id": "qwen/qwen-2.5-72b-instruct", + "developer": "qwen", + "additional_details": { + "raw_id": "qwen-2.5-72b-instruct", + "raw_name": "Qwen2.5 72B Instruct", + "raw_model_id": "qwen-2.5-72b-instruct", + "raw_model_name": "Qwen2.5 72B Instruct", + "raw_organization_id": "qwen", + "raw_organization_name": "Alibaba Cloud / Qwen Team", + "raw_release_date": "2024-09-19", + "raw_announcement_date": "2024-09-19", + "raw_multimodal": "false", + "raw_param_count": "72700000000", + "raw_provider_slug": "qwen", + "raw_provider_name": "Alibaba Cloud / Qwen Team" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "alignbench::alignbench-qwen-2.5-72b-instruct", + "evaluation_name": "llm_stats.alignbench", + "source_data": { + "dataset_name": "AlignBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-72b-instruct", + "https://llm-stats.com/benchmarks/alignbench", + "https://api.llm-stats.com/leaderboard/benchmarks/alignbench" + ], + "additional_details": { + "raw_benchmark_id": "alignbench", + "raw_model_id": "qwen-2.5-72b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "AlignBench is a comprehensive multi-dimensional benchmark for evaluating Chinese alignment of Large Language Models. It contains 8 main categories: Fundamental Language Ability, Advanced Chinese Understanding, Open-ended Questions, Writing Ability, Logical Reasoning, Mathematics, Task-oriented Role Play, and Professional Knowledge. The benchmark includes 683 real-scenario rooted queries with human-verified references and uses a rule-calibrated multi-dimensional LLM-as-Judge approach with Chain-of-Thought for evaluation.", + "metric_id": "llm_stats.alignbench.score", + "metric_name": "AlignBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "alignbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AlignBench", + "raw_categories": "[\"language\",\"math\",\"reasoning\",\"roleplay\",\"general\",\"creativity\",\"writing\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.816, + "details": { + "raw_score": "0.816", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-72b-instruct", + "raw_benchmark_id": "alignbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-72b-instruct\",\"https://llm-stats.com/benchmarks/alignbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/alignbench\"]", + "raw_score_id": "alignbench::qwen-2.5-72b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "arena-hard::arena-hard-qwen-2.5-72b-instruct", + "evaluation_name": "llm_stats.arena-hard", + "source_data": { + "dataset_name": "Arena Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-72b-instruct", + "https://llm-stats.com/benchmarks/arena-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/arena-hard" + ], + "additional_details": { + "raw_benchmark_id": "arena-hard", + "raw_model_id": "qwen-2.5-72b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Arena-Hard-Auto is an automatic evaluation benchmark for instruction-tuned LLMs consisting of 500 challenging real-world prompts curated by BenchBuilder. It includes open-ended software engineering problems, mathematical questions, and creative writing tasks. The benchmark uses LLM-as-a-Judge methodology with GPT-4.1 and Gemini-2.5 as automatic judges to approximate human preference. Arena-Hard achieves 98.6% correlation with human preference rankings and provides 3x higher separation of model performances compared to MT-Bench, making it highly effective for distinguishing between models of similar quality.", + "metric_id": "llm_stats.arena-hard.score", + "metric_name": "Arena Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arena-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Arena Hard", + "raw_categories": "[\"reasoning\",\"general\",\"creativity\",\"writing\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "26" + } + }, + "score_details": { + "score": 0.812, + "details": { + "raw_score": "0.812", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-72b-instruct", + "raw_benchmark_id": "arena-hard", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-72b-instruct\",\"https://llm-stats.com/benchmarks/arena-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/arena-hard\"]", + "raw_score_id": "arena-hard::qwen-2.5-72b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gsm8k::gsm8k-qwen-2.5-72b-instruct", + "evaluation_name": "llm_stats.gsm8k", + "source_data": { + "dataset_name": "GSM8k", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-72b-instruct", + "https://llm-stats.com/benchmarks/gsm8k", + "https://api.llm-stats.com/leaderboard/benchmarks/gsm8k" + ], + "additional_details": { + "raw_benchmark_id": "gsm8k", + "raw_model_id": "qwen-2.5-72b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Grade School Math 8K, a dataset of 8.5K high-quality linguistically diverse grade school math word problems requiring multi-step reasoning and elementary arithmetic operations.", + "metric_id": "llm_stats.gsm8k.score", + "metric_name": "GSM8k score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gsm8k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GSM8k", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "48" + } + }, + "score_details": { + "score": 0.958, + "details": { + "raw_score": "0.958", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-72b-instruct", + "raw_benchmark_id": "gsm8k", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-72b-instruct\",\"https://llm-stats.com/benchmarks/gsm8k\",\"https://api.llm-stats.com/leaderboard/benchmarks/gsm8k\"]", + "raw_score_id": "gsm8k::qwen-2.5-72b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "math::math-qwen-2.5-72b-instruct", + "evaluation_name": "llm_stats.math", + "source_data": { + "dataset_name": "MATH", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-72b-instruct", + "https://llm-stats.com/benchmarks/math", + "https://api.llm-stats.com/leaderboard/benchmarks/math" + ], + "additional_details": { + "raw_benchmark_id": "math", + "raw_model_id": "qwen-2.5-72b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MATH dataset contains 12,500 challenging competition mathematics problems from AMC 10, AMC 12, AIME, and other mathematics competitions. Each problem includes full step-by-step solutions and spans multiple difficulty levels (1-5) across seven mathematical subjects including Prealgebra, Algebra, Number Theory, Counting and Probability, Geometry, Intermediate Algebra, and Precalculus.", + "metric_id": "llm_stats.math.score", + "metric_name": "MATH score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "math", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MATH", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "71" + } + }, + "score_details": { + "score": 0.831, + "details": { + "raw_score": "0.831", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-72b-instruct", + "raw_benchmark_id": "math", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-72b-instruct\",\"https://llm-stats.com/benchmarks/math\",\"https://api.llm-stats.com/leaderboard/benchmarks/math\"]", + "raw_score_id": "math::qwen-2.5-72b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mbpp::mbpp-qwen-2.5-72b-instruct", + "evaluation_name": "llm_stats.mbpp", + "source_data": { + "dataset_name": "MBPP", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-72b-instruct", + "https://llm-stats.com/benchmarks/mbpp", + "https://api.llm-stats.com/leaderboard/benchmarks/mbpp" + ], + "additional_details": { + "raw_benchmark_id": "mbpp", + "raw_model_id": "qwen-2.5-72b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MBPP (Mostly Basic Python Problems) is a benchmark of 974 crowd-sourced Python programming problems designed to be solvable by entry-level programmers. Each problem consists of a task description, code solution, and 3 automated test cases covering programming fundamentals and standard library functionality.", + "metric_id": "llm_stats.mbpp.score", + "metric_name": "MBPP score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mbpp", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MBPP", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "33" + } + }, + "score_details": { + "score": 0.882, + "details": { + "raw_score": "0.882", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-72b-instruct", + "raw_benchmark_id": "mbpp", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-72b-instruct\",\"https://llm-stats.com/benchmarks/mbpp\",\"https://api.llm-stats.com/leaderboard/benchmarks/mbpp\"]", + "raw_score_id": "mbpp::qwen-2.5-72b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mt-bench::mt-bench-qwen-2.5-72b-instruct", + "evaluation_name": "llm_stats.mt-bench", + "source_data": { + "dataset_name": "MT-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-72b-instruct", + "https://llm-stats.com/benchmarks/mt-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/mt-bench" + ], + "additional_details": { + "raw_benchmark_id": "mt-bench", + "raw_model_id": "qwen-2.5-72b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MT-Bench is a challenging multi-turn benchmark that measures the ability of large language models to engage in coherent, informative, and engaging conversations. It uses strong LLMs as judges for scalable and explainable evaluation of multi-turn dialogue capabilities.", + "metric_id": "llm_stats.mt-bench.score", + "metric_name": "MT-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mt-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MT-Bench", + "raw_categories": "[\"reasoning\",\"roleplay\",\"general\",\"communication\",\"creativity\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.935, + "details": { + "raw_score": "0.935", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-72b-instruct", + "raw_benchmark_id": "mt-bench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-72b-instruct\",\"https://llm-stats.com/benchmarks/mt-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/mt-bench\"]", + "raw_score_id": "mt-bench::qwen-2.5-72b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "multipl-e::multipl-e-qwen-2.5-72b-instruct", + "evaluation_name": "llm_stats.multipl-e", + "source_data": { + "dataset_name": "MultiPL-E", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-72b-instruct", + "https://llm-stats.com/benchmarks/multipl-e", + "https://api.llm-stats.com/leaderboard/benchmarks/multipl-e" + ], + "additional_details": { + "raw_benchmark_id": "multipl-e", + "raw_model_id": "qwen-2.5-72b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MultiPL-E is a scalable and extensible system for translating unit test-driven code generation benchmarks to multiple programming languages. It extends HumanEval and MBPP Python benchmarks to 18 additional programming languages, enabling evaluation of neural code generation models across diverse programming paradigms and language features.", + "metric_id": "llm_stats.multipl-e.score", + "metric_name": "MultiPL-E score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "multipl-e", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MultiPL-E", + "raw_categories": "[\"language\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.751, + "details": { + "raw_score": "0.751", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-72b-instruct", + "raw_benchmark_id": "multipl-e", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-72b-instruct\",\"https://llm-stats.com/benchmarks/multipl-e\",\"https://api.llm-stats.com/leaderboard/benchmarks/multipl-e\"]", + "raw_score_id": "multipl-e::qwen-2.5-72b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/qwen/qwen-2.5-7b-instruct/123f24bb-c890-4660-9892-3bbb10e44ed0.json b/data/llm_stats/qwen/qwen-2.5-7b-instruct/123f24bb-c890-4660-9892-3bbb10e44ed0.json new file mode 100644 index 0000000000000000000000000000000000000000..6c6559f281136acf725bf548d533391a4ad6c6c9 --- /dev/null +++ b/data/llm_stats/qwen/qwen-2.5-7b-instruct/123f24bb-c890-4660-9892-3bbb10e44ed0.json @@ -0,0 +1,307 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/qwen_qwen-2.5-7b-instruct/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Qwen2.5 7B Instruct", + "id": "qwen/qwen-2.5-7b-instruct", + "developer": "qwen", + "additional_details": { + "raw_id": "qwen-2.5-7b-instruct", + "raw_name": "Qwen2.5 7B Instruct", + "raw_model_id": "qwen-2.5-7b-instruct", + "raw_model_name": "Qwen2.5 7B Instruct", + "raw_organization_id": "qwen", + "raw_organization_name": "Alibaba Cloud / Qwen Team", + "raw_release_date": "2024-09-19", + "raw_announcement_date": "2024-09-19", + "raw_multimodal": "false", + "raw_param_count": "7610000000", + "raw_provider_slug": "qwen", + "raw_provider_name": "Alibaba Cloud / Qwen Team" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "alignbench::alignbench-qwen-2.5-7b-instruct", + "evaluation_name": "llm_stats.alignbench", + "source_data": { + "dataset_name": "AlignBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-7b-instruct", + "https://llm-stats.com/benchmarks/alignbench", + "https://api.llm-stats.com/leaderboard/benchmarks/alignbench" + ], + "additional_details": { + "raw_benchmark_id": "alignbench", + "raw_model_id": "qwen-2.5-7b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "AlignBench is a comprehensive multi-dimensional benchmark for evaluating Chinese alignment of Large Language Models. It contains 8 main categories: Fundamental Language Ability, Advanced Chinese Understanding, Open-ended Questions, Writing Ability, Logical Reasoning, Mathematics, Task-oriented Role Play, and Professional Knowledge. The benchmark includes 683 real-scenario rooted queries with human-verified references and uses a rule-calibrated multi-dimensional LLM-as-Judge approach with Chain-of-Thought for evaluation.", + "metric_id": "llm_stats.alignbench.score", + "metric_name": "AlignBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "alignbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AlignBench", + "raw_categories": "[\"language\",\"math\",\"reasoning\",\"roleplay\",\"general\",\"creativity\",\"writing\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.733, + "details": { + "raw_score": "0.733", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-7b-instruct", + "raw_benchmark_id": "alignbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-7b-instruct\",\"https://llm-stats.com/benchmarks/alignbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/alignbench\"]", + "raw_score_id": "alignbench::qwen-2.5-7b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "arena-hard::arena-hard-qwen-2.5-7b-instruct", + "evaluation_name": "llm_stats.arena-hard", + "source_data": { + "dataset_name": "Arena Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-7b-instruct", + "https://llm-stats.com/benchmarks/arena-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/arena-hard" + ], + "additional_details": { + "raw_benchmark_id": "arena-hard", + "raw_model_id": "qwen-2.5-7b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Arena-Hard-Auto is an automatic evaluation benchmark for instruction-tuned LLMs consisting of 500 challenging real-world prompts curated by BenchBuilder. It includes open-ended software engineering problems, mathematical questions, and creative writing tasks. The benchmark uses LLM-as-a-Judge methodology with GPT-4.1 and Gemini-2.5 as automatic judges to approximate human preference. Arena-Hard achieves 98.6% correlation with human preference rankings and provides 3x higher separation of model performances compared to MT-Bench, making it highly effective for distinguishing between models of similar quality.", + "metric_id": "llm_stats.arena-hard.score", + "metric_name": "Arena Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arena-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Arena Hard", + "raw_categories": "[\"reasoning\",\"general\",\"creativity\",\"writing\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "26" + } + }, + "score_details": { + "score": 0.52, + "details": { + "raw_score": "0.52", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-7b-instruct", + "raw_benchmark_id": "arena-hard", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-7b-instruct\",\"https://llm-stats.com/benchmarks/arena-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/arena-hard\"]", + "raw_score_id": "arena-hard::qwen-2.5-7b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mbpp::mbpp-qwen-2.5-7b-instruct", + "evaluation_name": "llm_stats.mbpp", + "source_data": { + "dataset_name": "MBPP", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-7b-instruct", + "https://llm-stats.com/benchmarks/mbpp", + "https://api.llm-stats.com/leaderboard/benchmarks/mbpp" + ], + "additional_details": { + "raw_benchmark_id": "mbpp", + "raw_model_id": "qwen-2.5-7b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MBPP (Mostly Basic Python Problems) is a benchmark of 974 crowd-sourced Python programming problems designed to be solvable by entry-level programmers. Each problem consists of a task description, code solution, and 3 automated test cases covering programming fundamentals and standard library functionality.", + "metric_id": "llm_stats.mbpp.score", + "metric_name": "MBPP score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mbpp", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MBPP", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "33" + } + }, + "score_details": { + "score": 0.792, + "details": { + "raw_score": "0.792", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-7b-instruct", + "raw_benchmark_id": "mbpp", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-7b-instruct\",\"https://llm-stats.com/benchmarks/mbpp\",\"https://api.llm-stats.com/leaderboard/benchmarks/mbpp\"]", + "raw_score_id": "mbpp::qwen-2.5-7b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mt-bench::mt-bench-qwen-2.5-7b-instruct", + "evaluation_name": "llm_stats.mt-bench", + "source_data": { + "dataset_name": "MT-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-7b-instruct", + "https://llm-stats.com/benchmarks/mt-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/mt-bench" + ], + "additional_details": { + "raw_benchmark_id": "mt-bench", + "raw_model_id": "qwen-2.5-7b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MT-Bench is a challenging multi-turn benchmark that measures the ability of large language models to engage in coherent, informative, and engaging conversations. It uses strong LLMs as judges for scalable and explainable evaluation of multi-turn dialogue capabilities.", + "metric_id": "llm_stats.mt-bench.score", + "metric_name": "MT-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mt-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MT-Bench", + "raw_categories": "[\"reasoning\",\"roleplay\",\"general\",\"communication\",\"creativity\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.875, + "details": { + "raw_score": "0.875", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-7b-instruct", + "raw_benchmark_id": "mt-bench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-7b-instruct\",\"https://llm-stats.com/benchmarks/mt-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/mt-bench\"]", + "raw_score_id": "mt-bench::qwen-2.5-7b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "multipl-e::multipl-e-qwen-2.5-7b-instruct", + "evaluation_name": "llm_stats.multipl-e", + "source_data": { + "dataset_name": "MultiPL-E", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen-2.5-7b-instruct", + "https://llm-stats.com/benchmarks/multipl-e", + "https://api.llm-stats.com/leaderboard/benchmarks/multipl-e" + ], + "additional_details": { + "raw_benchmark_id": "multipl-e", + "raw_model_id": "qwen-2.5-7b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MultiPL-E is a scalable and extensible system for translating unit test-driven code generation benchmarks to multiple programming languages. It extends HumanEval and MBPP Python benchmarks to 18 additional programming languages, enabling evaluation of neural code generation models across diverse programming paradigms and language features.", + "metric_id": "llm_stats.multipl-e.score", + "metric_name": "MultiPL-E score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "multipl-e", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MultiPL-E", + "raw_categories": "[\"language\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.704, + "details": { + "raw_score": "0.704", + "raw_score_field": "score", + "raw_model_id": "qwen-2.5-7b-instruct", + "raw_benchmark_id": "multipl-e", + "source_urls_json": "[\"https://llm-stats.com/models/qwen-2.5-7b-instruct\",\"https://llm-stats.com/benchmarks/multipl-e\",\"https://api.llm-stats.com/leaderboard/benchmarks/multipl-e\"]", + "raw_score_id": "multipl-e::qwen-2.5-7b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/qwen/qwen2-72b-instruct/92d73112-2c43-46bf-a6df-f7b547f70f12.json b/data/llm_stats/qwen/qwen2-72b-instruct/92d73112-2c43-46bf-a6df-f7b547f70f12.json new file mode 100644 index 0000000000000000000000000000000000000000..f1b5fe388616c2f792a6f32f89474a184a0e3257 --- /dev/null +++ b/data/llm_stats/qwen/qwen2-72b-instruct/92d73112-2c43-46bf-a6df-f7b547f70f12.json @@ -0,0 +1,567 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/qwen_qwen2-72b-instruct/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Qwen2 72B Instruct", + "id": "qwen/qwen2-72b-instruct", + "developer": "qwen", + "additional_details": { + "raw_id": "qwen2-72b-instruct", + "raw_name": "Qwen2 72B Instruct", + "raw_model_id": "qwen2-72b-instruct", + "raw_model_name": "Qwen2 72B Instruct", + "raw_organization_id": "qwen", + "raw_organization_name": "Alibaba Cloud / Qwen Team", + "raw_release_date": "2024-07-23", + "raw_announcement_date": "2024-07-23", + "raw_multimodal": "false", + "raw_param_count": "72000000000", + "raw_provider_slug": "qwen", + "raw_provider_name": "Alibaba Cloud / Qwen Team" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "bbh::bbh-qwen2-72b-instruct", + "evaluation_name": "llm_stats.bbh", + "source_data": { + "dataset_name": "BBH", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen2-72b-instruct", + "https://llm-stats.com/benchmarks/bbh", + "https://api.llm-stats.com/leaderboard/benchmarks/bbh" + ], + "additional_details": { + "raw_benchmark_id": "bbh", + "raw_model_id": "qwen2-72b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Big-Bench Hard (BBH) is a suite of 23 challenging tasks selected from BIG-Bench for which prior language model evaluations did not outperform the average human-rater. These tasks require multi-step reasoning across diverse domains including arithmetic, logical reasoning, reading comprehension, and commonsense reasoning. The benchmark was designed to test capabilities believed to be beyond current language models and focuses on evaluating complex reasoning skills including temporal understanding, spatial reasoning, causal understanding, and deductive logical reasoning.", + "metric_id": "llm_stats.bbh.score", + "metric_name": "BBH score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bbh", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BBH", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.824, + "details": { + "raw_score": "0.824", + "raw_score_field": "score", + "raw_model_id": "qwen2-72b-instruct", + "raw_benchmark_id": "bbh", + "source_urls_json": "[\"https://llm-stats.com/models/qwen2-72b-instruct\",\"https://llm-stats.com/benchmarks/bbh\",\"https://api.llm-stats.com/leaderboard/benchmarks/bbh\"]", + "raw_score_id": "bbh::qwen2-72b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "c-eval::c-eval-qwen2-72b-instruct", + "evaluation_name": "llm_stats.c-eval", + "source_data": { + "dataset_name": "C-Eval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen2-72b-instruct", + "https://llm-stats.com/benchmarks/c-eval", + "https://api.llm-stats.com/leaderboard/benchmarks/c-eval" + ], + "additional_details": { + "raw_benchmark_id": "c-eval", + "raw_model_id": "qwen2-72b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "C-Eval is a comprehensive Chinese evaluation suite designed to assess advanced knowledge and reasoning abilities of foundation models in a Chinese context. It comprises 13,948 multiple-choice questions across 52 diverse disciplines spanning humanities, science, and engineering, with four difficulty levels: middle school, high school, college, and professional. The benchmark includes C-Eval Hard, a subset of very challenging subjects requiring advanced reasoning abilities.", + "metric_id": "llm_stats.c-eval.score", + "metric_name": "C-Eval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "c-eval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "C-Eval", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.838, + "details": { + "raw_score": "0.838", + "raw_score_field": "score", + "raw_model_id": "qwen2-72b-instruct", + "raw_benchmark_id": "c-eval", + "source_urls_json": "[\"https://llm-stats.com/models/qwen2-72b-instruct\",\"https://llm-stats.com/benchmarks/c-eval\",\"https://api.llm-stats.com/leaderboard/benchmarks/c-eval\"]", + "raw_score_id": "c-eval::qwen2-72b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "cmmlu::cmmlu-qwen2-72b-instruct", + "evaluation_name": "llm_stats.cmmlu", + "source_data": { + "dataset_name": "CMMLU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen2-72b-instruct", + "https://llm-stats.com/benchmarks/cmmlu", + "https://api.llm-stats.com/leaderboard/benchmarks/cmmlu" + ], + "additional_details": { + "raw_benchmark_id": "cmmlu", + "raw_model_id": "qwen2-72b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CMMLU (Chinese Massive Multitask Language Understanding) is a comprehensive Chinese benchmark that evaluates the knowledge and reasoning capabilities of large language models across 67 different subject topics. The benchmark covers natural sciences, social sciences, engineering, and humanities with multiple-choice questions ranging from basic to advanced professional levels.", + "metric_id": "llm_stats.cmmlu.score", + "metric_name": "CMMLU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "cmmlu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CMMLU", + "raw_categories": "[\"language\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.901, + "details": { + "raw_score": "0.901", + "raw_score_field": "score", + "raw_model_id": "qwen2-72b-instruct", + "raw_benchmark_id": "cmmlu", + "source_urls_json": "[\"https://llm-stats.com/models/qwen2-72b-instruct\",\"https://llm-stats.com/benchmarks/cmmlu\",\"https://api.llm-stats.com/leaderboard/benchmarks/cmmlu\"]", + "raw_score_id": "cmmlu::qwen2-72b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "evalplus::evalplus-qwen2-72b-instruct", + "evaluation_name": "llm_stats.evalplus", + "source_data": { + "dataset_name": "EvalPlus", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen2-72b-instruct", + "https://llm-stats.com/benchmarks/evalplus", + "https://api.llm-stats.com/leaderboard/benchmarks/evalplus" + ], + "additional_details": { + "raw_benchmark_id": "evalplus", + "raw_model_id": "qwen2-72b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A rigorous code synthesis evaluation framework that augments existing datasets with extensive test cases generated by LLM and mutation-based strategies to better assess functional correctness of generated code, including HumanEval+ with 80x more test cases", + "metric_id": "llm_stats.evalplus.score", + "metric_name": "EvalPlus score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "evalplus", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "EvalPlus", + "raw_categories": "[\"reasoning\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.79, + "details": { + "raw_score": "0.79", + "raw_score_field": "score", + "raw_model_id": "qwen2-72b-instruct", + "raw_benchmark_id": "evalplus", + "source_urls_json": "[\"https://llm-stats.com/models/qwen2-72b-instruct\",\"https://llm-stats.com/benchmarks/evalplus\",\"https://api.llm-stats.com/leaderboard/benchmarks/evalplus\"]", + "raw_score_id": "evalplus::qwen2-72b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hellaswag::hellaswag-qwen2-72b-instruct", + "evaluation_name": "llm_stats.hellaswag", + "source_data": { + "dataset_name": "HellaSwag", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen2-72b-instruct", + "https://llm-stats.com/benchmarks/hellaswag", + "https://api.llm-stats.com/leaderboard/benchmarks/hellaswag" + ], + "additional_details": { + "raw_benchmark_id": "hellaswag", + "raw_model_id": "qwen2-72b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A challenging commonsense natural language inference dataset that uses Adversarial Filtering to create questions trivial for humans (>95% accuracy) but difficult for state-of-the-art models, requiring completion of sentence endings based on physical situations and everyday activities", + "metric_id": "llm_stats.hellaswag.score", + "metric_name": "HellaSwag score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hellaswag", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HellaSwag", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.876, + "details": { + "raw_score": "0.876", + "raw_score_field": "score", + "raw_model_id": "qwen2-72b-instruct", + "raw_benchmark_id": "hellaswag", + "source_urls_json": "[\"https://llm-stats.com/models/qwen2-72b-instruct\",\"https://llm-stats.com/benchmarks/hellaswag\",\"https://api.llm-stats.com/leaderboard/benchmarks/hellaswag\"]", + "raw_score_id": "hellaswag::qwen2-72b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mbpp::mbpp-qwen2-72b-instruct", + "evaluation_name": "llm_stats.mbpp", + "source_data": { + "dataset_name": "MBPP", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen2-72b-instruct", + "https://llm-stats.com/benchmarks/mbpp", + "https://api.llm-stats.com/leaderboard/benchmarks/mbpp" + ], + "additional_details": { + "raw_benchmark_id": "mbpp", + "raw_model_id": "qwen2-72b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MBPP (Mostly Basic Python Problems) is a benchmark of 974 crowd-sourced Python programming problems designed to be solvable by entry-level programmers. Each problem consists of a task description, code solution, and 3 automated test cases covering programming fundamentals and standard library functionality.", + "metric_id": "llm_stats.mbpp.score", + "metric_name": "MBPP score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mbpp", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MBPP", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "33" + } + }, + "score_details": { + "score": 0.802, + "details": { + "raw_score": "0.802", + "raw_score_field": "score", + "raw_model_id": "qwen2-72b-instruct", + "raw_benchmark_id": "mbpp", + "source_urls_json": "[\"https://llm-stats.com/models/qwen2-72b-instruct\",\"https://llm-stats.com/benchmarks/mbpp\",\"https://api.llm-stats.com/leaderboard/benchmarks/mbpp\"]", + "raw_score_id": "mbpp::qwen2-72b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "multipl-e::multipl-e-qwen2-72b-instruct", + "evaluation_name": "llm_stats.multipl-e", + "source_data": { + "dataset_name": "MultiPL-E", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen2-72b-instruct", + "https://llm-stats.com/benchmarks/multipl-e", + "https://api.llm-stats.com/leaderboard/benchmarks/multipl-e" + ], + "additional_details": { + "raw_benchmark_id": "multipl-e", + "raw_model_id": "qwen2-72b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MultiPL-E is a scalable and extensible system for translating unit test-driven code generation benchmarks to multiple programming languages. It extends HumanEval and MBPP Python benchmarks to 18 additional programming languages, enabling evaluation of neural code generation models across diverse programming paradigms and language features.", + "metric_id": "llm_stats.multipl-e.score", + "metric_name": "MultiPL-E score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "multipl-e", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MultiPL-E", + "raw_categories": "[\"language\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.692, + "details": { + "raw_score": "0.692", + "raw_score_field": "score", + "raw_model_id": "qwen2-72b-instruct", + "raw_benchmark_id": "multipl-e", + "source_urls_json": "[\"https://llm-stats.com/models/qwen2-72b-instruct\",\"https://llm-stats.com/benchmarks/multipl-e\",\"https://api.llm-stats.com/leaderboard/benchmarks/multipl-e\"]", + "raw_score_id": "multipl-e::qwen2-72b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "theoremqa::theoremqa-qwen2-72b-instruct", + "evaluation_name": "llm_stats.theoremqa", + "source_data": { + "dataset_name": "TheoremQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen2-72b-instruct", + "https://llm-stats.com/benchmarks/theoremqa", + "https://api.llm-stats.com/leaderboard/benchmarks/theoremqa" + ], + "additional_details": { + "raw_benchmark_id": "theoremqa", + "raw_model_id": "qwen2-72b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A theorem-driven question answering dataset containing 800 high-quality questions covering 350+ theorems from Math, Physics, EE&CS, and Finance. Designed to evaluate AI models' capabilities to apply theorems to solve challenging university-level science problems.", + "metric_id": "llm_stats.theoremqa.score", + "metric_name": "TheoremQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "theoremqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TheoremQA", + "raw_categories": "[\"math\",\"physics\",\"reasoning\",\"finance\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.444, + "details": { + "raw_score": "0.444", + "raw_score_field": "score", + "raw_model_id": "qwen2-72b-instruct", + "raw_benchmark_id": "theoremqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen2-72b-instruct\",\"https://llm-stats.com/benchmarks/theoremqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/theoremqa\"]", + "raw_score_id": "theoremqa::qwen2-72b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "truthfulqa::truthfulqa-qwen2-72b-instruct", + "evaluation_name": "llm_stats.truthfulqa", + "source_data": { + "dataset_name": "TruthfulQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen2-72b-instruct", + "https://llm-stats.com/benchmarks/truthfulqa", + "https://api.llm-stats.com/leaderboard/benchmarks/truthfulqa" + ], + "additional_details": { + "raw_benchmark_id": "truthfulqa", + "raw_model_id": "qwen2-72b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "TruthfulQA is a benchmark to measure whether language models are truthful in generating answers to questions. It comprises 817 questions that span 38 categories, including health, law, finance and politics. The questions are crafted such that some humans would answer falsely due to a false belief or misconception, testing models' ability to avoid generating false answers learned from human texts.", + "metric_id": "llm_stats.truthfulqa.score", + "metric_name": "TruthfulQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "truthfulqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TruthfulQA", + "raw_categories": "[\"legal\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.548, + "details": { + "raw_score": "0.548", + "raw_score_field": "score", + "raw_model_id": "qwen2-72b-instruct", + "raw_benchmark_id": "truthfulqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen2-72b-instruct\",\"https://llm-stats.com/benchmarks/truthfulqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/truthfulqa\"]", + "raw_score_id": "truthfulqa::qwen2-72b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "winogrande::winogrande-qwen2-72b-instruct", + "evaluation_name": "llm_stats.winogrande", + "source_data": { + "dataset_name": "Winogrande", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen2-72b-instruct", + "https://llm-stats.com/benchmarks/winogrande", + "https://api.llm-stats.com/leaderboard/benchmarks/winogrande" + ], + "additional_details": { + "raw_benchmark_id": "winogrande", + "raw_model_id": "qwen2-72b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WinoGrande: An Adversarial Winograd Schema Challenge at Scale. A large-scale dataset of 44,000 pronoun resolution problems designed to test machine commonsense reasoning. Uses adversarial filtering to reduce spurious biases and provides a more robust evaluation of whether AI systems truly understand commonsense or exploit statistical shortcuts. Current best AI methods achieve 59.4-79.1% accuracy, significantly below human performance of 94.0%.", + "metric_id": "llm_stats.winogrande.score", + "metric_name": "Winogrande score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "winogrande", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Winogrande", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "22" + } + }, + "score_details": { + "score": 0.851, + "details": { + "raw_score": "0.851", + "raw_score_field": "score", + "raw_model_id": "qwen2-72b-instruct", + "raw_benchmark_id": "winogrande", + "source_urls_json": "[\"https://llm-stats.com/models/qwen2-72b-instruct\",\"https://llm-stats.com/benchmarks/winogrande\",\"https://api.llm-stats.com/leaderboard/benchmarks/winogrande\"]", + "raw_score_id": "winogrande::qwen2-72b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/qwen/qwen2-7b-instruct/d9df1549-5457-42ab-8aad-d8389c06b94a.json b/data/llm_stats/qwen/qwen2-7b-instruct/d9df1549-5457-42ab-8aad-d8389c06b94a.json new file mode 100644 index 0000000000000000000000000000000000000000..8bb3ab4c0651a44f9cba90844a908f45598e5b44 --- /dev/null +++ b/data/llm_stats/qwen/qwen2-7b-instruct/d9df1549-5457-42ab-8aad-d8389c06b94a.json @@ -0,0 +1,359 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/qwen_qwen2-7b-instruct/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Qwen2 7B Instruct", + "id": "qwen/qwen2-7b-instruct", + "developer": "qwen", + "additional_details": { + "raw_id": "qwen2-7b-instruct", + "raw_name": "Qwen2 7B Instruct", + "raw_model_id": "qwen2-7b-instruct", + "raw_model_name": "Qwen2 7B Instruct", + "raw_organization_id": "qwen", + "raw_organization_name": "Alibaba Cloud / Qwen Team", + "raw_release_date": "2024-07-23", + "raw_announcement_date": "2024-07-23", + "raw_multimodal": "false", + "raw_param_count": "7620000000", + "raw_provider_slug": "qwen", + "raw_provider_name": "Alibaba Cloud / Qwen Team" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "alignbench::alignbench-qwen2-7b-instruct", + "evaluation_name": "llm_stats.alignbench", + "source_data": { + "dataset_name": "AlignBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen2-7b-instruct", + "https://llm-stats.com/benchmarks/alignbench", + "https://api.llm-stats.com/leaderboard/benchmarks/alignbench" + ], + "additional_details": { + "raw_benchmark_id": "alignbench", + "raw_model_id": "qwen2-7b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "AlignBench is a comprehensive multi-dimensional benchmark for evaluating Chinese alignment of Large Language Models. It contains 8 main categories: Fundamental Language Ability, Advanced Chinese Understanding, Open-ended Questions, Writing Ability, Logical Reasoning, Mathematics, Task-oriented Role Play, and Professional Knowledge. The benchmark includes 683 real-scenario rooted queries with human-verified references and uses a rule-calibrated multi-dimensional LLM-as-Judge approach with Chain-of-Thought for evaluation.", + "metric_id": "llm_stats.alignbench.score", + "metric_name": "AlignBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "alignbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AlignBench", + "raw_categories": "[\"language\",\"math\",\"reasoning\",\"roleplay\",\"general\",\"creativity\",\"writing\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.721, + "details": { + "raw_score": "0.721", + "raw_score_field": "score", + "raw_model_id": "qwen2-7b-instruct", + "raw_benchmark_id": "alignbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen2-7b-instruct\",\"https://llm-stats.com/benchmarks/alignbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/alignbench\"]", + "raw_score_id": "alignbench::qwen2-7b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "c-eval::c-eval-qwen2-7b-instruct", + "evaluation_name": "llm_stats.c-eval", + "source_data": { + "dataset_name": "C-Eval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen2-7b-instruct", + "https://llm-stats.com/benchmarks/c-eval", + "https://api.llm-stats.com/leaderboard/benchmarks/c-eval" + ], + "additional_details": { + "raw_benchmark_id": "c-eval", + "raw_model_id": "qwen2-7b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "C-Eval is a comprehensive Chinese evaluation suite designed to assess advanced knowledge and reasoning abilities of foundation models in a Chinese context. It comprises 13,948 multiple-choice questions across 52 diverse disciplines spanning humanities, science, and engineering, with four difficulty levels: middle school, high school, college, and professional. The benchmark includes C-Eval Hard, a subset of very challenging subjects requiring advanced reasoning abilities.", + "metric_id": "llm_stats.c-eval.score", + "metric_name": "C-Eval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "c-eval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "C-Eval", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.772, + "details": { + "raw_score": "0.772", + "raw_score_field": "score", + "raw_model_id": "qwen2-7b-instruct", + "raw_benchmark_id": "c-eval", + "source_urls_json": "[\"https://llm-stats.com/models/qwen2-7b-instruct\",\"https://llm-stats.com/benchmarks/c-eval\",\"https://api.llm-stats.com/leaderboard/benchmarks/c-eval\"]", + "raw_score_id": "c-eval::qwen2-7b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "evalplus::evalplus-qwen2-7b-instruct", + "evaluation_name": "llm_stats.evalplus", + "source_data": { + "dataset_name": "EvalPlus", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen2-7b-instruct", + "https://llm-stats.com/benchmarks/evalplus", + "https://api.llm-stats.com/leaderboard/benchmarks/evalplus" + ], + "additional_details": { + "raw_benchmark_id": "evalplus", + "raw_model_id": "qwen2-7b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A rigorous code synthesis evaluation framework that augments existing datasets with extensive test cases generated by LLM and mutation-based strategies to better assess functional correctness of generated code, including HumanEval+ with 80x more test cases", + "metric_id": "llm_stats.evalplus.score", + "metric_name": "EvalPlus score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "evalplus", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "EvalPlus", + "raw_categories": "[\"reasoning\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.703, + "details": { + "raw_score": "0.703", + "raw_score_field": "score", + "raw_model_id": "qwen2-7b-instruct", + "raw_benchmark_id": "evalplus", + "source_urls_json": "[\"https://llm-stats.com/models/qwen2-7b-instruct\",\"https://llm-stats.com/benchmarks/evalplus\",\"https://api.llm-stats.com/leaderboard/benchmarks/evalplus\"]", + "raw_score_id": "evalplus::qwen2-7b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mt-bench::mt-bench-qwen2-7b-instruct", + "evaluation_name": "llm_stats.mt-bench", + "source_data": { + "dataset_name": "MT-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen2-7b-instruct", + "https://llm-stats.com/benchmarks/mt-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/mt-bench" + ], + "additional_details": { + "raw_benchmark_id": "mt-bench", + "raw_model_id": "qwen2-7b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MT-Bench is a challenging multi-turn benchmark that measures the ability of large language models to engage in coherent, informative, and engaging conversations. It uses strong LLMs as judges for scalable and explainable evaluation of multi-turn dialogue capabilities.", + "metric_id": "llm_stats.mt-bench.score", + "metric_name": "MT-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mt-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MT-Bench", + "raw_categories": "[\"reasoning\",\"roleplay\",\"general\",\"communication\",\"creativity\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.841, + "details": { + "raw_score": "0.841", + "raw_score_field": "score", + "raw_model_id": "qwen2-7b-instruct", + "raw_benchmark_id": "mt-bench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen2-7b-instruct\",\"https://llm-stats.com/benchmarks/mt-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/mt-bench\"]", + "raw_score_id": "mt-bench::qwen2-7b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "multipl-e::multipl-e-qwen2-7b-instruct", + "evaluation_name": "llm_stats.multipl-e", + "source_data": { + "dataset_name": "MultiPL-E", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen2-7b-instruct", + "https://llm-stats.com/benchmarks/multipl-e", + "https://api.llm-stats.com/leaderboard/benchmarks/multipl-e" + ], + "additional_details": { + "raw_benchmark_id": "multipl-e", + "raw_model_id": "qwen2-7b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MultiPL-E is a scalable and extensible system for translating unit test-driven code generation benchmarks to multiple programming languages. It extends HumanEval and MBPP Python benchmarks to 18 additional programming languages, enabling evaluation of neural code generation models across diverse programming paradigms and language features.", + "metric_id": "llm_stats.multipl-e.score", + "metric_name": "MultiPL-E score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "multipl-e", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MultiPL-E", + "raw_categories": "[\"language\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.591, + "details": { + "raw_score": "0.591", + "raw_score_field": "score", + "raw_model_id": "qwen2-7b-instruct", + "raw_benchmark_id": "multipl-e", + "source_urls_json": "[\"https://llm-stats.com/models/qwen2-7b-instruct\",\"https://llm-stats.com/benchmarks/multipl-e\",\"https://api.llm-stats.com/leaderboard/benchmarks/multipl-e\"]", + "raw_score_id": "multipl-e::qwen2-7b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "theoremqa::theoremqa-qwen2-7b-instruct", + "evaluation_name": "llm_stats.theoremqa", + "source_data": { + "dataset_name": "TheoremQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen2-7b-instruct", + "https://llm-stats.com/benchmarks/theoremqa", + "https://api.llm-stats.com/leaderboard/benchmarks/theoremqa" + ], + "additional_details": { + "raw_benchmark_id": "theoremqa", + "raw_model_id": "qwen2-7b-instruct", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A theorem-driven question answering dataset containing 800 high-quality questions covering 350+ theorems from Math, Physics, EE&CS, and Finance. Designed to evaluate AI models' capabilities to apply theorems to solve challenging university-level science problems.", + "metric_id": "llm_stats.theoremqa.score", + "metric_name": "TheoremQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "theoremqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TheoremQA", + "raw_categories": "[\"math\",\"physics\",\"reasoning\",\"finance\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.253, + "details": { + "raw_score": "0.253", + "raw_score_field": "score", + "raw_model_id": "qwen2-7b-instruct", + "raw_benchmark_id": "theoremqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen2-7b-instruct\",\"https://llm-stats.com/benchmarks/theoremqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/theoremqa\"]", + "raw_score_id": "theoremqa::qwen2-7b-instruct", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/qwen/qwen3-235b-a22b/baf10e13-e1b8-47a0-9355-ffa8f11b7271.json b/data/llm_stats/qwen/qwen3-235b-a22b/baf10e13-e1b8-47a0-9355-ffa8f11b7271.json new file mode 100644 index 0000000000000000000000000000000000000000..f635755b5762553ba6e3d15524e3cd39c075b119 --- /dev/null +++ b/data/llm_stats/qwen/qwen3-235b-a22b/baf10e13-e1b8-47a0-9355-ffa8f11b7271.json @@ -0,0 +1,827 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/qwen_qwen3-235b-a22b/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Qwen3 235B A22B", + "id": "qwen/qwen3-235b-a22b", + "developer": "qwen", + "additional_details": { + "raw_id": "qwen3-235b-a22b", + "raw_name": "Qwen3 235B A22B", + "raw_model_id": "qwen3-235b-a22b", + "raw_model_name": "Qwen3 235B A22B", + "raw_organization_id": "qwen", + "raw_organization_name": "Alibaba Cloud / Qwen Team", + "raw_release_date": "2025-04-29", + "raw_announcement_date": "2025-04-29", + "raw_multimodal": "false", + "raw_param_count": "235000000000", + "raw_provider_slug": "qwen", + "raw_provider_name": "Alibaba Cloud / Qwen Team" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aider::aider-qwen3-235b-a22b", + "evaluation_name": "llm_stats.aider", + "source_data": { + "dataset_name": "Aider", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3-235b-a22b", + "https://llm-stats.com/benchmarks/aider", + "https://api.llm-stats.com/leaderboard/benchmarks/aider" + ], + "additional_details": { + "raw_benchmark_id": "aider", + "raw_model_id": "qwen3-235b-a22b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Aider is a comprehensive code editing benchmark based on 133 practice exercises from Exercism's Python repository, designed to evaluate AI models' ability to translate natural language coding requests into executable code that passes unit tests. The benchmark measures end-to-end code editing capabilities, including GPT's ability to edit existing code and format code changes for automated saving to local files. The Aider Polyglot variant extends this evaluation across 225 challenging exercises spanning C++, Go, Java, JavaScript, Python, and Rust, making it a standard benchmark for assessing multilingual code editing performance in AI research.", + "metric_id": "llm_stats.aider.score", + "metric_name": "Aider score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aider", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Aider", + "raw_categories": "[\"reasoning\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.618, + "details": { + "raw_score": "0.618", + "raw_score_field": "score", + "raw_model_id": "qwen3-235b-a22b", + "raw_benchmark_id": "aider", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3-235b-a22b\",\"https://llm-stats.com/benchmarks/aider\",\"https://api.llm-stats.com/leaderboard/benchmarks/aider\"]", + "raw_score_id": "aider::qwen3-235b-a22b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "aime-2024::aime-2024-qwen3-235b-a22b", + "evaluation_name": "llm_stats.aime-2024", + "source_data": { + "dataset_name": "AIME 2024", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3-235b-a22b", + "https://llm-stats.com/benchmarks/aime-2024", + "https://api.llm-stats.com/leaderboard/benchmarks/aime-2024" + ], + "additional_details": { + "raw_benchmark_id": "aime-2024", + "raw_model_id": "qwen3-235b-a22b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "American Invitational Mathematics Examination 2024, consisting of 30 challenging mathematical reasoning problems from AIME I and AIME II competitions. Each problem requires an integer answer between 0-999 and tests advanced mathematical reasoning across algebra, geometry, combinatorics, and number theory. Used as a benchmark for evaluating mathematical reasoning capabilities in large language models at Olympiad-level difficulty.", + "metric_id": "llm_stats.aime-2024.score", + "metric_name": "AIME 2024 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aime-2024", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AIME 2024", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "53" + } + }, + "score_details": { + "score": 0.857, + "details": { + "raw_score": "0.857", + "raw_score_field": "score", + "raw_model_id": "qwen3-235b-a22b", + "raw_benchmark_id": "aime-2024", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3-235b-a22b\",\"https://llm-stats.com/benchmarks/aime-2024\",\"https://api.llm-stats.com/leaderboard/benchmarks/aime-2024\"]", + "raw_score_id": "aime-2024::qwen3-235b-a22b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "arena-hard::arena-hard-qwen3-235b-a22b", + "evaluation_name": "llm_stats.arena-hard", + "source_data": { + "dataset_name": "Arena Hard", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3-235b-a22b", + "https://llm-stats.com/benchmarks/arena-hard", + "https://api.llm-stats.com/leaderboard/benchmarks/arena-hard" + ], + "additional_details": { + "raw_benchmark_id": "arena-hard", + "raw_model_id": "qwen3-235b-a22b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Arena-Hard-Auto is an automatic evaluation benchmark for instruction-tuned LLMs consisting of 500 challenging real-world prompts curated by BenchBuilder. It includes open-ended software engineering problems, mathematical questions, and creative writing tasks. The benchmark uses LLM-as-a-Judge methodology with GPT-4.1 and Gemini-2.5 as automatic judges to approximate human preference. Arena-Hard achieves 98.6% correlation with human preference rankings and provides 3x higher separation of model performances compared to MT-Bench, making it highly effective for distinguishing between models of similar quality.", + "metric_id": "llm_stats.arena-hard.score", + "metric_name": "Arena Hard score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arena-hard", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Arena Hard", + "raw_categories": "[\"reasoning\",\"general\",\"creativity\",\"writing\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "26" + } + }, + "score_details": { + "score": 0.956, + "details": { + "raw_score": "0.956", + "raw_score_field": "score", + "raw_model_id": "qwen3-235b-a22b", + "raw_benchmark_id": "arena-hard", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3-235b-a22b\",\"https://llm-stats.com/benchmarks/arena-hard\",\"https://api.llm-stats.com/leaderboard/benchmarks/arena-hard\"]", + "raw_score_id": "arena-hard::qwen3-235b-a22b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bbh::bbh-qwen3-235b-a22b", + "evaluation_name": "llm_stats.bbh", + "source_data": { + "dataset_name": "BBH", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3-235b-a22b", + "https://llm-stats.com/benchmarks/bbh", + "https://api.llm-stats.com/leaderboard/benchmarks/bbh" + ], + "additional_details": { + "raw_benchmark_id": "bbh", + "raw_model_id": "qwen3-235b-a22b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Big-Bench Hard (BBH) is a suite of 23 challenging tasks selected from BIG-Bench for which prior language model evaluations did not outperform the average human-rater. These tasks require multi-step reasoning across diverse domains including arithmetic, logical reasoning, reading comprehension, and commonsense reasoning. The benchmark was designed to test capabilities believed to be beyond current language models and focuses on evaluating complex reasoning skills including temporal understanding, spatial reasoning, causal understanding, and deductive logical reasoning.", + "metric_id": "llm_stats.bbh.score", + "metric_name": "BBH score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bbh", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BBH", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.8887, + "details": { + "raw_score": "0.8887", + "raw_score_field": "score", + "raw_model_id": "qwen3-235b-a22b", + "raw_benchmark_id": "bbh", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3-235b-a22b\",\"https://llm-stats.com/benchmarks/bbh\",\"https://api.llm-stats.com/leaderboard/benchmarks/bbh\"]", + "raw_score_id": "bbh::qwen3-235b-a22b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bfcl::bfcl-qwen3-235b-a22b", + "evaluation_name": "llm_stats.bfcl", + "source_data": { + "dataset_name": "BFCL", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3-235b-a22b", + "https://llm-stats.com/benchmarks/bfcl", + "https://api.llm-stats.com/leaderboard/benchmarks/bfcl" + ], + "additional_details": { + "raw_benchmark_id": "bfcl", + "raw_model_id": "qwen3-235b-a22b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The Berkeley Function Calling Leaderboard (BFCL) is the first comprehensive and executable function call evaluation dedicated to assessing Large Language Models' ability to invoke functions. It evaluates serial and parallel function calls across multiple programming languages (Python, Java, JavaScript, REST API) using a novel Abstract Syntax Tree (AST) evaluation method. The benchmark consists of over 2,000 question-function-answer pairs covering diverse application domains and complex use cases including multiple function calls, parallel function calls, and multi-turn interactions.", + "metric_id": "llm_stats.bfcl.score", + "metric_name": "BFCL score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bfcl", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BFCL", + "raw_categories": "[\"reasoning\",\"general\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.708, + "details": { + "raw_score": "0.708", + "raw_score_field": "score", + "raw_model_id": "qwen3-235b-a22b", + "raw_benchmark_id": "bfcl", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3-235b-a22b\",\"https://llm-stats.com/benchmarks/bfcl\",\"https://api.llm-stats.com/leaderboard/benchmarks/bfcl\"]", + "raw_score_id": "bfcl::qwen3-235b-a22b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "crux-o::crux-o-qwen3-235b-a22b", + "evaluation_name": "llm_stats.crux-o", + "source_data": { + "dataset_name": "CRUX-O", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3-235b-a22b", + "https://llm-stats.com/benchmarks/crux-o", + "https://api.llm-stats.com/leaderboard/benchmarks/crux-o" + ], + "additional_details": { + "raw_benchmark_id": "crux-o", + "raw_model_id": "qwen3-235b-a22b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CRUXEval-O (output prediction) is part of the CRUXEval benchmark consisting of 800 Python functions (3-13 lines) designed to evaluate AI models' capabilities in code reasoning, understanding, and execution. The benchmark tests models' ability to predict correct function outputs given function code and inputs, focusing on short problems that a good human programmer should be able to solve in a minute.", + "metric_id": "llm_stats.crux-o.score", + "metric_name": "CRUX-O score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "crux-o", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CRUX-O", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.79, + "details": { + "raw_score": "0.79", + "raw_score_field": "score", + "raw_model_id": "qwen3-235b-a22b", + "raw_benchmark_id": "crux-o", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3-235b-a22b\",\"https://llm-stats.com/benchmarks/crux-o\",\"https://api.llm-stats.com/leaderboard/benchmarks/crux-o\"]", + "raw_score_id": "crux-o::qwen3-235b-a22b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "evalplus::evalplus-qwen3-235b-a22b", + "evaluation_name": "llm_stats.evalplus", + "source_data": { + "dataset_name": "EvalPlus", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3-235b-a22b", + "https://llm-stats.com/benchmarks/evalplus", + "https://api.llm-stats.com/leaderboard/benchmarks/evalplus" + ], + "additional_details": { + "raw_benchmark_id": "evalplus", + "raw_model_id": "qwen3-235b-a22b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A rigorous code synthesis evaluation framework that augments existing datasets with extensive test cases generated by LLM and mutation-based strategies to better assess functional correctness of generated code, including HumanEval+ with 80x more test cases", + "metric_id": "llm_stats.evalplus.score", + "metric_name": "EvalPlus score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "evalplus", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "EvalPlus", + "raw_categories": "[\"reasoning\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.776, + "details": { + "raw_score": "0.776", + "raw_score_field": "score", + "raw_model_id": "qwen3-235b-a22b", + "raw_benchmark_id": "evalplus", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3-235b-a22b\",\"https://llm-stats.com/benchmarks/evalplus\",\"https://api.llm-stats.com/leaderboard/benchmarks/evalplus\"]", + "raw_score_id": "evalplus::qwen3-235b-a22b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gsm8k::gsm8k-qwen3-235b-a22b", + "evaluation_name": "llm_stats.gsm8k", + "source_data": { + "dataset_name": "GSM8k", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3-235b-a22b", + "https://llm-stats.com/benchmarks/gsm8k", + "https://api.llm-stats.com/leaderboard/benchmarks/gsm8k" + ], + "additional_details": { + "raw_benchmark_id": "gsm8k", + "raw_model_id": "qwen3-235b-a22b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Grade School Math 8K, a dataset of 8.5K high-quality linguistically diverse grade school math word problems requiring multi-step reasoning and elementary arithmetic operations.", + "metric_id": "llm_stats.gsm8k.score", + "metric_name": "GSM8k score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gsm8k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GSM8k", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "48" + } + }, + "score_details": { + "score": 0.9439, + "details": { + "raw_score": "0.9439", + "raw_score_field": "score", + "raw_model_id": "qwen3-235b-a22b", + "raw_benchmark_id": "gsm8k", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3-235b-a22b\",\"https://llm-stats.com/benchmarks/gsm8k\",\"https://api.llm-stats.com/leaderboard/benchmarks/gsm8k\"]", + "raw_score_id": "gsm8k::qwen3-235b-a22b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "include::include-qwen3-235b-a22b", + "evaluation_name": "llm_stats.include", + "source_data": { + "dataset_name": "Include", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3-235b-a22b", + "https://llm-stats.com/benchmarks/include", + "https://api.llm-stats.com/leaderboard/benchmarks/include" + ], + "additional_details": { + "raw_benchmark_id": "include", + "raw_model_id": "qwen3-235b-a22b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Include benchmark - specific documentation not found in official sources", + "metric_id": "llm_stats.include.score", + "metric_name": "Include score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "include", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Include", + "raw_categories": "[\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "31" + } + }, + "score_details": { + "score": 0.7346, + "details": { + "raw_score": "0.7346", + "raw_score_field": "score", + "raw_model_id": "qwen3-235b-a22b", + "raw_benchmark_id": "include", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3-235b-a22b\",\"https://llm-stats.com/benchmarks/include\",\"https://api.llm-stats.com/leaderboard/benchmarks/include\"]", + "raw_score_id": "include::qwen3-235b-a22b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "livebench::livebench-qwen3-235b-a22b", + "evaluation_name": "llm_stats.livebench", + "source_data": { + "dataset_name": "LiveBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3-235b-a22b", + "https://llm-stats.com/benchmarks/livebench", + "https://api.llm-stats.com/leaderboard/benchmarks/livebench" + ], + "additional_details": { + "raw_benchmark_id": "livebench", + "raw_model_id": "qwen3-235b-a22b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LiveBench is a challenging, contamination-limited LLM benchmark that addresses test set contamination by releasing new questions monthly based on recently-released datasets, arXiv papers, news articles, and IMDb movie synopses. It comprises tasks across math, coding, reasoning, language, instruction following, and data analysis with verifiable, objective ground-truth answers.", + "metric_id": "llm_stats.livebench.score", + "metric_name": "LiveBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "livebench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LiveBench", + "raw_categories": "[\"math\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "38" + } + }, + "score_details": { + "score": 0.771, + "details": { + "raw_score": "0.771", + "raw_score_field": "score", + "raw_model_id": "qwen3-235b-a22b", + "raw_benchmark_id": "livebench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3-235b-a22b\",\"https://llm-stats.com/benchmarks/livebench\",\"https://api.llm-stats.com/leaderboard/benchmarks/livebench\"]", + "raw_score_id": "livebench::qwen3-235b-a22b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mbpp::mbpp-qwen3-235b-a22b", + "evaluation_name": "llm_stats.mbpp", + "source_data": { + "dataset_name": "MBPP", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3-235b-a22b", + "https://llm-stats.com/benchmarks/mbpp", + "https://api.llm-stats.com/leaderboard/benchmarks/mbpp" + ], + "additional_details": { + "raw_benchmark_id": "mbpp", + "raw_model_id": "qwen3-235b-a22b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MBPP (Mostly Basic Python Problems) is a benchmark of 974 crowd-sourced Python programming problems designed to be solvable by entry-level programmers. Each problem consists of a task description, code solution, and 3 automated test cases covering programming fundamentals and standard library functionality.", + "metric_id": "llm_stats.mbpp.score", + "metric_name": "MBPP score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mbpp", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MBPP", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "33" + } + }, + "score_details": { + "score": 0.814, + "details": { + "raw_score": "0.814", + "raw_score_field": "score", + "raw_model_id": "qwen3-235b-a22b", + "raw_benchmark_id": "mbpp", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3-235b-a22b\",\"https://llm-stats.com/benchmarks/mbpp\",\"https://api.llm-stats.com/leaderboard/benchmarks/mbpp\"]", + "raw_score_id": "mbpp::qwen3-235b-a22b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mgsm::mgsm-qwen3-235b-a22b", + "evaluation_name": "llm_stats.mgsm", + "source_data": { + "dataset_name": "MGSM", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3-235b-a22b", + "https://llm-stats.com/benchmarks/mgsm", + "https://api.llm-stats.com/leaderboard/benchmarks/mgsm" + ], + "additional_details": { + "raw_benchmark_id": "mgsm", + "raw_model_id": "qwen3-235b-a22b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MGSM (Multilingual Grade School Math) is a benchmark of grade-school math problems. Contains 250 grade-school math problems manually translated from the GSM8K dataset into ten typologically diverse languages: Spanish, French, German, Russian, Chinese, Japanese, Thai, Swahili, Bengali, and Telugu. Evaluates multilingual mathematical reasoning capabilities.", + "metric_id": "llm_stats.mgsm.score", + "metric_name": "MGSM score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mgsm", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MGSM", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "31" + } + }, + "score_details": { + "score": 0.8353, + "details": { + "raw_score": "0.8353", + "raw_score_field": "score", + "raw_model_id": "qwen3-235b-a22b", + "raw_benchmark_id": "mgsm", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3-235b-a22b\",\"https://llm-stats.com/benchmarks/mgsm\",\"https://api.llm-stats.com/leaderboard/benchmarks/mgsm\"]", + "raw_score_id": "mgsm::qwen3-235b-a22b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu::mmlu-qwen3-235b-a22b", + "evaluation_name": "llm_stats.mmlu", + "source_data": { + "dataset_name": "MMLU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3-235b-a22b", + "https://llm-stats.com/benchmarks/mmlu", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu" + ], + "additional_details": { + "raw_benchmark_id": "mmlu", + "raw_model_id": "qwen3-235b-a22b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Massive Multitask Language Understanding benchmark testing knowledge across 57 diverse subjects including STEM, humanities, social sciences, and professional domains", + "metric_id": "llm_stats.mmlu.score", + "metric_name": "MMLU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU", + "raw_categories": "[\"language\",\"legal\",\"math\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "100" + } + }, + "score_details": { + "score": 0.8781, + "details": { + "raw_score": "0.8781", + "raw_score_field": "score", + "raw_model_id": "qwen3-235b-a22b", + "raw_benchmark_id": "mmlu", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3-235b-a22b\",\"https://llm-stats.com/benchmarks/mmlu\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu\"]", + "raw_score_id": "mmlu::qwen3-235b-a22b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "multilf::multilf-qwen3-235b-a22b", + "evaluation_name": "llm_stats.multilf", + "source_data": { + "dataset_name": "MultiLF", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3-235b-a22b", + "https://llm-stats.com/benchmarks/multilf", + "https://api.llm-stats.com/leaderboard/benchmarks/multilf" + ], + "additional_details": { + "raw_benchmark_id": "multilf", + "raw_model_id": "qwen3-235b-a22b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MultiLF benchmark", + "metric_id": "llm_stats.multilf.score", + "metric_name": "MultiLF score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "multilf", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MultiLF", + "raw_categories": "[\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.719, + "details": { + "raw_score": "0.719", + "raw_score_field": "score", + "raw_model_id": "qwen3-235b-a22b", + "raw_benchmark_id": "multilf", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3-235b-a22b\",\"https://llm-stats.com/benchmarks/multilf\",\"https://api.llm-stats.com/leaderboard/benchmarks/multilf\"]", + "raw_score_id": "multilf::qwen3-235b-a22b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "multipl-e::multipl-e-qwen3-235b-a22b", + "evaluation_name": "llm_stats.multipl-e", + "source_data": { + "dataset_name": "MultiPL-E", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3-235b-a22b", + "https://llm-stats.com/benchmarks/multipl-e", + "https://api.llm-stats.com/leaderboard/benchmarks/multipl-e" + ], + "additional_details": { + "raw_benchmark_id": "multipl-e", + "raw_model_id": "qwen3-235b-a22b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MultiPL-E is a scalable and extensible system for translating unit test-driven code generation benchmarks to multiple programming languages. It extends HumanEval and MBPP Python benchmarks to 18 additional programming languages, enabling evaluation of neural code generation models across diverse programming paradigms and language features.", + "metric_id": "llm_stats.multipl-e.score", + "metric_name": "MultiPL-E score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "multipl-e", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MultiPL-E", + "raw_categories": "[\"language\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.6594, + "details": { + "raw_score": "0.6594", + "raw_score_field": "score", + "raw_model_id": "qwen3-235b-a22b", + "raw_benchmark_id": "multipl-e", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3-235b-a22b\",\"https://llm-stats.com/benchmarks/multipl-e\",\"https://api.llm-stats.com/leaderboard/benchmarks/multipl-e\"]", + "raw_score_id": "multipl-e::qwen3-235b-a22b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/qwen/qwen3.5-0.8b/0dd73020-7b47-49aa-9e30-06e4d118de7b.json b/data/llm_stats/qwen/qwen3.5-0.8b/0dd73020-7b47-49aa-9e30-06e4d118de7b.json new file mode 100644 index 0000000000000000000000000000000000000000..cc7d367eda3c07c5f177d5ed3addbd56ad2606ae --- /dev/null +++ b/data/llm_stats/qwen/qwen3.5-0.8b/0dd73020-7b47-49aa-9e30-06e4d118de7b.json @@ -0,0 +1,411 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/qwen_qwen3.5-0.8b/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Qwen3.5-0.8B", + "id": "qwen/qwen3.5-0.8b", + "developer": "qwen", + "additional_details": { + "raw_id": "qwen3.5-0.8b", + "raw_name": "Qwen3.5-0.8B", + "raw_model_id": "qwen3.5-0.8b", + "raw_model_name": "Qwen3.5-0.8B", + "raw_organization_id": "qwen", + "raw_organization_name": "Alibaba Cloud / Qwen Team", + "raw_release_date": "2026-03-02", + "raw_announcement_date": "2026-03-02", + "raw_multimodal": "true", + "raw_param_count": "800000000", + "raw_provider_slug": "qwen", + "raw_provider_name": "Alibaba Cloud / Qwen Team" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aa-lcr::aa-lcr-qwen3.5-0.8b", + "evaluation_name": "llm_stats.aa-lcr", + "source_data": { + "dataset_name": "AA-LCR", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-0.8b", + "https://llm-stats.com/benchmarks/aa-lcr", + "https://api.llm-stats.com/leaderboard/benchmarks/aa-lcr" + ], + "additional_details": { + "raw_benchmark_id": "aa-lcr", + "raw_model_id": "qwen3.5-0.8b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Agent Arena Long Context Reasoning benchmark", + "metric_id": "llm_stats.aa-lcr.score", + "metric_name": "AA-LCR score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aa-lcr", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AA-LCR", + "raw_categories": "[\"long_context\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "14" + } + }, + "score_details": { + "score": 0.047, + "details": { + "raw_score": "0.047", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-0.8b", + "raw_benchmark_id": "aa-lcr", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-0.8b\",\"https://llm-stats.com/benchmarks/aa-lcr\",\"https://api.llm-stats.com/leaderboard/benchmarks/aa-lcr\"]", + "raw_score_id": "aa-lcr::qwen3.5-0.8b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bfcl-v4::bfcl-v4-qwen3.5-0.8b", + "evaluation_name": "llm_stats.bfcl-v4", + "source_data": { + "dataset_name": "BFCL-V4", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-0.8b", + "https://llm-stats.com/benchmarks/bfcl-v4", + "https://api.llm-stats.com/leaderboard/benchmarks/bfcl-v4" + ], + "additional_details": { + "raw_benchmark_id": "bfcl-v4", + "raw_model_id": "qwen3.5-0.8b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Berkeley Function Calling Leaderboard V4 (BFCL-V4) evaluates LLMs on their ability to accurately call functions and APIs, including simple, multiple, parallel, and nested function calls across diverse programming scenarios.", + "metric_id": "llm_stats.bfcl-v4.score", + "metric_name": "BFCL-V4 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bfcl-v4", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BFCL-V4", + "raw_categories": "[\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.253, + "details": { + "raw_score": "0.253", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-0.8b", + "raw_benchmark_id": "bfcl-v4", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-0.8b\",\"https://llm-stats.com/benchmarks/bfcl-v4\",\"https://api.llm-stats.com/leaderboard/benchmarks/bfcl-v4\"]", + "raw_score_id": "bfcl-v4::qwen3.5-0.8b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "c-eval::c-eval-qwen3.5-0.8b", + "evaluation_name": "llm_stats.c-eval", + "source_data": { + "dataset_name": "C-Eval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-0.8b", + "https://llm-stats.com/benchmarks/c-eval", + "https://api.llm-stats.com/leaderboard/benchmarks/c-eval" + ], + "additional_details": { + "raw_benchmark_id": "c-eval", + "raw_model_id": "qwen3.5-0.8b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "C-Eval is a comprehensive Chinese evaluation suite designed to assess advanced knowledge and reasoning abilities of foundation models in a Chinese context. It comprises 13,948 multiple-choice questions across 52 diverse disciplines spanning humanities, science, and engineering, with four difficulty levels: middle school, high school, college, and professional. The benchmark includes C-Eval Hard, a subset of very challenging subjects requiring advanced reasoning abilities.", + "metric_id": "llm_stats.c-eval.score", + "metric_name": "C-Eval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "c-eval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "C-Eval", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.505, + "details": { + "raw_score": "0.505", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-0.8b", + "raw_benchmark_id": "c-eval", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-0.8b\",\"https://llm-stats.com/benchmarks/c-eval\",\"https://api.llm-stats.com/leaderboard/benchmarks/c-eval\"]", + "raw_score_id": "c-eval::qwen3.5-0.8b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "global-piqa::global-piqa-qwen3.5-0.8b", + "evaluation_name": "llm_stats.global-piqa", + "source_data": { + "dataset_name": "Global PIQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-0.8b", + "https://llm-stats.com/benchmarks/global-piqa", + "https://api.llm-stats.com/leaderboard/benchmarks/global-piqa" + ], + "additional_details": { + "raw_benchmark_id": "global-piqa", + "raw_model_id": "qwen3.5-0.8b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Global PIQA is a multilingual commonsense reasoning benchmark that evaluates physical interaction knowledge across 100 languages and cultures. It tests AI systems' understanding of physical world knowledge in diverse cultural contexts through multiple choice questions about everyday situations requiring physical commonsense.", + "metric_id": "llm_stats.global-piqa.score", + "metric_name": "Global PIQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "global-piqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Global PIQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.594, + "details": { + "raw_score": "0.594", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-0.8b", + "raw_benchmark_id": "global-piqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-0.8b\",\"https://llm-stats.com/benchmarks/global-piqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/global-piqa\"]", + "raw_score_id": "global-piqa::qwen3.5-0.8b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "longbench-v2::longbench-v2-qwen3.5-0.8b", + "evaluation_name": "llm_stats.longbench-v2", + "source_data": { + "dataset_name": "LongBench v2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-0.8b", + "https://llm-stats.com/benchmarks/longbench-v2", + "https://api.llm-stats.com/leaderboard/benchmarks/longbench-v2" + ], + "additional_details": { + "raw_benchmark_id": "longbench-v2", + "raw_model_id": "qwen3.5-0.8b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LongBench v2 is a benchmark designed to assess the ability of LLMs to handle long-context problems requiring deep understanding and reasoning across real-world multitasks. It consists of 503 challenging multiple-choice questions with contexts ranging from 8k to 2M words across six major task categories: single-document QA, multi-document QA, long in-context learning, long-dialogue history understanding, code repository understanding, and long structured data understanding.", + "metric_id": "llm_stats.longbench-v2.score", + "metric_name": "LongBench v2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "longbench-v2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LongBench v2", + "raw_categories": "[\"long_context\",\"reasoning\",\"structured_output\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.261, + "details": { + "raw_score": "0.261", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-0.8b", + "raw_benchmark_id": "longbench-v2", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-0.8b\",\"https://llm-stats.com/benchmarks/longbench-v2\",\"https://api.llm-stats.com/leaderboard/benchmarks/longbench-v2\"]", + "raw_score_id": "longbench-v2::qwen3.5-0.8b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "maxife::maxife-qwen3.5-0.8b", + "evaluation_name": "llm_stats.maxife", + "source_data": { + "dataset_name": "MAXIFE", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-0.8b", + "https://llm-stats.com/benchmarks/maxife", + "https://api.llm-stats.com/leaderboard/benchmarks/maxife" + ], + "additional_details": { + "raw_benchmark_id": "maxife", + "raw_model_id": "qwen3.5-0.8b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MAXIFE is a multilingual benchmark evaluating LLMs on instruction following and execution across multiple languages and cultural contexts.", + "metric_id": "llm_stats.maxife.score", + "metric_name": "MAXIFE score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "maxife", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MAXIFE", + "raw_categories": "[\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.392, + "details": { + "raw_score": "0.392", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-0.8b", + "raw_benchmark_id": "maxife", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-0.8b\",\"https://llm-stats.com/benchmarks/maxife\",\"https://api.llm-stats.com/leaderboard/benchmarks/maxife\"]", + "raw_score_id": "maxife::qwen3.5-0.8b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "nova-63::nova-63-qwen3.5-0.8b", + "evaluation_name": "llm_stats.nova-63", + "source_data": { + "dataset_name": "NOVA-63", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-0.8b", + "https://llm-stats.com/benchmarks/nova-63", + "https://api.llm-stats.com/leaderboard/benchmarks/nova-63" + ], + "additional_details": { + "raw_benchmark_id": "nova-63", + "raw_model_id": "qwen3.5-0.8b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "NOVA-63 is a multilingual evaluation benchmark covering 63 languages, designed to assess LLM performance across diverse linguistic contexts and tasks.", + "metric_id": "llm_stats.nova-63.score", + "metric_name": "NOVA-63 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "nova-63", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "NOVA-63", + "raw_categories": "[\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.424, + "details": { + "raw_score": "0.424", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-0.8b", + "raw_benchmark_id": "nova-63", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-0.8b\",\"https://llm-stats.com/benchmarks/nova-63\",\"https://api.llm-stats.com/leaderboard/benchmarks/nova-63\"]", + "raw_score_id": "nova-63::qwen3.5-0.8b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/qwen/qwen3.5-122b-a10b/15981a89-0c1f-4e10-916a-a599f3c65b4b.json b/data/llm_stats/qwen/qwen3.5-122b-a10b/15981a89-0c1f-4e10-916a-a599f3c65b4b.json new file mode 100644 index 0000000000000000000000000000000000000000..d5958c36a16fc414ffd840c65559fccfafbbe9a9 --- /dev/null +++ b/data/llm_stats/qwen/qwen3.5-122b-a10b/15981a89-0c1f-4e10-916a-a599f3c65b4b.json @@ -0,0 +1,3843 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/qwen_qwen3.5-122b-a10b/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Qwen3.5-122B-A10B", + "id": "qwen/qwen3.5-122b-a10b", + "developer": "qwen", + "additional_details": { + "raw_id": "qwen3.5-122b-a10b", + "raw_name": "Qwen3.5-122B-A10B", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_model_name": "Qwen3.5-122B-A10B", + "raw_organization_id": "qwen", + "raw_organization_name": "Alibaba Cloud / Qwen Team", + "raw_release_date": "2026-02-24", + "raw_announcement_date": "2026-02-24", + "raw_multimodal": "true", + "raw_param_count": "122000000000", + "raw_provider_slug": "qwen", + "raw_provider_name": "Alibaba Cloud / Qwen Team" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aa-lcr::aa-lcr-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.aa-lcr", + "source_data": { + "dataset_name": "AA-LCR", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/aa-lcr", + "https://api.llm-stats.com/leaderboard/benchmarks/aa-lcr" + ], + "additional_details": { + "raw_benchmark_id": "aa-lcr", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Agent Arena Long Context Reasoning benchmark", + "metric_id": "llm_stats.aa-lcr.score", + "metric_name": "AA-LCR score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aa-lcr", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AA-LCR", + "raw_categories": "[\"long_context\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "14" + } + }, + "score_details": { + "score": 0.669, + "details": { + "raw_score": "0.669", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "aa-lcr", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/aa-lcr\",\"https://api.llm-stats.com/leaderboard/benchmarks/aa-lcr\"]", + "raw_score_id": "aa-lcr::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ai2d::ai2d-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.ai2d", + "source_data": { + "dataset_name": "AI2D", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/ai2d", + "https://api.llm-stats.com/leaderboard/benchmarks/ai2d" + ], + "additional_details": { + "raw_benchmark_id": "ai2d", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "AI2D is a dataset of 4,903 illustrative diagrams from grade school natural sciences (such as food webs, human physiology, and life cycles) with over 15,000 multiple choice questions and answers. The benchmark evaluates diagram understanding and visual reasoning capabilities, requiring models to interpret diagrammatic elements, relationships, and structure to answer questions about scientific concepts represented in visual form.", + "metric_id": "llm_stats.ai2d.score", + "metric_name": "AI2D score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ai2d", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AI2D", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "32" + } + }, + "score_details": { + "score": 0.933, + "details": { + "raw_score": "0.933", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "ai2d", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/ai2d\",\"https://api.llm-stats.com/leaderboard/benchmarks/ai2d\"]", + "raw_score_id": "ai2d::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "androidworld-sr::androidworld-sr-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.androidworld-sr", + "source_data": { + "dataset_name": "AndroidWorld_SR", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/androidworld-sr", + "https://api.llm-stats.com/leaderboard/benchmarks/androidworld-sr" + ], + "additional_details": { + "raw_benchmark_id": "androidworld-sr", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "AndroidWorld Success Rate (SR) benchmark - A dynamic benchmarking environment for autonomous agents operating on Android devices. Evaluates agents on 116 programmatic tasks across 20 real-world Android apps using multimodal inputs (screen screenshots, accessibility trees, and natural language instructions). Measures success rate of agents completing tasks like sending messages, creating calendar events, and navigating mobile interfaces. Published at ICLR 2025. Best current performance: 30.6% success rate (M3A agent) vs 80.0% human performance.", + "metric_id": "llm_stats.androidworld-sr.score", + "metric_name": "AndroidWorld_SR score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "androidworld-sr", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AndroidWorld_SR", + "raw_categories": "[\"multimodal\",\"reasoning\",\"general\",\"agents\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.664, + "details": { + "raw_score": "0.664", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "androidworld-sr", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/androidworld-sr\",\"https://api.llm-stats.com/leaderboard/benchmarks/androidworld-sr\"]", + "raw_score_id": "androidworld-sr::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "babyvision::babyvision-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.babyvision", + "source_data": { + "dataset_name": "BabyVision", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/babyvision", + "https://api.llm-stats.com/leaderboard/benchmarks/babyvision" + ], + "additional_details": { + "raw_benchmark_id": "babyvision", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A benchmark for early-stage visual reasoning and perception on child-like vision tasks.", + "metric_id": "llm_stats.babyvision.score", + "metric_name": "BabyVision score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "babyvision", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BabyVision", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.402, + "details": { + "raw_score": "0.402", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "babyvision", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/babyvision\",\"https://api.llm-stats.com/leaderboard/benchmarks/babyvision\"]", + "raw_score_id": "babyvision::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bfcl-v4::bfcl-v4-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.bfcl-v4", + "source_data": { + "dataset_name": "BFCL-V4", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/bfcl-v4", + "https://api.llm-stats.com/leaderboard/benchmarks/bfcl-v4" + ], + "additional_details": { + "raw_benchmark_id": "bfcl-v4", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Berkeley Function Calling Leaderboard V4 (BFCL-V4) evaluates LLMs on their ability to accurately call functions and APIs, including simple, multiple, parallel, and nested function calls across diverse programming scenarios.", + "metric_id": "llm_stats.bfcl-v4.score", + "metric_name": "BFCL-V4 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bfcl-v4", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BFCL-V4", + "raw_categories": "[\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.722, + "details": { + "raw_score": "0.722", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "bfcl-v4", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/bfcl-v4\",\"https://api.llm-stats.com/leaderboard/benchmarks/bfcl-v4\"]", + "raw_score_id": "bfcl-v4::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "browsecomp-zh::browsecomp-zh-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.browsecomp-zh", + "source_data": { + "dataset_name": "BrowseComp-zh", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/browsecomp-zh", + "https://api.llm-stats.com/leaderboard/benchmarks/browsecomp-zh" + ], + "additional_details": { + "raw_benchmark_id": "browsecomp-zh", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A high-difficulty benchmark purpose-built to comprehensively evaluate LLM agents on the Chinese web, consisting of 289 multi-hop questions spanning 11 diverse domains including Film & TV, Technology, Medicine, and History. Questions are reverse-engineered from short, objective, and easily verifiable answers, requiring sophisticated reasoning and information reconciliation beyond basic retrieval. The benchmark addresses linguistic, infrastructural, and censorship-related complexities in Chinese web environments.", + "metric_id": "llm_stats.browsecomp-zh.score", + "metric_name": "BrowseComp-zh score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "browsecomp-zh", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BrowseComp-zh", + "raw_categories": "[\"reasoning\",\"search\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.699, + "details": { + "raw_score": "0.699", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "browsecomp-zh", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/browsecomp-zh\",\"https://api.llm-stats.com/leaderboard/benchmarks/browsecomp-zh\"]", + "raw_score_id": "browsecomp-zh::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "c-eval::c-eval-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.c-eval", + "source_data": { + "dataset_name": "C-Eval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/c-eval", + "https://api.llm-stats.com/leaderboard/benchmarks/c-eval" + ], + "additional_details": { + "raw_benchmark_id": "c-eval", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "C-Eval is a comprehensive Chinese evaluation suite designed to assess advanced knowledge and reasoning abilities of foundation models in a Chinese context. It comprises 13,948 multiple-choice questions across 52 diverse disciplines spanning humanities, science, and engineering, with four difficulty levels: middle school, high school, college, and professional. The benchmark includes C-Eval Hard, a subset of very challenging subjects requiring advanced reasoning abilities.", + "metric_id": "llm_stats.c-eval.score", + "metric_name": "C-Eval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "c-eval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "C-Eval", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.919, + "details": { + "raw_score": "0.919", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "c-eval", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/c-eval\",\"https://api.llm-stats.com/leaderboard/benchmarks/c-eval\"]", + "raw_score_id": "c-eval::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "cc-ocr::cc-ocr-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.cc-ocr", + "source_data": { + "dataset_name": "CC-OCR", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/cc-ocr", + "https://api.llm-stats.com/leaderboard/benchmarks/cc-ocr" + ], + "additional_details": { + "raw_benchmark_id": "cc-ocr", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A comprehensive OCR benchmark for evaluating Large Multimodal Models (LMMs) in literacy. Comprises four OCR-centric tracks: multi-scene text reading, multilingual text reading, document parsing, and key information extraction. Contains 39 subsets with 7,058 fully annotated images, 41% sourced from real applications. Tests capabilities including text grounding, multi-orientation text recognition, and detecting hallucination/repetition across diverse visual challenges.", + "metric_id": "llm_stats.cc-ocr.score", + "metric_name": "CC-OCR score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "cc-ocr", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CC-OCR", + "raw_categories": "[\"multimodal\",\"structured_output\",\"text-to-image\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.818, + "details": { + "raw_score": "0.818", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "cc-ocr", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/cc-ocr\",\"https://api.llm-stats.com/leaderboard/benchmarks/cc-ocr\"]", + "raw_score_id": "cc-ocr::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "codeforces::codeforces-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.codeforces", + "source_data": { + "dataset_name": "CodeForces", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/codeforces", + "https://api.llm-stats.com/leaderboard/benchmarks/codeforces" + ], + "additional_details": { + "raw_benchmark_id": "codeforces", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A competitive programming benchmark using problems from the CodeForces platform. The benchmark evaluates code generation capabilities of LLMs on algorithmic problems with difficulty ratings ranging from 800 to 2400. Problems cover diverse algorithmic categories including dynamic programming, graph algorithms, data structures, and mathematical problems with standardized evaluation through direct platform submission.", + "metric_id": "llm_stats.codeforces.score", + "metric_name": "CodeForces score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "codeforces", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CodeForces", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.851, + "details": { + "raw_score": "0.851", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "codeforces", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/codeforces\",\"https://api.llm-stats.com/leaderboard/benchmarks/codeforces\"]", + "raw_score_id": "codeforces::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "countbench::countbench-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.countbench", + "source_data": { + "dataset_name": "CountBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/countbench", + "https://api.llm-stats.com/leaderboard/benchmarks/countbench" + ], + "additional_details": { + "raw_benchmark_id": "countbench", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CountBench evaluates object counting capabilities in visual understanding.", + "metric_id": "llm_stats.countbench.score", + "metric_name": "CountBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "countbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CountBench", + "raw_categories": "[\"reasoning\",\"spatial_reasoning\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.97, + "details": { + "raw_score": "0.97", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "countbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/countbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/countbench\"]", + "raw_score_id": "countbench::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "deep-planning::deep-planning-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.deep-planning", + "source_data": { + "dataset_name": "DeepPlanning", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/deep-planning", + "https://api.llm-stats.com/leaderboard/benchmarks/deep-planning" + ], + "additional_details": { + "raw_benchmark_id": "deep-planning", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "DeepPlanning evaluates LLMs on complex multi-step planning tasks requiring long-horizon reasoning, goal decomposition, and strategic decision-making.", + "metric_id": "llm_stats.deep-planning.score", + "metric_name": "DeepPlanning score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "deep-planning", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DeepPlanning", + "raw_categories": "[\"reasoning\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.241, + "details": { + "raw_score": "0.241", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "deep-planning", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/deep-planning\",\"https://api.llm-stats.com/leaderboard/benchmarks/deep-planning\"]", + "raw_score_id": "deep-planning::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "dynamath::dynamath-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.dynamath", + "source_data": { + "dataset_name": "DynaMath", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/dynamath", + "https://api.llm-stats.com/leaderboard/benchmarks/dynamath" + ], + "additional_details": { + "raw_benchmark_id": "dynamath", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multimodal mathematics and reasoning benchmark focused on dynamic visual problem solving.", + "metric_id": "llm_stats.dynamath.score", + "metric_name": "DynaMath score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "dynamath", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DynaMath", + "raw_categories": "[\"math\",\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.859, + "details": { + "raw_score": "0.859", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "dynamath", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/dynamath\",\"https://api.llm-stats.com/leaderboard/benchmarks/dynamath\"]", + "raw_score_id": "dynamath::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "embspatialbench::embspatialbench-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.embspatialbench", + "source_data": { + "dataset_name": "EmbSpatialBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/embspatialbench", + "https://api.llm-stats.com/leaderboard/benchmarks/embspatialbench" + ], + "additional_details": { + "raw_benchmark_id": "embspatialbench", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "EmbSpatialBench evaluates embodied spatial understanding and reasoning capabilities.", + "metric_id": "llm_stats.embspatialbench.score", + "metric_name": "EmbSpatialBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "embspatialbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "EmbSpatialBench", + "raw_categories": "[\"spatial_reasoning\",\"embodied\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.839, + "details": { + "raw_score": "0.839", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "embspatialbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/embspatialbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/embspatialbench\"]", + "raw_score_id": "embspatialbench::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "erqa::erqa-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.erqa", + "source_data": { + "dataset_name": "ERQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/erqa", + "https://api.llm-stats.com/leaderboard/benchmarks/erqa" + ], + "additional_details": { + "raw_benchmark_id": "erqa", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Embodied Reasoning Question Answering benchmark consisting of 400 multiple-choice visual questions across spatial reasoning, trajectory reasoning, action reasoning, state estimation, and multi-view reasoning for evaluating AI capabilities in physical world interactions", + "metric_id": "llm_stats.erqa.score", + "metric_name": "ERQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "erqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ERQA", + "raw_categories": "[\"reasoning\",\"spatial_reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "22" + } + }, + "score_details": { + "score": 0.62, + "details": { + "raw_score": "0.62", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "erqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/erqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/erqa\"]", + "raw_score_id": "erqa::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "fullstackbench-en::fullstackbench-en-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.fullstackbench-en", + "source_data": { + "dataset_name": "FullStackBench en", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/fullstackbench-en", + "https://api.llm-stats.com/leaderboard/benchmarks/fullstackbench-en" + ], + "additional_details": { + "raw_benchmark_id": "fullstackbench-en", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "English subset of FullStackBench for evaluating end-to-end software engineering and full-stack development capability.", + "metric_id": "llm_stats.fullstackbench-en.score", + "metric_name": "FullStackBench en score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "fullstackbench-en", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FullStackBench en", + "raw_categories": "[\"reasoning\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.626, + "details": { + "raw_score": "0.626", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "fullstackbench-en", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/fullstackbench-en\",\"https://api.llm-stats.com/leaderboard/benchmarks/fullstackbench-en\"]", + "raw_score_id": "fullstackbench-en::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "fullstackbench-zh::fullstackbench-zh-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.fullstackbench-zh", + "source_data": { + "dataset_name": "FullStackBench zh", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/fullstackbench-zh", + "https://api.llm-stats.com/leaderboard/benchmarks/fullstackbench-zh" + ], + "additional_details": { + "raw_benchmark_id": "fullstackbench-zh", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Chinese subset of FullStackBench for evaluating end-to-end software engineering and full-stack development capability.", + "metric_id": "llm_stats.fullstackbench-zh.score", + "metric_name": "FullStackBench zh score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "fullstackbench-zh", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FullStackBench zh", + "raw_categories": "[\"reasoning\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.587, + "details": { + "raw_score": "0.587", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "fullstackbench-zh", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/fullstackbench-zh\",\"https://api.llm-stats.com/leaderboard/benchmarks/fullstackbench-zh\"]", + "raw_score_id": "fullstackbench-zh::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "global-piqa::global-piqa-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.global-piqa", + "source_data": { + "dataset_name": "Global PIQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/global-piqa", + "https://api.llm-stats.com/leaderboard/benchmarks/global-piqa" + ], + "additional_details": { + "raw_benchmark_id": "global-piqa", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Global PIQA is a multilingual commonsense reasoning benchmark that evaluates physical interaction knowledge across 100 languages and cultures. It tests AI systems' understanding of physical world knowledge in diverse cultural contexts through multiple choice questions about everyday situations requiring physical commonsense.", + "metric_id": "llm_stats.global-piqa.score", + "metric_name": "Global PIQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "global-piqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Global PIQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.884, + "details": { + "raw_score": "0.884", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "global-piqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/global-piqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/global-piqa\"]", + "raw_score_id": "global-piqa::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hallusion-bench::hallusion-bench-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.hallusion-bench", + "source_data": { + "dataset_name": "Hallusion Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/hallusion-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/hallusion-bench" + ], + "additional_details": { + "raw_benchmark_id": "hallusion-bench", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A comprehensive benchmark designed to evaluate image-context reasoning in large visual-language models (LVLMs) by challenging models with 346 images and 1,129 carefully crafted questions to assess language hallucination and visual illusion", + "metric_id": "llm_stats.hallusion-bench.score", + "metric_name": "Hallusion Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hallusion-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Hallusion Bench", + "raw_categories": "[\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.676, + "details": { + "raw_score": "0.676", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "hallusion-bench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/hallusion-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/hallusion-bench\"]", + "raw_score_id": "hallusion-bench::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hmmt-2025::hmmt-2025-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.hmmt-2025", + "source_data": { + "dataset_name": "HMMT 2025", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/hmmt-2025", + "https://api.llm-stats.com/leaderboard/benchmarks/hmmt-2025" + ], + "additional_details": { + "raw_benchmark_id": "hmmt-2025", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Harvard-MIT Mathematics Tournament 2025 - A prestigious student-organized mathematics competition for high school students featuring two tournaments (November 2025 at MIT and February 2026 at Harvard) with individual tests, team rounds, and guts rounds", + "metric_id": "llm_stats.hmmt-2025.score", + "metric_name": "HMMT 2025 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hmmt-2025", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HMMT 2025", + "raw_categories": "[\"math\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "33" + } + }, + "score_details": { + "score": 0.914, + "details": { + "raw_score": "0.914", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "hmmt-2025", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/hmmt-2025\",\"https://api.llm-stats.com/leaderboard/benchmarks/hmmt-2025\"]", + "raw_score_id": "hmmt-2025::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hmmt25::hmmt25-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.hmmt25", + "source_data": { + "dataset_name": "HMMT25", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/hmmt25", + "https://api.llm-stats.com/leaderboard/benchmarks/hmmt25" + ], + "additional_details": { + "raw_benchmark_id": "hmmt25", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Harvard-MIT Mathematics Tournament 2025 - A prestigious student-organized mathematics competition for high school students featuring two tournaments (November 2025 at MIT and February 2026 at Harvard) with individual tests, team rounds, and guts rounds", + "metric_id": "llm_stats.hmmt25.score", + "metric_name": "HMMT25 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hmmt25", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HMMT25", + "raw_categories": "[\"math\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "25" + } + }, + "score_details": { + "score": 0.903, + "details": { + "raw_score": "0.903", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "hmmt25", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/hmmt25\",\"https://api.llm-stats.com/leaderboard/benchmarks/hmmt25\"]", + "raw_score_id": "hmmt25::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hypersim::hypersim-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.hypersim", + "source_data": { + "dataset_name": "Hypersim", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/hypersim", + "https://api.llm-stats.com/leaderboard/benchmarks/hypersim" + ], + "additional_details": { + "raw_benchmark_id": "hypersim", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Hypersim evaluates 3D grounding and depth understanding in synthetic indoor scenes.", + "metric_id": "llm_stats.hypersim.score", + "metric_name": "Hypersim score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hypersim", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Hypersim", + "raw_categories": "[\"spatial_reasoning\",\"3d\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.127, + "details": { + "raw_score": "0.127", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "hypersim", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/hypersim\",\"https://api.llm-stats.com/leaderboard/benchmarks/hypersim\"]", + "raw_score_id": "hypersim::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ifbench::ifbench-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.ifbench", + "source_data": { + "dataset_name": "IFBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/ifbench", + "https://api.llm-stats.com/leaderboard/benchmarks/ifbench" + ], + "additional_details": { + "raw_benchmark_id": "ifbench", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Instruction Following Benchmark evaluating model's ability to follow complex instructions", + "metric_id": "llm_stats.ifbench.score", + "metric_name": "IFBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ifbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IFBench", + "raw_categories": "[\"instruction_following\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.761, + "details": { + "raw_score": "0.761", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "ifbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/ifbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/ifbench\"]", + "raw_score_id": "ifbench::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ifeval::ifeval-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.ifeval", + "source_data": { + "dataset_name": "IFEval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/ifeval", + "https://api.llm-stats.com/leaderboard/benchmarks/ifeval" + ], + "additional_details": { + "raw_benchmark_id": "ifeval", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Instruction-Following Evaluation (IFEval) benchmark for large language models, focusing on verifiable instructions with 25 types of instructions and around 500 prompts containing one or more verifiable constraints", + "metric_id": "llm_stats.ifeval.score", + "metric_name": "IFEval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ifeval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IFEval", + "raw_categories": "[\"instruction_following\",\"structured_output\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "65" + } + }, + "score_details": { + "score": 0.934, + "details": { + "raw_score": "0.934", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "ifeval", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/ifeval\",\"https://api.llm-stats.com/leaderboard/benchmarks/ifeval\"]", + "raw_score_id": "ifeval::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "include::include-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.include", + "source_data": { + "dataset_name": "Include", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/include", + "https://api.llm-stats.com/leaderboard/benchmarks/include" + ], + "additional_details": { + "raw_benchmark_id": "include", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Include benchmark - specific documentation not found in official sources", + "metric_id": "llm_stats.include.score", + "metric_name": "Include score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "include", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Include", + "raw_categories": "[\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "31" + } + }, + "score_details": { + "score": 0.828, + "details": { + "raw_score": "0.828", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "include", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/include\",\"https://api.llm-stats.com/leaderboard/benchmarks/include\"]", + "raw_score_id": "include::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "lingoqa::lingoqa-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.lingoqa", + "source_data": { + "dataset_name": "LingoQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/lingoqa", + "https://api.llm-stats.com/leaderboard/benchmarks/lingoqa" + ], + "additional_details": { + "raw_benchmark_id": "lingoqa", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A benchmark for multimodal spatial-language understanding and visual-linguistic question answering.", + "metric_id": "llm_stats.lingoqa.score", + "metric_name": "LingoQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "lingoqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LingoQA", + "raw_categories": "[\"language\",\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.808, + "details": { + "raw_score": "0.808", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "lingoqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/lingoqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/lingoqa\"]", + "raw_score_id": "lingoqa::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "longbench-v2::longbench-v2-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.longbench-v2", + "source_data": { + "dataset_name": "LongBench v2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/longbench-v2", + "https://api.llm-stats.com/leaderboard/benchmarks/longbench-v2" + ], + "additional_details": { + "raw_benchmark_id": "longbench-v2", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LongBench v2 is a benchmark designed to assess the ability of LLMs to handle long-context problems requiring deep understanding and reasoning across real-world multitasks. It consists of 503 challenging multiple-choice questions with contexts ranging from 8k to 2M words across six major task categories: single-document QA, multi-document QA, long in-context learning, long-dialogue history understanding, code repository understanding, and long structured data understanding.", + "metric_id": "llm_stats.longbench-v2.score", + "metric_name": "LongBench v2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "longbench-v2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LongBench v2", + "raw_categories": "[\"long_context\",\"reasoning\",\"structured_output\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.602, + "details": { + "raw_score": "0.602", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "longbench-v2", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/longbench-v2\",\"https://api.llm-stats.com/leaderboard/benchmarks/longbench-v2\"]", + "raw_score_id": "longbench-v2::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "lvbench::lvbench-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.lvbench", + "source_data": { + "dataset_name": "LVBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/lvbench", + "https://api.llm-stats.com/leaderboard/benchmarks/lvbench" + ], + "additional_details": { + "raw_benchmark_id": "lvbench", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LVBench is an extreme long video understanding benchmark designed to evaluate multimodal models on videos up to two hours in duration. It contains 6 major categories and 21 subcategories, with videos averaging five times longer than existing datasets. The benchmark addresses applications requiring comprehension of extremely long videos.", + "metric_id": "llm_stats.lvbench.score", + "metric_name": "LVBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "lvbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LVBench", + "raw_categories": "[\"long_context\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.744, + "details": { + "raw_score": "0.744", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "lvbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/lvbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/lvbench\"]", + "raw_score_id": "lvbench::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mathvision::mathvision-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.mathvision", + "source_data": { + "dataset_name": "MathVision", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/mathvision", + "https://api.llm-stats.com/leaderboard/benchmarks/mathvision" + ], + "additional_details": { + "raw_benchmark_id": "mathvision", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MATH-Vision is a dataset designed to measure multimodal mathematical reasoning capabilities. It focuses on evaluating how well models can solve mathematical problems that require both visual understanding and mathematical reasoning, bridging the gap between visual and mathematical domains.", + "metric_id": "llm_stats.mathvision.score", + "metric_name": "MathVision score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mathvision", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MathVision", + "raw_categories": "[\"math\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "31" + } + }, + "score_details": { + "score": 0.862, + "details": { + "raw_score": "0.862", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "mathvision", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/mathvision\",\"https://api.llm-stats.com/leaderboard/benchmarks/mathvision\"]", + "raw_score_id": "mathvision::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mathvista-mini::mathvista-mini-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.mathvista-mini", + "source_data": { + "dataset_name": "MathVista-Mini", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/mathvista-mini", + "https://api.llm-stats.com/leaderboard/benchmarks/mathvista-mini" + ], + "additional_details": { + "raw_benchmark_id": "mathvista-mini", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MathVista-Mini is a smaller version of the MathVista benchmark that evaluates mathematical reasoning in visual contexts. It consists of examples derived from multimodal datasets involving mathematics, combining challenges from diverse mathematical and visual tasks to assess foundation models' ability to solve problems requiring both visual understanding and mathematical reasoning.", + "metric_id": "llm_stats.mathvista-mini.score", + "metric_name": "MathVista-Mini score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mathvista-mini", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MathVista-Mini", + "raw_categories": "[\"math\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.874, + "details": { + "raw_score": "0.874", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "mathvista-mini", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/mathvista-mini\",\"https://api.llm-stats.com/leaderboard/benchmarks/mathvista-mini\"]", + "raw_score_id": "mathvista-mini::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "maxife::maxife-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.maxife", + "source_data": { + "dataset_name": "MAXIFE", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/maxife", + "https://api.llm-stats.com/leaderboard/benchmarks/maxife" + ], + "additional_details": { + "raw_benchmark_id": "maxife", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MAXIFE is a multilingual benchmark evaluating LLMs on instruction following and execution across multiple languages and cultural contexts.", + "metric_id": "llm_stats.maxife.score", + "metric_name": "MAXIFE score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "maxife", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MAXIFE", + "raw_categories": "[\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.879, + "details": { + "raw_score": "0.879", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "maxife", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/maxife\",\"https://api.llm-stats.com/leaderboard/benchmarks/maxife\"]", + "raw_score_id": "maxife::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "medxpertqa::medxpertqa-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.medxpertqa", + "source_data": { + "dataset_name": "MedXpertQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/medxpertqa", + "https://api.llm-stats.com/leaderboard/benchmarks/medxpertqa" + ], + "additional_details": { + "raw_benchmark_id": "medxpertqa", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A comprehensive benchmark to evaluate expert-level medical knowledge and advanced reasoning, featuring 4,460 questions spanning 17 specialties and 11 body systems. Includes both text-only and multimodal subsets with expert-level exam questions incorporating diverse medical images and rich clinical information.", + "metric_id": "llm_stats.medxpertqa.score", + "metric_name": "MedXpertQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "medxpertqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MedXpertQA", + "raw_categories": "[\"multimodal\",\"reasoning\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.673, + "details": { + "raw_score": "0.673", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "medxpertqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/medxpertqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/medxpertqa\"]", + "raw_score_id": "medxpertqa::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mlvu::mlvu-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.mlvu", + "source_data": { + "dataset_name": "MLVU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/mlvu", + "https://api.llm-stats.com/leaderboard/benchmarks/mlvu" + ], + "additional_details": { + "raw_benchmark_id": "mlvu", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A comprehensive benchmark for multi-task long video understanding that evaluates multimodal large language models on videos ranging from 3 minutes to 2 hours across 9 distinct tasks including reasoning, captioning, recognition, and summarization.", + "metric_id": "llm_stats.mlvu.score", + "metric_name": "MLVU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mlvu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MLVU", + "raw_categories": "[\"long_context\",\"multimodal\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.873, + "details": { + "raw_score": "0.873", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "mlvu", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/mlvu\",\"https://api.llm-stats.com/leaderboard/benchmarks/mlvu\"]", + "raw_score_id": "mlvu::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmbench-v1.1::mmbench-v1.1-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.mmbench-v1.1", + "source_data": { + "dataset_name": "MMBench-V1.1", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/mmbench-v1.1", + "https://api.llm-stats.com/leaderboard/benchmarks/mmbench-v1.1" + ], + "additional_details": { + "raw_benchmark_id": "mmbench-v1.1", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Version 1.1 of MMBench, an improved bilingual benchmark for assessing multi-modal capabilities of vision-language models through multiple-choice questions in both English and Chinese, providing systematic evaluation across diverse vision-language tasks.", + "metric_id": "llm_stats.mmbench-v1.1.score", + "metric_name": "MMBench-V1.1 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmbench-v1.1", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMBench-V1.1", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.928, + "details": { + "raw_score": "0.928", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "mmbench-v1.1", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/mmbench-v1.1\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmbench-v1.1\"]", + "raw_score_id": "mmbench-v1.1::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlongbench-doc::mmlongbench-doc-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.mmlongbench-doc", + "source_data": { + "dataset_name": "MMLongBench-Doc", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/mmlongbench-doc", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlongbench-doc" + ], + "additional_details": { + "raw_benchmark_id": "mmlongbench-doc", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MMLongBench-Doc evaluates long document understanding capabilities in vision-language models.", + "metric_id": "llm_stats.mmlongbench-doc.score", + "metric_name": "MMLongBench-Doc score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlongbench-doc", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLongBench-Doc", + "raw_categories": "[\"long_context\",\"multimodal\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.59, + "details": { + "raw_score": "0.59", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "mmlongbench-doc", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/mmlongbench-doc\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlongbench-doc\"]", + "raw_score_id": "mmlongbench-doc::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu-pro::mmlu-pro-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.mmlu-pro", + "source_data": { + "dataset_name": "MMLU-Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/mmlu-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu-pro" + ], + "additional_details": { + "raw_benchmark_id": "mmlu-pro", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A more robust and challenging multi-task language understanding benchmark that extends MMLU by expanding multiple-choice options from 4 to 10, eliminating trivial questions, and focusing on reasoning-intensive tasks. Features over 12,000 curated questions across 14 domains and causes a 16-33% accuracy drop compared to original MMLU.", + "metric_id": "llm_stats.mmlu-pro.score", + "metric_name": "MMLU-Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU-Pro", + "raw_categories": "[\"language\",\"legal\",\"math\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "129" + } + }, + "score_details": { + "score": 0.867, + "details": { + "raw_score": "0.867", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "mmlu-pro", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/mmlu-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu-pro\"]", + "raw_score_id": "mmlu-pro::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu-prox::mmlu-prox-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.mmlu-prox", + "source_data": { + "dataset_name": "MMLU-ProX", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/mmlu-prox", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu-prox" + ], + "additional_details": { + "raw_benchmark_id": "mmlu-prox", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Extended version of MMLU-Pro providing additional challenging multiple-choice questions for evaluating language models across diverse academic and professional domains. Built on the foundation of the Massive Multitask Language Understanding benchmark framework.", + "metric_id": "llm_stats.mmlu-prox.score", + "metric_name": "MMLU-ProX score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu-prox", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU-ProX", + "raw_categories": "[\"language\",\"legal\",\"math\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "32" + } + }, + "score_details": { + "score": 0.822, + "details": { + "raw_score": "0.822", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "mmlu-prox", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/mmlu-prox\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu-prox\"]", + "raw_score_id": "mmlu-prox::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu-redux::mmlu-redux-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.mmlu-redux", + "source_data": { + "dataset_name": "MMLU-Redux", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/mmlu-redux", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu-redux" + ], + "additional_details": { + "raw_benchmark_id": "mmlu-redux", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "An improved version of the MMLU benchmark featuring manually re-annotated questions to identify and correct errors in the original dataset. Provides more reliable evaluation metrics for language models by addressing dataset quality issues found in the original MMLU.", + "metric_id": "llm_stats.mmlu-redux.score", + "metric_name": "MMLU-Redux score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu-redux", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU-Redux", + "raw_categories": "[\"language\",\"math\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "48" + } + }, + "score_details": { + "score": 0.94, + "details": { + "raw_score": "0.94", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "mmlu-redux", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/mmlu-redux\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu-redux\"]", + "raw_score_id": "mmlu-redux::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmmu-pro::mmmu-pro-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.mmmu-pro", + "source_data": { + "dataset_name": "MMMU-Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/mmmu-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/mmmu-pro" + ], + "additional_details": { + "raw_benchmark_id": "mmmu-pro", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A more robust multi-discipline multimodal understanding benchmark that enhances MMMU through a three-step process: filtering text-only answerable questions, augmenting candidate options, and introducing vision-only input settings. Achieves significantly lower model performance (16.8-26.9%) compared to original MMMU, providing more rigorous evaluation that closely mimics real-world scenarios.", + "metric_id": "llm_stats.mmmu-pro.score", + "metric_name": "MMMU-Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmmu-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMMU-Pro", + "raw_categories": "[\"multimodal\",\"reasoning\",\"general\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "60" + } + }, + "score_details": { + "score": 0.769, + "details": { + "raw_score": "0.769", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "mmmu-pro", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/mmmu-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmmu-pro\"]", + "raw_score_id": "mmmu-pro::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmmu::mmmu-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.mmmu", + "source_data": { + "dataset_name": "MMMU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/mmmu", + "https://api.llm-stats.com/leaderboard/benchmarks/mmmu" + ], + "additional_details": { + "raw_benchmark_id": "mmmu", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MMMU (Massive Multi-discipline Multimodal Understanding) is a benchmark designed to evaluate multimodal models on college-level subject knowledge and deliberate reasoning. Contains 11.5K meticulously collected multimodal questions from college exams, quizzes, and textbooks, covering six core disciplines: Art & Design, Business, Science, Health & Medicine, Humanities & Social Science, and Tech & Engineering across 30 subjects and 183 subfields.", + "metric_id": "llm_stats.mmmu.score", + "metric_name": "MMMU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmmu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMMU", + "raw_categories": "[\"multimodal\",\"reasoning\",\"general\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "63" + } + }, + "score_details": { + "score": 0.839, + "details": { + "raw_score": "0.839", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "mmmu", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/mmmu\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmmu\"]", + "raw_score_id": "mmmu::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmstar::mmstar-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.mmstar", + "source_data": { + "dataset_name": "MMStar", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/mmstar", + "https://api.llm-stats.com/leaderboard/benchmarks/mmstar" + ], + "additional_details": { + "raw_benchmark_id": "mmstar", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MMStar is an elite vision-indispensable multimodal benchmark comprising 1,500 challenge samples meticulously selected by humans to evaluate 6 core capabilities and 18 detailed axes. The benchmark addresses issues of visual content unnecessity and unintentional data leakage in existing multimodal evaluations.", + "metric_id": "llm_stats.mmstar.score", + "metric_name": "MMStar score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmstar", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMStar", + "raw_categories": "[\"multimodal\",\"reasoning\",\"general\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "22" + } + }, + "score_details": { + "score": 0.829, + "details": { + "raw_score": "0.829", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "mmstar", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/mmstar\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmstar\"]", + "raw_score_id": "mmstar::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmvu::mmvu-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.mmvu", + "source_data": { + "dataset_name": "MMVU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/mmvu", + "https://api.llm-stats.com/leaderboard/benchmarks/mmvu" + ], + "additional_details": { + "raw_benchmark_id": "mmvu", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MMVU (Multimodal Multi-disciplinary Video Understanding) is a benchmark for evaluating multimodal models on video understanding tasks across multiple disciplines, testing comprehension and reasoning capabilities on video content.", + "metric_id": "llm_stats.mmvu.score", + "metric_name": "MMVU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmvu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMVU", + "raw_categories": "[\"multimodal\",\"reasoning\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.747, + "details": { + "raw_score": "0.747", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "mmvu", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/mmvu\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmvu\"]", + "raw_score_id": "mmvu::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "multichallenge::multichallenge-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.multichallenge", + "source_data": { + "dataset_name": "Multi-Challenge", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/multichallenge", + "https://api.llm-stats.com/leaderboard/benchmarks/multichallenge" + ], + "additional_details": { + "raw_benchmark_id": "multichallenge", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MultiChallenge is a realistic multi-turn conversation evaluation benchmark that challenges frontier LLMs across four key categories: instruction retention (maintaining instructions throughout conversations), inference memory (recalling and connecting details from previous turns), reliable versioned editing (adapting to evolving instructions during collaborative editing), and self-coherence (avoiding contradictions in responses). The benchmark evaluates models on sustained, contextually complex dialogues across diverse topics including travel planning, technical documentation, and professional communication.", + "metric_id": "llm_stats.multichallenge.score", + "metric_name": "Multi-Challenge score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "multichallenge", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Multi-Challenge", + "raw_categories": "[\"reasoning\",\"communication\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "29" + } + }, + "score_details": { + "score": 0.615, + "details": { + "raw_score": "0.615", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "multichallenge", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/multichallenge\",\"https://api.llm-stats.com/leaderboard/benchmarks/multichallenge\"]", + "raw_score_id": "multichallenge::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mvbench::mvbench-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.mvbench", + "source_data": { + "dataset_name": "MVBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/mvbench", + "https://api.llm-stats.com/leaderboard/benchmarks/mvbench" + ], + "additional_details": { + "raw_benchmark_id": "mvbench", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A comprehensive multi-modal video understanding benchmark covering 20 challenging video tasks that require temporal understanding beyond single-frame analysis. Tasks span from perception to cognition, including action recognition, temporal reasoning, spatial reasoning, object interaction, scene transition, and counterfactual inference. Uses a novel static-to-dynamic method to systematically generate video tasks from existing annotations.", + "metric_id": "llm_stats.mvbench.score", + "metric_name": "MVBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mvbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MVBench", + "raw_categories": "[\"multimodal\",\"reasoning\",\"spatial_reasoning\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "17" + } + }, + "score_details": { + "score": 0.766, + "details": { + "raw_score": "0.766", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "mvbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/mvbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/mvbench\"]", + "raw_score_id": "mvbench::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "nova-63::nova-63-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.nova-63", + "source_data": { + "dataset_name": "NOVA-63", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/nova-63", + "https://api.llm-stats.com/leaderboard/benchmarks/nova-63" + ], + "additional_details": { + "raw_benchmark_id": "nova-63", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "NOVA-63 is a multilingual evaluation benchmark covering 63 languages, designed to assess LLM performance across diverse linguistic contexts and tasks.", + "metric_id": "llm_stats.nova-63.score", + "metric_name": "NOVA-63 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "nova-63", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "NOVA-63", + "raw_categories": "[\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.586, + "details": { + "raw_score": "0.586", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "nova-63", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/nova-63\",\"https://api.llm-stats.com/leaderboard/benchmarks/nova-63\"]", + "raw_score_id": "nova-63::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "nuscene::nuscene-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.nuscene", + "source_data": { + "dataset_name": "Nuscene", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/nuscene", + "https://api.llm-stats.com/leaderboard/benchmarks/nuscene" + ], + "additional_details": { + "raw_benchmark_id": "nuscene", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multimodal benchmark for scene understanding and reasoning over the nuScenes autonomous driving domain.", + "metric_id": "llm_stats.nuscene.score", + "metric_name": "Nuscene score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "nuscene", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Nuscene", + "raw_categories": "[\"multimodal\",\"reasoning\",\"spatial\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.154, + "details": { + "raw_score": "0.154", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "nuscene", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/nuscene\",\"https://api.llm-stats.com/leaderboard/benchmarks/nuscene\"]", + "raw_score_id": "nuscene::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ocrbench::ocrbench-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.ocrbench", + "source_data": { + "dataset_name": "OCRBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/ocrbench", + "https://api.llm-stats.com/leaderboard/benchmarks/ocrbench" + ], + "additional_details": { + "raw_benchmark_id": "ocrbench", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OCRBench: Comprehensive evaluation benchmark for assessing Optical Character Recognition (OCR) capabilities in Large Multimodal Models across text recognition, scene text VQA, and document understanding tasks", + "metric_id": "llm_stats.ocrbench.score", + "metric_name": "OCRBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ocrbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OCRBench", + "raw_categories": "[\"image_to_text\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "22" + } + }, + "score_details": { + "score": 0.921, + "details": { + "raw_score": "0.921", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "ocrbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/ocrbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/ocrbench\"]", + "raw_score_id": "ocrbench::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "odinw::odinw-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.odinw", + "source_data": { + "dataset_name": "ODinW", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/odinw", + "https://api.llm-stats.com/leaderboard/benchmarks/odinw" + ], + "additional_details": { + "raw_benchmark_id": "odinw", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Object Detection in the Wild (ODinW) benchmark for evaluating object detection models' task-level transfer ability across diverse real-world datasets in terms of prediction accuracy and adaptation efficiency", + "metric_id": "llm_stats.odinw.score", + "metric_name": "ODinW score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "odinw", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ODinW", + "raw_categories": "[\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.445, + "details": { + "raw_score": "0.445", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "odinw", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/odinw\",\"https://api.llm-stats.com/leaderboard/benchmarks/odinw\"]", + "raw_score_id": "odinw::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ojbench::ojbench-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.ojbench", + "source_data": { + "dataset_name": "OJBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/ojbench", + "https://api.llm-stats.com/leaderboard/benchmarks/ojbench" + ], + "additional_details": { + "raw_benchmark_id": "ojbench", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OJBench is a competition-level code benchmark designed to assess the competitive-level code reasoning abilities of large language models. It comprises 232 programming competition problems from NOI and ICPC, categorized into Easy, Medium, and Hard difficulty levels. The benchmark evaluates models' ability to solve complex competitive programming challenges using Python and C++.", + "metric_id": "llm_stats.ojbench.score", + "metric_name": "OJBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ojbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OJBench", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.395, + "details": { + "raw_score": "0.395", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "ojbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/ojbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/ojbench\"]", + "raw_score_id": "ojbench::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "omnidocbench-1.5::omnidocbench-1.5-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.omnidocbench-1.5", + "source_data": { + "dataset_name": "OmniDocBench 1.5", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/omnidocbench-1.5", + "https://api.llm-stats.com/leaderboard/benchmarks/omnidocbench-1.5" + ], + "additional_details": { + "raw_benchmark_id": "omnidocbench-1.5", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OmniDocBench 1.5 is a comprehensive benchmark for evaluating multimodal large language models on document understanding tasks, including OCR, document parsing, information extraction, and visual question answering across diverse document types. Lower Overall Edit Distance scores are better.", + "metric_id": "llm_stats.omnidocbench-1.5.score", + "metric_name": "OmniDocBench 1.5 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "omnidocbench-1.5", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OmniDocBench 1.5", + "raw_categories": "[\"multimodal\",\"reasoning\",\"structured_output\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.898, + "details": { + "raw_score": "0.898", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "omnidocbench-1.5", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/omnidocbench-1.5\",\"https://api.llm-stats.com/leaderboard/benchmarks/omnidocbench-1.5\"]", + "raw_score_id": "omnidocbench-1.5::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "osworld-verified::osworld-verified-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.osworld-verified", + "source_data": { + "dataset_name": "OSWorld-Verified", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/osworld-verified", + "https://api.llm-stats.com/leaderboard/benchmarks/osworld-verified" + ], + "additional_details": { + "raw_benchmark_id": "osworld-verified", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OSWorld-Verified is a verified subset of OSWorld, a scalable real computer environment for multimodal agents supporting task setup, execution-based evaluation, and interactive learning across Ubuntu, Windows, and macOS.", + "metric_id": "llm_stats.osworld-verified.score", + "metric_name": "OSWorld-Verified score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "osworld-verified", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OSWorld-Verified", + "raw_categories": "[\"multimodal\",\"general\",\"agents\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.58, + "details": { + "raw_score": "0.58", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "osworld-verified", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/osworld-verified\",\"https://api.llm-stats.com/leaderboard/benchmarks/osworld-verified\"]", + "raw_score_id": "osworld-verified::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "pmc-vqa::pmc-vqa-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.pmc-vqa", + "source_data": { + "dataset_name": "PMC-VQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/pmc-vqa", + "https://api.llm-stats.com/leaderboard/benchmarks/pmc-vqa" + ], + "additional_details": { + "raw_benchmark_id": "pmc-vqa", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A medical visual question answering benchmark built on biomedical literature and medical figures.", + "metric_id": "llm_stats.pmc-vqa.score", + "metric_name": "PMC-VQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "pmc-vqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PMC-VQA", + "raw_categories": "[\"multimodal\",\"reasoning\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.633, + "details": { + "raw_score": "0.633", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "pmc-vqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/pmc-vqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/pmc-vqa\"]", + "raw_score_id": "pmc-vqa::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "polymath::polymath-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.polymath", + "source_data": { + "dataset_name": "PolyMATH", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/polymath", + "https://api.llm-stats.com/leaderboard/benchmarks/polymath" + ], + "additional_details": { + "raw_benchmark_id": "polymath", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Polymath is a challenging multi-modal mathematical reasoning benchmark designed to evaluate the general cognitive reasoning abilities of Multi-modal Large Language Models (MLLMs). The benchmark comprises 5,000 manually collected high-quality images of cognitive textual and visual challenges across 10 distinct categories, including pattern recognition, spatial reasoning, and relative reasoning.", + "metric_id": "llm_stats.polymath.score", + "metric_name": "PolyMATH score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "polymath", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PolyMATH", + "raw_categories": "[\"math\",\"multimodal\",\"reasoning\",\"spatial_reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.689, + "details": { + "raw_score": "0.689", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "polymath", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/polymath\",\"https://api.llm-stats.com/leaderboard/benchmarks/polymath\"]", + "raw_score_id": "polymath::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "realworldqa::realworldqa-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.realworldqa", + "source_data": { + "dataset_name": "RealWorldQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/realworldqa", + "https://api.llm-stats.com/leaderboard/benchmarks/realworldqa" + ], + "additional_details": { + "raw_benchmark_id": "realworldqa", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "RealWorldQA is a benchmark designed to evaluate basic real-world spatial understanding capabilities of multimodal models. The initial release consists of over 700 anonymized images taken from vehicles and other real-world scenarios, each accompanied by a question and easily verifiable answer. Released by xAI as part of their Grok-1.5 Vision preview to test models' ability to understand natural scenes and spatial relationships in everyday visual contexts.", + "metric_id": "llm_stats.realworldqa.score", + "metric_name": "RealWorldQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "realworldqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "RealWorldQA", + "raw_categories": "[\"spatial_reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "25" + } + }, + "score_details": { + "score": 0.851, + "details": { + "raw_score": "0.851", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "realworldqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/realworldqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/realworldqa\"]", + "raw_score_id": "realworldqa::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "refcoco-avg::refcoco-avg-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.refcoco-avg", + "source_data": { + "dataset_name": "RefCOCO-avg", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/refcoco-avg", + "https://api.llm-stats.com/leaderboard/benchmarks/refcoco-avg" + ], + "additional_details": { + "raw_benchmark_id": "refcoco-avg", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "RefCOCO-avg measures object grounding accuracy averaged across RefCOCO, RefCOCO+, and RefCOCOg benchmarks.", + "metric_id": "llm_stats.refcoco-avg.score", + "metric_name": "RefCOCO-avg score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "refcoco-avg", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "RefCOCO-avg", + "raw_categories": "[\"spatial_reasoning\",\"grounding\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.913, + "details": { + "raw_score": "0.913", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "refcoco-avg", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/refcoco-avg\",\"https://api.llm-stats.com/leaderboard/benchmarks/refcoco-avg\"]", + "raw_score_id": "refcoco-avg::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "refspatialbench::refspatialbench-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.refspatialbench", + "source_data": { + "dataset_name": "RefSpatialBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/refspatialbench", + "https://api.llm-stats.com/leaderboard/benchmarks/refspatialbench" + ], + "additional_details": { + "raw_benchmark_id": "refspatialbench", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "RefSpatialBench evaluates spatial reference understanding and grounding.", + "metric_id": "llm_stats.refspatialbench.score", + "metric_name": "RefSpatialBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "refspatialbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "RefSpatialBench", + "raw_categories": "[\"spatial_reasoning\",\"grounding\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.693, + "details": { + "raw_score": "0.693", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "refspatialbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/refspatialbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/refspatialbench\"]", + "raw_score_id": "refspatialbench::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "screenspot-pro::screenspot-pro-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.screenspot-pro", + "source_data": { + "dataset_name": "ScreenSpot Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/screenspot-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/screenspot-pro" + ], + "additional_details": { + "raw_benchmark_id": "screenspot-pro", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ScreenSpot-Pro is a novel GUI grounding benchmark designed to rigorously evaluate the grounding capabilities of multimodal large language models (MLLMs) in professional high-resolution computing environments. The benchmark comprises 1,581 instructions across 23 applications spanning 5 industries and 3 operating systems, featuring authentic high-resolution images from professional domains with expert annotations. Unlike previous benchmarks that focus on cropped screenshots in consumer applications, ScreenSpot-Pro addresses the complexity and diversity of real-world professional software scenarios, revealing significant performance gaps in current MLLM GUI perception capabilities.", + "metric_id": "llm_stats.screenspot-pro.score", + "metric_name": "ScreenSpot Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "screenspot-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ScreenSpot Pro", + "raw_categories": "[\"multimodal\",\"spatial_reasoning\",\"grounding\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.704, + "details": { + "raw_score": "0.704", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "screenspot-pro", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/screenspot-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/screenspot-pro\"]", + "raw_score_id": "screenspot-pro::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "seal-0::seal-0-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.seal-0", + "source_data": { + "dataset_name": "Seal-0", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/seal-0", + "https://api.llm-stats.com/leaderboard/benchmarks/seal-0" + ], + "additional_details": { + "raw_benchmark_id": "seal-0", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Seal-0 is a benchmark for evaluating agentic search capabilities, testing models' ability to navigate and retrieve information using tools.", + "metric_id": "llm_stats.seal-0.score", + "metric_name": "Seal-0 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "seal-0", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Seal-0", + "raw_categories": "[\"reasoning\",\"search\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.441, + "details": { + "raw_score": "0.441", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "seal-0", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/seal-0\",\"https://api.llm-stats.com/leaderboard/benchmarks/seal-0\"]", + "raw_score_id": "seal-0::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "simplevqa::simplevqa-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.simplevqa", + "source_data": { + "dataset_name": "SimpleVQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/simplevqa", + "https://api.llm-stats.com/leaderboard/benchmarks/simplevqa" + ], + "additional_details": { + "raw_benchmark_id": "simplevqa", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SimpleVQA is a visual question answering benchmark focused on simple queries.", + "metric_id": "llm_stats.simplevqa.score", + "metric_name": "SimpleVQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "simplevqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SimpleVQA", + "raw_categories": "[\"image_to_text\",\"multimodal\",\"general\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.617, + "details": { + "raw_score": "0.617", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "simplevqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/simplevqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/simplevqa\"]", + "raw_score_id": "simplevqa::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "slakevqa::slakevqa-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.slakevqa", + "source_data": { + "dataset_name": "SlakeVQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/slakevqa", + "https://api.llm-stats.com/leaderboard/benchmarks/slakevqa" + ], + "additional_details": { + "raw_benchmark_id": "slakevqa", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A semantically-labeled knowledge-enhanced dataset for medical visual question answering. Contains 642 radiology images (CT scans, MRI scans, X-rays) covering five body parts and 14,028 bilingual English-Chinese question-answer pairs annotated by experienced physicians. Features comprehensive semantic labels and a structural medical knowledge base with both vision-only and knowledge-based questions requiring external medical knowledge reasoning.", + "metric_id": "llm_stats.slakevqa.score", + "metric_name": "SlakeVQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "slakevqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SlakeVQA", + "raw_categories": "[\"image_to_text\",\"multimodal\",\"reasoning\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.816, + "details": { + "raw_score": "0.816", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "slakevqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/slakevqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/slakevqa\"]", + "raw_score_id": "slakevqa::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "sunrgbd::sunrgbd-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.sunrgbd", + "source_data": { + "dataset_name": "SUNRGBD", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/sunrgbd", + "https://api.llm-stats.com/leaderboard/benchmarks/sunrgbd" + ], + "additional_details": { + "raw_benchmark_id": "sunrgbd", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SUNRGBD evaluates RGB-D scene understanding and 3D grounding capabilities.", + "metric_id": "llm_stats.sunrgbd.score", + "metric_name": "SUNRGBD score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "sunrgbd", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SUNRGBD", + "raw_categories": "[\"spatial_reasoning\",\"3d\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.362, + "details": { + "raw_score": "0.362", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "sunrgbd", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/sunrgbd\",\"https://api.llm-stats.com/leaderboard/benchmarks/sunrgbd\"]", + "raw_score_id": "sunrgbd::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "supergpqa::supergpqa-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.supergpqa", + "source_data": { + "dataset_name": "SuperGPQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/supergpqa", + "https://api.llm-stats.com/leaderboard/benchmarks/supergpqa" + ], + "additional_details": { + "raw_benchmark_id": "supergpqa", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SuperGPQA is a comprehensive benchmark that evaluates large language models across 285 graduate-level academic disciplines. The benchmark contains 25,957 questions covering 13 broad disciplinary areas including Engineering, Medicine, Science, and Law, with specialized fields in light industry, agriculture, and service-oriented domains. It employs a Human-LLM collaborative filtering mechanism with over 80 expert annotators to create challenging questions that assess graduate-level knowledge and reasoning capabilities.", + "metric_id": "llm_stats.supergpqa.score", + "metric_name": "SuperGPQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "supergpqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SuperGPQA", + "raw_categories": "[\"legal\",\"math\",\"physics\",\"reasoning\",\"finance\",\"general\",\"healthcare\",\"chemistry\",\"economics\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "34" + } + }, + "score_details": { + "score": 0.671, + "details": { + "raw_score": "0.671", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "supergpqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/supergpqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/supergpqa\"]", + "raw_score_id": "supergpqa::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "t2-bench::t2-bench-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.t2-bench", + "source_data": { + "dataset_name": "t2-bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/t2-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/t2-bench" + ], + "additional_details": { + "raw_benchmark_id": "t2-bench", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "t2-bench is a benchmark for evaluating agentic tool use capabilities, measuring how well models can select, sequence, and utilize tools to solve complex tasks. It tests autonomous planning and execution in multi-step scenarios.", + "metric_id": "llm_stats.t2-bench.score", + "metric_name": "t2-bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "t2-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "t2-bench", + "raw_categories": "[\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.795, + "details": { + "raw_score": "0.795", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "t2-bench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/t2-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/t2-bench\"]", + "raw_score_id": "t2-bench::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "tir-bench::tir-bench-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.tir-bench", + "source_data": { + "dataset_name": "TIR-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/tir-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/tir-bench" + ], + "additional_details": { + "raw_benchmark_id": "tir-bench", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A tool-calling and multimodal interaction benchmark for testing visual instruction following and execution reliability.", + "metric_id": "llm_stats.tir-bench.score", + "metric_name": "TIR-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "tir-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TIR-Bench", + "raw_categories": "[\"multimodal\",\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.532, + "details": { + "raw_score": "0.532", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "tir-bench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/tir-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/tir-bench\"]", + "raw_score_id": "tir-bench::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "v-star::v-star-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.v-star", + "source_data": { + "dataset_name": "V*", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/v-star", + "https://api.llm-stats.com/leaderboard/benchmarks/v-star" + ], + "additional_details": { + "raw_benchmark_id": "v-star", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A visual reasoning benchmark evaluating multimodal inference under challenging spatial and grounded tasks.", + "metric_id": "llm_stats.v-star.score", + "metric_name": "V* score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "v-star", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "V*", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.932, + "details": { + "raw_score": "0.932", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "v-star", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/v-star\",\"https://api.llm-stats.com/leaderboard/benchmarks/v-star\"]", + "raw_score_id": "v-star::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "videomme-w-o-sub.::videomme-w-o-sub.-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.videomme-w-o-sub.", + "source_data": { + "dataset_name": "VideoMME w/o sub.", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/videomme-w-o-sub.", + "https://api.llm-stats.com/leaderboard/benchmarks/videomme-w-o-sub." + ], + "additional_details": { + "raw_benchmark_id": "videomme-w-o-sub.", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Video-MME is a comprehensive evaluation benchmark for multi-modal large language models in video analysis. It features 900 videos across 6 primary visual domains with 30 subfields, ranging from 11 seconds to 1 hour in duration, with 2,700 question-answer pairs. The benchmark evaluates MLLMs' capabilities in processing sequential visual data and multi-modal content including video frames, subtitles, and audio.", + "metric_id": "llm_stats.videomme-w-o-sub..score", + "metric_name": "VideoMME w/o sub. score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "videomme-w-o-sub.", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VideoMME w/o sub.", + "raw_categories": "[\"multimodal\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.839, + "details": { + "raw_score": "0.839", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "videomme-w-o-sub.", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/videomme-w-o-sub.\",\"https://api.llm-stats.com/leaderboard/benchmarks/videomme-w-o-sub.\"]", + "raw_score_id": "videomme-w-o-sub.::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "videomme-w-sub.::videomme-w-sub.-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.videomme-w-sub.", + "source_data": { + "dataset_name": "VideoMME w sub.", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/videomme-w-sub.", + "https://api.llm-stats.com/leaderboard/benchmarks/videomme-w-sub." + ], + "additional_details": { + "raw_benchmark_id": "videomme-w-sub.", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The first-ever comprehensive evaluation benchmark of Multi-modal LLMs in Video analysis. Features 900 videos (254 hours) with 2,700 question-answer pairs covering 6 primary visual domains and 30 subfields. Evaluates temporal understanding across short (11 seconds) to long (1 hour) videos with multi-modal inputs including video frames, subtitles, and audio.", + "metric_id": "llm_stats.videomme-w-sub..score", + "metric_name": "VideoMME w sub. score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "videomme-w-sub.", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VideoMME w sub.", + "raw_categories": "[\"multimodal\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.873, + "details": { + "raw_score": "0.873", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "videomme-w-sub.", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/videomme-w-sub.\",\"https://api.llm-stats.com/leaderboard/benchmarks/videomme-w-sub.\"]", + "raw_score_id": "videomme-w-sub.::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "videommmu::videommmu-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.videommmu", + "source_data": { + "dataset_name": "VideoMMMU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/videommmu", + "https://api.llm-stats.com/leaderboard/benchmarks/videommmu" + ], + "additional_details": { + "raw_benchmark_id": "videommmu", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Video-MMMU evaluates Large Multimodal Models' ability to acquire knowledge from expert-level professional videos across six disciplines through three cognitive stages: perception, comprehension, and adaptation. Contains 300 videos and 900 human-annotated questions spanning Art, Business, Science, Medicine, Humanities, and Engineering.", + "metric_id": "llm_stats.videommmu.score", + "metric_name": "VideoMMMU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "videommmu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VideoMMMU", + "raw_categories": "[\"multimodal\",\"reasoning\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "26" + } + }, + "score_details": { + "score": 0.82, + "details": { + "raw_score": "0.82", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "videommmu", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/videommmu\",\"https://api.llm-stats.com/leaderboard/benchmarks/videommmu\"]", + "raw_score_id": "videommmu::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "vita-bench::vita-bench-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.vita-bench", + "source_data": { + "dataset_name": "VITA-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/vita-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/vita-bench" + ], + "additional_details": { + "raw_benchmark_id": "vita-bench", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VITA-Bench evaluates AI agents on real-world virtual task automation, measuring their ability to complete complex multi-step tasks in simulated environments.", + "metric_id": "llm_stats.vita-bench.score", + "metric_name": "VITA-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "vita-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VITA-Bench", + "raw_categories": "[\"reasoning\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.336, + "details": { + "raw_score": "0.336", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "vita-bench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/vita-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/vita-bench\"]", + "raw_score_id": "vita-bench::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "vlmsareblind::vlmsareblind-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.vlmsareblind", + "source_data": { + "dataset_name": "VLMsAreBlind", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/vlmsareblind", + "https://api.llm-stats.com/leaderboard/benchmarks/vlmsareblind" + ], + "additional_details": { + "raw_benchmark_id": "vlmsareblind", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A vision-language benchmark that probes blind spots and brittle reasoning in multimodal models.", + "metric_id": "llm_stats.vlmsareblind.score", + "metric_name": "VLMsAreBlind score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "vlmsareblind", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VLMsAreBlind", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.967, + "details": { + "raw_score": "0.967", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "vlmsareblind", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/vlmsareblind\",\"https://api.llm-stats.com/leaderboard/benchmarks/vlmsareblind\"]", + "raw_score_id": "vlmsareblind::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "widesearch::widesearch-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.widesearch", + "source_data": { + "dataset_name": "WideSearch", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/widesearch", + "https://api.llm-stats.com/leaderboard/benchmarks/widesearch" + ], + "additional_details": { + "raw_benchmark_id": "widesearch", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WideSearch is an agentic search benchmark that evaluates models' ability to perform broad, parallel search operations across multiple sources. It tests wide-coverage information retrieval and synthesis capabilities.", + "metric_id": "llm_stats.widesearch.score", + "metric_name": "WideSearch score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "widesearch", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WideSearch", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.605, + "details": { + "raw_score": "0.605", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "widesearch", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/widesearch\",\"https://api.llm-stats.com/leaderboard/benchmarks/widesearch\"]", + "raw_score_id": "widesearch::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "wmt24::wmt24-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.wmt24", + "source_data": { + "dataset_name": "WMT24++", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/wmt24", + "https://api.llm-stats.com/leaderboard/benchmarks/wmt24++" + ], + "additional_details": { + "raw_benchmark_id": "wmt24++", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WMT24++ is a comprehensive multilingual machine translation benchmark that expands the WMT24 dataset to cover 55 languages and dialects. It includes human-written references and post-edits across four domains (literary, news, social, and speech) to evaluate machine translation systems and large language models across diverse linguistic contexts.", + "metric_id": "llm_stats.wmt24.score", + "metric_name": "WMT24++ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "wmt24++", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WMT24++", + "raw_categories": "[\"language\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.783, + "details": { + "raw_score": "0.783", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "wmt24++", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/wmt24\",\"https://api.llm-stats.com/leaderboard/benchmarks/wmt24++\"]", + "raw_score_id": "wmt24++::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "zerobench-sub::zerobench-sub-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.zerobench-sub", + "source_data": { + "dataset_name": "ZEROBench-Sub", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/zerobench-sub", + "https://api.llm-stats.com/leaderboard/benchmarks/zerobench-sub" + ], + "additional_details": { + "raw_benchmark_id": "zerobench-sub", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ZEROBench-Sub is a subset of the ZEROBench benchmark.", + "metric_id": "llm_stats.zerobench-sub.score", + "metric_name": "ZEROBench-Sub score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "zerobench-sub", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ZEROBench-Sub", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.362, + "details": { + "raw_score": "0.362", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "zerobench-sub", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/zerobench-sub\",\"https://api.llm-stats.com/leaderboard/benchmarks/zerobench-sub\"]", + "raw_score_id": "zerobench-sub::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "zerobench::zerobench-qwen3.5-122b-a10b", + "evaluation_name": "llm_stats.zerobench", + "source_data": { + "dataset_name": "ZEROBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-122b-a10b", + "https://llm-stats.com/benchmarks/zerobench", + "https://api.llm-stats.com/leaderboard/benchmarks/zerobench" + ], + "additional_details": { + "raw_benchmark_id": "zerobench", + "raw_model_id": "qwen3.5-122b-a10b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ZEROBench is a challenging vision benchmark designed to test models on zero-shot visual understanding tasks.", + "metric_id": "llm_stats.zerobench.score", + "metric_name": "ZEROBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "zerobench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ZEROBench", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.09, + "details": { + "raw_score": "0.09", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-122b-a10b", + "raw_benchmark_id": "zerobench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-122b-a10b\",\"https://llm-stats.com/benchmarks/zerobench\",\"https://api.llm-stats.com/leaderboard/benchmarks/zerobench\"]", + "raw_score_id": "zerobench::qwen3.5-122b-a10b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/qwen/qwen3.5-27b/a6b13bd7-a9d2-4e07-868f-0cc9f6238eb3.json b/data/llm_stats/qwen/qwen3.5-27b/a6b13bd7-a9d2-4e07-868f-0cc9f6238eb3.json new file mode 100644 index 0000000000000000000000000000000000000000..27a03fbeaadf0fe729f0a5cfa3c92753f071357b --- /dev/null +++ b/data/llm_stats/qwen/qwen3.5-27b/a6b13bd7-a9d2-4e07-868f-0cc9f6238eb3.json @@ -0,0 +1,3948 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/qwen_qwen3.5-27b/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.5-27b", + "id": "qwen/qwen3.5-27b", + "developer": "qwen", + "additional_details": { + "raw_id": "qwen3.5-27b", + "raw_model_name": "qwen3.5-27b", + "raw_display_name": "Qwen3.5-27B", + "raw_organization_id": "qwen", + "raw_organization_name": "Alibaba Cloud / Qwen Team", + "raw_context_length": "262144", + "raw_input_modalities": "[\"image\",\"text\",\"video\"]", + "raw_output_modalities": "[\"text\"]", + "raw_input_price": "0.3", + "raw_output_price": "2.4", + "raw_model_id": "qwen3.5-27b", + "raw_provider_slug": "qwen", + "raw_provider_name": "ZeroEval" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aa-lcr::aa-lcr-qwen3.5-27b", + "evaluation_name": "llm_stats.aa-lcr", + "source_data": { + "dataset_name": "AA-LCR", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/aa-lcr", + "https://api.llm-stats.com/leaderboard/benchmarks/aa-lcr" + ], + "additional_details": { + "raw_benchmark_id": "aa-lcr", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Agent Arena Long Context Reasoning benchmark", + "metric_id": "llm_stats.aa-lcr.score", + "metric_name": "AA-LCR score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aa-lcr", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AA-LCR", + "raw_categories": "[\"long_context\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "14" + } + }, + "score_details": { + "score": 0.661, + "details": { + "raw_score": "0.661", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "aa-lcr", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/aa-lcr\",\"https://api.llm-stats.com/leaderboard/benchmarks/aa-lcr\"]", + "raw_score_id": "aa-lcr::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ai2d::ai2d-qwen3.5-27b", + "evaluation_name": "llm_stats.ai2d", + "source_data": { + "dataset_name": "AI2D", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/ai2d", + "https://api.llm-stats.com/leaderboard/benchmarks/ai2d" + ], + "additional_details": { + "raw_benchmark_id": "ai2d", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "AI2D is a dataset of 4,903 illustrative diagrams from grade school natural sciences (such as food webs, human physiology, and life cycles) with over 15,000 multiple choice questions and answers. The benchmark evaluates diagram understanding and visual reasoning capabilities, requiring models to interpret diagrammatic elements, relationships, and structure to answer questions about scientific concepts represented in visual form.", + "metric_id": "llm_stats.ai2d.score", + "metric_name": "AI2D score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ai2d", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AI2D", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "32" + } + }, + "score_details": { + "score": 0.929, + "details": { + "raw_score": "0.929", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "ai2d", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/ai2d\",\"https://api.llm-stats.com/leaderboard/benchmarks/ai2d\"]", + "raw_score_id": "ai2d::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "androidworld-sr::androidworld-sr-qwen3.5-27b", + "evaluation_name": "llm_stats.androidworld-sr", + "source_data": { + "dataset_name": "AndroidWorld_SR", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/androidworld-sr", + "https://api.llm-stats.com/leaderboard/benchmarks/androidworld-sr" + ], + "additional_details": { + "raw_benchmark_id": "androidworld-sr", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "AndroidWorld Success Rate (SR) benchmark - A dynamic benchmarking environment for autonomous agents operating on Android devices. Evaluates agents on 116 programmatic tasks across 20 real-world Android apps using multimodal inputs (screen screenshots, accessibility trees, and natural language instructions). Measures success rate of agents completing tasks like sending messages, creating calendar events, and navigating mobile interfaces. Published at ICLR 2025. Best current performance: 30.6% success rate (M3A agent) vs 80.0% human performance.", + "metric_id": "llm_stats.androidworld-sr.score", + "metric_name": "AndroidWorld_SR score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "androidworld-sr", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AndroidWorld_SR", + "raw_categories": "[\"multimodal\",\"reasoning\",\"general\",\"agents\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.642, + "details": { + "raw_score": "0.642", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "androidworld-sr", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/androidworld-sr\",\"https://api.llm-stats.com/leaderboard/benchmarks/androidworld-sr\"]", + "raw_score_id": "androidworld-sr::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "babyvision::babyvision-qwen3.5-27b", + "evaluation_name": "llm_stats.babyvision", + "source_data": { + "dataset_name": "BabyVision", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/babyvision", + "https://api.llm-stats.com/leaderboard/benchmarks/babyvision" + ], + "additional_details": { + "raw_benchmark_id": "babyvision", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A benchmark for early-stage visual reasoning and perception on child-like vision tasks.", + "metric_id": "llm_stats.babyvision.score", + "metric_name": "BabyVision score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "babyvision", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BabyVision", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.446, + "details": { + "raw_score": "0.446", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "babyvision", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/babyvision\",\"https://api.llm-stats.com/leaderboard/benchmarks/babyvision\"]", + "raw_score_id": "babyvision::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bfcl-v4::bfcl-v4-qwen3.5-27b", + "evaluation_name": "llm_stats.bfcl-v4", + "source_data": { + "dataset_name": "BFCL-V4", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/bfcl-v4", + "https://api.llm-stats.com/leaderboard/benchmarks/bfcl-v4" + ], + "additional_details": { + "raw_benchmark_id": "bfcl-v4", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Berkeley Function Calling Leaderboard V4 (BFCL-V4) evaluates LLMs on their ability to accurately call functions and APIs, including simple, multiple, parallel, and nested function calls across diverse programming scenarios.", + "metric_id": "llm_stats.bfcl-v4.score", + "metric_name": "BFCL-V4 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bfcl-v4", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BFCL-V4", + "raw_categories": "[\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.685, + "details": { + "raw_score": "0.685", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "bfcl-v4", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/bfcl-v4\",\"https://api.llm-stats.com/leaderboard/benchmarks/bfcl-v4\"]", + "raw_score_id": "bfcl-v4::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "browsecomp-zh::browsecomp-zh-qwen3.5-27b", + "evaluation_name": "llm_stats.browsecomp-zh", + "source_data": { + "dataset_name": "BrowseComp-zh", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/browsecomp-zh", + "https://api.llm-stats.com/leaderboard/benchmarks/browsecomp-zh" + ], + "additional_details": { + "raw_benchmark_id": "browsecomp-zh", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A high-difficulty benchmark purpose-built to comprehensively evaluate LLM agents on the Chinese web, consisting of 289 multi-hop questions spanning 11 diverse domains including Film & TV, Technology, Medicine, and History. Questions are reverse-engineered from short, objective, and easily verifiable answers, requiring sophisticated reasoning and information reconciliation beyond basic retrieval. The benchmark addresses linguistic, infrastructural, and censorship-related complexities in Chinese web environments.", + "metric_id": "llm_stats.browsecomp-zh.score", + "metric_name": "BrowseComp-zh score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "browsecomp-zh", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BrowseComp-zh", + "raw_categories": "[\"reasoning\",\"search\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.621, + "details": { + "raw_score": "0.621", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "browsecomp-zh", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/browsecomp-zh\",\"https://api.llm-stats.com/leaderboard/benchmarks/browsecomp-zh\"]", + "raw_score_id": "browsecomp-zh::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "c-eval::c-eval-qwen3.5-27b", + "evaluation_name": "llm_stats.c-eval", + "source_data": { + "dataset_name": "C-Eval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/c-eval", + "https://api.llm-stats.com/leaderboard/benchmarks/c-eval" + ], + "additional_details": { + "raw_benchmark_id": "c-eval", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "C-Eval is a comprehensive Chinese evaluation suite designed to assess advanced knowledge and reasoning abilities of foundation models in a Chinese context. It comprises 13,948 multiple-choice questions across 52 diverse disciplines spanning humanities, science, and engineering, with four difficulty levels: middle school, high school, college, and professional. The benchmark includes C-Eval Hard, a subset of very challenging subjects requiring advanced reasoning abilities.", + "metric_id": "llm_stats.c-eval.score", + "metric_name": "C-Eval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "c-eval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "C-Eval", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.905, + "details": { + "raw_score": "0.905", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "c-eval", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/c-eval\",\"https://api.llm-stats.com/leaderboard/benchmarks/c-eval\"]", + "raw_score_id": "c-eval::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "cc-ocr::cc-ocr-qwen3.5-27b", + "evaluation_name": "llm_stats.cc-ocr", + "source_data": { + "dataset_name": "CC-OCR", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/cc-ocr", + "https://api.llm-stats.com/leaderboard/benchmarks/cc-ocr" + ], + "additional_details": { + "raw_benchmark_id": "cc-ocr", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A comprehensive OCR benchmark for evaluating Large Multimodal Models (LMMs) in literacy. Comprises four OCR-centric tracks: multi-scene text reading, multilingual text reading, document parsing, and key information extraction. Contains 39 subsets with 7,058 fully annotated images, 41% sourced from real applications. Tests capabilities including text grounding, multi-orientation text recognition, and detecting hallucination/repetition across diverse visual challenges.", + "metric_id": "llm_stats.cc-ocr.score", + "metric_name": "CC-OCR score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "cc-ocr", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CC-OCR", + "raw_categories": "[\"multimodal\",\"structured_output\",\"text-to-image\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.81, + "details": { + "raw_score": "0.81", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "cc-ocr", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/cc-ocr\",\"https://api.llm-stats.com/leaderboard/benchmarks/cc-ocr\"]", + "raw_score_id": "cc-ocr::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "charxiv-r::charxiv-r-qwen3.5-27b", + "evaluation_name": "llm_stats.charxiv-r", + "source_data": { + "dataset_name": "CharXiv-R", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/charxiv-r", + "https://api.llm-stats.com/leaderboard/benchmarks/charxiv-r" + ], + "additional_details": { + "raw_benchmark_id": "charxiv-r", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CharXiv-R is the reasoning component of the CharXiv benchmark, focusing on complex reasoning questions that require synthesizing information across visual chart elements. It evaluates multimodal large language models on their ability to understand and reason about scientific charts from arXiv papers through various reasoning tasks.", + "metric_id": "llm_stats.charxiv-r.score", + "metric_name": "CharXiv-R score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "charxiv-r", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CharXiv-R", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "43" + } + }, + "score_details": { + "score": 0.795, + "details": { + "raw_score": "0.795", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "charxiv-r", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/charxiv-r\",\"https://api.llm-stats.com/leaderboard/benchmarks/charxiv-r\"]", + "raw_score_id": "charxiv-r::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "codeforces::codeforces-qwen3.5-27b", + "evaluation_name": "llm_stats.codeforces", + "source_data": { + "dataset_name": "CodeForces", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/codeforces", + "https://api.llm-stats.com/leaderboard/benchmarks/codeforces" + ], + "additional_details": { + "raw_benchmark_id": "codeforces", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A competitive programming benchmark using problems from the CodeForces platform. The benchmark evaluates code generation capabilities of LLMs on algorithmic problems with difficulty ratings ranging from 800 to 2400. Problems cover diverse algorithmic categories including dynamic programming, graph algorithms, data structures, and mathematical problems with standardized evaluation through direct platform submission.", + "metric_id": "llm_stats.codeforces.score", + "metric_name": "CodeForces score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "codeforces", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CodeForces", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.807, + "details": { + "raw_score": "0.807", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "codeforces", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/codeforces\",\"https://api.llm-stats.com/leaderboard/benchmarks/codeforces\"]", + "raw_score_id": "codeforces::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "countbench::countbench-qwen3.5-27b", + "evaluation_name": "llm_stats.countbench", + "source_data": { + "dataset_name": "CountBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/countbench", + "https://api.llm-stats.com/leaderboard/benchmarks/countbench" + ], + "additional_details": { + "raw_benchmark_id": "countbench", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CountBench evaluates object counting capabilities in visual understanding.", + "metric_id": "llm_stats.countbench.score", + "metric_name": "CountBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "countbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CountBench", + "raw_categories": "[\"reasoning\",\"spatial_reasoning\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.978, + "details": { + "raw_score": "0.978", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "countbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/countbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/countbench\"]", + "raw_score_id": "countbench::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "deep-planning::deep-planning-qwen3.5-27b", + "evaluation_name": "llm_stats.deep-planning", + "source_data": { + "dataset_name": "DeepPlanning", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/deep-planning", + "https://api.llm-stats.com/leaderboard/benchmarks/deep-planning" + ], + "additional_details": { + "raw_benchmark_id": "deep-planning", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "DeepPlanning evaluates LLMs on complex multi-step planning tasks requiring long-horizon reasoning, goal decomposition, and strategic decision-making.", + "metric_id": "llm_stats.deep-planning.score", + "metric_name": "DeepPlanning score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "deep-planning", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DeepPlanning", + "raw_categories": "[\"reasoning\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.226, + "details": { + "raw_score": "0.226", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "deep-planning", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/deep-planning\",\"https://api.llm-stats.com/leaderboard/benchmarks/deep-planning\"]", + "raw_score_id": "deep-planning::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "dynamath::dynamath-qwen3.5-27b", + "evaluation_name": "llm_stats.dynamath", + "source_data": { + "dataset_name": "DynaMath", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/dynamath", + "https://api.llm-stats.com/leaderboard/benchmarks/dynamath" + ], + "additional_details": { + "raw_benchmark_id": "dynamath", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multimodal mathematics and reasoning benchmark focused on dynamic visual problem solving.", + "metric_id": "llm_stats.dynamath.score", + "metric_name": "DynaMath score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "dynamath", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DynaMath", + "raw_categories": "[\"math\",\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.877, + "details": { + "raw_score": "0.877", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "dynamath", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/dynamath\",\"https://api.llm-stats.com/leaderboard/benchmarks/dynamath\"]", + "raw_score_id": "dynamath::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "embspatialbench::embspatialbench-qwen3.5-27b", + "evaluation_name": "llm_stats.embspatialbench", + "source_data": { + "dataset_name": "EmbSpatialBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/embspatialbench", + "https://api.llm-stats.com/leaderboard/benchmarks/embspatialbench" + ], + "additional_details": { + "raw_benchmark_id": "embspatialbench", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "EmbSpatialBench evaluates embodied spatial understanding and reasoning capabilities.", + "metric_id": "llm_stats.embspatialbench.score", + "metric_name": "EmbSpatialBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "embspatialbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "EmbSpatialBench", + "raw_categories": "[\"spatial_reasoning\",\"embodied\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.845, + "details": { + "raw_score": "0.845", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "embspatialbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/embspatialbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/embspatialbench\"]", + "raw_score_id": "embspatialbench::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "erqa::erqa-qwen3.5-27b", + "evaluation_name": "llm_stats.erqa", + "source_data": { + "dataset_name": "ERQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/erqa", + "https://api.llm-stats.com/leaderboard/benchmarks/erqa" + ], + "additional_details": { + "raw_benchmark_id": "erqa", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Embodied Reasoning Question Answering benchmark consisting of 400 multiple-choice visual questions across spatial reasoning, trajectory reasoning, action reasoning, state estimation, and multi-view reasoning for evaluating AI capabilities in physical world interactions", + "metric_id": "llm_stats.erqa.score", + "metric_name": "ERQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "erqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ERQA", + "raw_categories": "[\"reasoning\",\"spatial_reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "22" + } + }, + "score_details": { + "score": 0.605, + "details": { + "raw_score": "0.605", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "erqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/erqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/erqa\"]", + "raw_score_id": "erqa::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "fullstackbench-en::fullstackbench-en-qwen3.5-27b", + "evaluation_name": "llm_stats.fullstackbench-en", + "source_data": { + "dataset_name": "FullStackBench en", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/fullstackbench-en", + "https://api.llm-stats.com/leaderboard/benchmarks/fullstackbench-en" + ], + "additional_details": { + "raw_benchmark_id": "fullstackbench-en", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "English subset of FullStackBench for evaluating end-to-end software engineering and full-stack development capability.", + "metric_id": "llm_stats.fullstackbench-en.score", + "metric_name": "FullStackBench en score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "fullstackbench-en", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FullStackBench en", + "raw_categories": "[\"reasoning\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.601, + "details": { + "raw_score": "0.601", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "fullstackbench-en", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/fullstackbench-en\",\"https://api.llm-stats.com/leaderboard/benchmarks/fullstackbench-en\"]", + "raw_score_id": "fullstackbench-en::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "fullstackbench-zh::fullstackbench-zh-qwen3.5-27b", + "evaluation_name": "llm_stats.fullstackbench-zh", + "source_data": { + "dataset_name": "FullStackBench zh", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/fullstackbench-zh", + "https://api.llm-stats.com/leaderboard/benchmarks/fullstackbench-zh" + ], + "additional_details": { + "raw_benchmark_id": "fullstackbench-zh", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Chinese subset of FullStackBench for evaluating end-to-end software engineering and full-stack development capability.", + "metric_id": "llm_stats.fullstackbench-zh.score", + "metric_name": "FullStackBench zh score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "fullstackbench-zh", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FullStackBench zh", + "raw_categories": "[\"reasoning\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.574, + "details": { + "raw_score": "0.574", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "fullstackbench-zh", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/fullstackbench-zh\",\"https://api.llm-stats.com/leaderboard/benchmarks/fullstackbench-zh\"]", + "raw_score_id": "fullstackbench-zh::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "global-piqa::global-piqa-qwen3.5-27b", + "evaluation_name": "llm_stats.global-piqa", + "source_data": { + "dataset_name": "Global PIQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/global-piqa", + "https://api.llm-stats.com/leaderboard/benchmarks/global-piqa" + ], + "additional_details": { + "raw_benchmark_id": "global-piqa", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Global PIQA is a multilingual commonsense reasoning benchmark that evaluates physical interaction knowledge across 100 languages and cultures. It tests AI systems' understanding of physical world knowledge in diverse cultural contexts through multiple choice questions about everyday situations requiring physical commonsense.", + "metric_id": "llm_stats.global-piqa.score", + "metric_name": "Global PIQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "global-piqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Global PIQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.875, + "details": { + "raw_score": "0.875", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "global-piqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/global-piqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/global-piqa\"]", + "raw_score_id": "global-piqa::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hallusion-bench::hallusion-bench-qwen3.5-27b", + "evaluation_name": "llm_stats.hallusion-bench", + "source_data": { + "dataset_name": "Hallusion Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/hallusion-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/hallusion-bench" + ], + "additional_details": { + "raw_benchmark_id": "hallusion-bench", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A comprehensive benchmark designed to evaluate image-context reasoning in large visual-language models (LVLMs) by challenging models with 346 images and 1,129 carefully crafted questions to assess language hallucination and visual illusion", + "metric_id": "llm_stats.hallusion-bench.score", + "metric_name": "Hallusion Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hallusion-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Hallusion Bench", + "raw_categories": "[\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.7, + "details": { + "raw_score": "0.7", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "hallusion-bench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/hallusion-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/hallusion-bench\"]", + "raw_score_id": "hallusion-bench::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hmmt-2025::hmmt-2025-qwen3.5-27b", + "evaluation_name": "llm_stats.hmmt-2025", + "source_data": { + "dataset_name": "HMMT 2025", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/hmmt-2025", + "https://api.llm-stats.com/leaderboard/benchmarks/hmmt-2025" + ], + "additional_details": { + "raw_benchmark_id": "hmmt-2025", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Harvard-MIT Mathematics Tournament 2025 - A prestigious student-organized mathematics competition for high school students featuring two tournaments (November 2025 at MIT and February 2026 at Harvard) with individual tests, team rounds, and guts rounds", + "metric_id": "llm_stats.hmmt-2025.score", + "metric_name": "HMMT 2025 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hmmt-2025", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HMMT 2025", + "raw_categories": "[\"math\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "33" + } + }, + "score_details": { + "score": 0.92, + "details": { + "raw_score": "0.92", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "hmmt-2025", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/hmmt-2025\",\"https://api.llm-stats.com/leaderboard/benchmarks/hmmt-2025\"]", + "raw_score_id": "hmmt-2025::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hmmt25::hmmt25-qwen3.5-27b", + "evaluation_name": "llm_stats.hmmt25", + "source_data": { + "dataset_name": "HMMT25", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/hmmt25", + "https://api.llm-stats.com/leaderboard/benchmarks/hmmt25" + ], + "additional_details": { + "raw_benchmark_id": "hmmt25", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Harvard-MIT Mathematics Tournament 2025 - A prestigious student-organized mathematics competition for high school students featuring two tournaments (November 2025 at MIT and February 2026 at Harvard) with individual tests, team rounds, and guts rounds", + "metric_id": "llm_stats.hmmt25.score", + "metric_name": "HMMT25 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hmmt25", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HMMT25", + "raw_categories": "[\"math\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "25" + } + }, + "score_details": { + "score": 0.898, + "details": { + "raw_score": "0.898", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "hmmt25", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/hmmt25\",\"https://api.llm-stats.com/leaderboard/benchmarks/hmmt25\"]", + "raw_score_id": "hmmt25::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "humanity-s-last-exam::humanity-s-last-exam-qwen3.5-27b", + "evaluation_name": "llm_stats.humanity-s-last-exam", + "source_data": { + "dataset_name": "Humanity's Last Exam", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/humanity-s-last-exam", + "https://api.llm-stats.com/leaderboard/benchmarks/humanity's-last-exam" + ], + "additional_details": { + "raw_benchmark_id": "humanity's-last-exam", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Humanity's Last Exam (HLE) is a multi-modal academic benchmark with 2,500 questions across mathematics, humanities, and natural sciences, designed to test LLM capabilities at the frontier of human knowledge with unambiguous, verifiable solutions", + "metric_id": "llm_stats.humanity-s-last-exam.score", + "metric_name": "Humanity's Last Exam score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "humanity's-last-exam", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Humanity's Last Exam", + "raw_categories": "[\"math\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "88" + } + }, + "score_details": { + "score": 0.485, + "details": { + "raw_score": "0.485", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "humanity's-last-exam", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/humanity-s-last-exam\",\"https://api.llm-stats.com/leaderboard/benchmarks/humanity's-last-exam\"]", + "raw_score_id": "humanity's-last-exam::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hypersim::hypersim-qwen3.5-27b", + "evaluation_name": "llm_stats.hypersim", + "source_data": { + "dataset_name": "Hypersim", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/hypersim", + "https://api.llm-stats.com/leaderboard/benchmarks/hypersim" + ], + "additional_details": { + "raw_benchmark_id": "hypersim", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Hypersim evaluates 3D grounding and depth understanding in synthetic indoor scenes.", + "metric_id": "llm_stats.hypersim.score", + "metric_name": "Hypersim score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hypersim", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Hypersim", + "raw_categories": "[\"spatial_reasoning\",\"3d\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.13, + "details": { + "raw_score": "0.13", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "hypersim", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/hypersim\",\"https://api.llm-stats.com/leaderboard/benchmarks/hypersim\"]", + "raw_score_id": "hypersim::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ifbench::ifbench-qwen3.5-27b", + "evaluation_name": "llm_stats.ifbench", + "source_data": { + "dataset_name": "IFBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/ifbench", + "https://api.llm-stats.com/leaderboard/benchmarks/ifbench" + ], + "additional_details": { + "raw_benchmark_id": "ifbench", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Instruction Following Benchmark evaluating model's ability to follow complex instructions", + "metric_id": "llm_stats.ifbench.score", + "metric_name": "IFBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ifbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IFBench", + "raw_categories": "[\"instruction_following\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.765, + "details": { + "raw_score": "0.765", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "ifbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/ifbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/ifbench\"]", + "raw_score_id": "ifbench::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ifeval::ifeval-qwen3.5-27b", + "evaluation_name": "llm_stats.ifeval", + "source_data": { + "dataset_name": "IFEval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/ifeval", + "https://api.llm-stats.com/leaderboard/benchmarks/ifeval" + ], + "additional_details": { + "raw_benchmark_id": "ifeval", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Instruction-Following Evaluation (IFEval) benchmark for large language models, focusing on verifiable instructions with 25 types of instructions and around 500 prompts containing one or more verifiable constraints", + "metric_id": "llm_stats.ifeval.score", + "metric_name": "IFEval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ifeval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IFEval", + "raw_categories": "[\"instruction_following\",\"structured_output\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "65" + } + }, + "score_details": { + "score": 0.95, + "details": { + "raw_score": "0.95", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "ifeval", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/ifeval\",\"https://api.llm-stats.com/leaderboard/benchmarks/ifeval\"]", + "raw_score_id": "ifeval::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "include::include-qwen3.5-27b", + "evaluation_name": "llm_stats.include", + "source_data": { + "dataset_name": "Include", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/include", + "https://api.llm-stats.com/leaderboard/benchmarks/include" + ], + "additional_details": { + "raw_benchmark_id": "include", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Include benchmark - specific documentation not found in official sources", + "metric_id": "llm_stats.include.score", + "metric_name": "Include score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "include", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Include", + "raw_categories": "[\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "31" + } + }, + "score_details": { + "score": 0.816, + "details": { + "raw_score": "0.816", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "include", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/include\",\"https://api.llm-stats.com/leaderboard/benchmarks/include\"]", + "raw_score_id": "include::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "lingoqa::lingoqa-qwen3.5-27b", + "evaluation_name": "llm_stats.lingoqa", + "source_data": { + "dataset_name": "LingoQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/lingoqa", + "https://api.llm-stats.com/leaderboard/benchmarks/lingoqa" + ], + "additional_details": { + "raw_benchmark_id": "lingoqa", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A benchmark for multimodal spatial-language understanding and visual-linguistic question answering.", + "metric_id": "llm_stats.lingoqa.score", + "metric_name": "LingoQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "lingoqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LingoQA", + "raw_categories": "[\"language\",\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.82, + "details": { + "raw_score": "0.82", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "lingoqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/lingoqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/lingoqa\"]", + "raw_score_id": "lingoqa::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "livecodebench-v6::livecodebench-v6-qwen3.5-27b", + "evaluation_name": "llm_stats.livecodebench-v6", + "source_data": { + "dataset_name": "LiveCodeBench v6", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/livecodebench-v6", + "https://api.llm-stats.com/leaderboard/benchmarks/livecodebench-v6" + ], + "additional_details": { + "raw_benchmark_id": "livecodebench-v6", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LiveCodeBench is a holistic and contamination-free evaluation benchmark for large language models for code. It continuously collects new problems from programming contests (LeetCode, AtCoder, CodeForces) and evaluates four different scenarios: code generation, self-repair, code execution, and test output prediction. Problems are annotated with release dates to enable evaluation on unseen problems released after a model's training cutoff.", + "metric_id": "llm_stats.livecodebench-v6.score", + "metric_name": "LiveCodeBench v6 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "livecodebench-v6", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LiveCodeBench v6", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "53" + } + }, + "score_details": { + "score": 0.807, + "details": { + "raw_score": "0.807", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "livecodebench-v6", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/livecodebench-v6\",\"https://api.llm-stats.com/leaderboard/benchmarks/livecodebench-v6\"]", + "raw_score_id": "livecodebench-v6::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "longbench-v2::longbench-v2-qwen3.5-27b", + "evaluation_name": "llm_stats.longbench-v2", + "source_data": { + "dataset_name": "LongBench v2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/longbench-v2", + "https://api.llm-stats.com/leaderboard/benchmarks/longbench-v2" + ], + "additional_details": { + "raw_benchmark_id": "longbench-v2", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LongBench v2 is a benchmark designed to assess the ability of LLMs to handle long-context problems requiring deep understanding and reasoning across real-world multitasks. It consists of 503 challenging multiple-choice questions with contexts ranging from 8k to 2M words across six major task categories: single-document QA, multi-document QA, long in-context learning, long-dialogue history understanding, code repository understanding, and long structured data understanding.", + "metric_id": "llm_stats.longbench-v2.score", + "metric_name": "LongBench v2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "longbench-v2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LongBench v2", + "raw_categories": "[\"long_context\",\"reasoning\",\"structured_output\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.606, + "details": { + "raw_score": "0.606", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "longbench-v2", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/longbench-v2\",\"https://api.llm-stats.com/leaderboard/benchmarks/longbench-v2\"]", + "raw_score_id": "longbench-v2::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "lvbench::lvbench-qwen3.5-27b", + "evaluation_name": "llm_stats.lvbench", + "source_data": { + "dataset_name": "LVBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/lvbench", + "https://api.llm-stats.com/leaderboard/benchmarks/lvbench" + ], + "additional_details": { + "raw_benchmark_id": "lvbench", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LVBench is an extreme long video understanding benchmark designed to evaluate multimodal models on videos up to two hours in duration. It contains 6 major categories and 21 subcategories, with videos averaging five times longer than existing datasets. The benchmark addresses applications requiring comprehension of extremely long videos.", + "metric_id": "llm_stats.lvbench.score", + "metric_name": "LVBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "lvbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LVBench", + "raw_categories": "[\"long_context\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.736, + "details": { + "raw_score": "0.736", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "lvbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/lvbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/lvbench\"]", + "raw_score_id": "lvbench::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mathvision::mathvision-qwen3.5-27b", + "evaluation_name": "llm_stats.mathvision", + "source_data": { + "dataset_name": "MathVision", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/mathvision", + "https://api.llm-stats.com/leaderboard/benchmarks/mathvision" + ], + "additional_details": { + "raw_benchmark_id": "mathvision", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MATH-Vision is a dataset designed to measure multimodal mathematical reasoning capabilities. It focuses on evaluating how well models can solve mathematical problems that require both visual understanding and mathematical reasoning, bridging the gap between visual and mathematical domains.", + "metric_id": "llm_stats.mathvision.score", + "metric_name": "MathVision score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mathvision", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MathVision", + "raw_categories": "[\"math\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "31" + } + }, + "score_details": { + "score": 0.86, + "details": { + "raw_score": "0.86", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "mathvision", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/mathvision\",\"https://api.llm-stats.com/leaderboard/benchmarks/mathvision\"]", + "raw_score_id": "mathvision::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mathvista-mini::mathvista-mini-qwen3.5-27b", + "evaluation_name": "llm_stats.mathvista-mini", + "source_data": { + "dataset_name": "MathVista-Mini", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/mathvista-mini", + "https://api.llm-stats.com/leaderboard/benchmarks/mathvista-mini" + ], + "additional_details": { + "raw_benchmark_id": "mathvista-mini", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MathVista-Mini is a smaller version of the MathVista benchmark that evaluates mathematical reasoning in visual contexts. It consists of examples derived from multimodal datasets involving mathematics, combining challenges from diverse mathematical and visual tasks to assess foundation models' ability to solve problems requiring both visual understanding and mathematical reasoning.", + "metric_id": "llm_stats.mathvista-mini.score", + "metric_name": "MathVista-Mini score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mathvista-mini", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MathVista-Mini", + "raw_categories": "[\"math\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.878, + "details": { + "raw_score": "0.878", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "mathvista-mini", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/mathvista-mini\",\"https://api.llm-stats.com/leaderboard/benchmarks/mathvista-mini\"]", + "raw_score_id": "mathvista-mini::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "maxife::maxife-qwen3.5-27b", + "evaluation_name": "llm_stats.maxife", + "source_data": { + "dataset_name": "MAXIFE", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/maxife", + "https://api.llm-stats.com/leaderboard/benchmarks/maxife" + ], + "additional_details": { + "raw_benchmark_id": "maxife", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MAXIFE is a multilingual benchmark evaluating LLMs on instruction following and execution across multiple languages and cultural contexts.", + "metric_id": "llm_stats.maxife.score", + "metric_name": "MAXIFE score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "maxife", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MAXIFE", + "raw_categories": "[\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.88, + "details": { + "raw_score": "0.88", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "maxife", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/maxife\",\"https://api.llm-stats.com/leaderboard/benchmarks/maxife\"]", + "raw_score_id": "maxife::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "medxpertqa::medxpertqa-qwen3.5-27b", + "evaluation_name": "llm_stats.medxpertqa", + "source_data": { + "dataset_name": "MedXpertQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/medxpertqa", + "https://api.llm-stats.com/leaderboard/benchmarks/medxpertqa" + ], + "additional_details": { + "raw_benchmark_id": "medxpertqa", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A comprehensive benchmark to evaluate expert-level medical knowledge and advanced reasoning, featuring 4,460 questions spanning 17 specialties and 11 body systems. Includes both text-only and multimodal subsets with expert-level exam questions incorporating diverse medical images and rich clinical information.", + "metric_id": "llm_stats.medxpertqa.score", + "metric_name": "MedXpertQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "medxpertqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MedXpertQA", + "raw_categories": "[\"multimodal\",\"reasoning\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.624, + "details": { + "raw_score": "0.624", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "medxpertqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/medxpertqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/medxpertqa\"]", + "raw_score_id": "medxpertqa::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mlvu::mlvu-qwen3.5-27b", + "evaluation_name": "llm_stats.mlvu", + "source_data": { + "dataset_name": "MLVU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/mlvu", + "https://api.llm-stats.com/leaderboard/benchmarks/mlvu" + ], + "additional_details": { + "raw_benchmark_id": "mlvu", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A comprehensive benchmark for multi-task long video understanding that evaluates multimodal large language models on videos ranging from 3 minutes to 2 hours across 9 distinct tasks including reasoning, captioning, recognition, and summarization.", + "metric_id": "llm_stats.mlvu.score", + "metric_name": "MLVU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mlvu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MLVU", + "raw_categories": "[\"long_context\",\"multimodal\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.859, + "details": { + "raw_score": "0.859", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "mlvu", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/mlvu\",\"https://api.llm-stats.com/leaderboard/benchmarks/mlvu\"]", + "raw_score_id": "mlvu::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmbench-v1.1::mmbench-v1.1-qwen3.5-27b", + "evaluation_name": "llm_stats.mmbench-v1.1", + "source_data": { + "dataset_name": "MMBench-V1.1", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/mmbench-v1.1", + "https://api.llm-stats.com/leaderboard/benchmarks/mmbench-v1.1" + ], + "additional_details": { + "raw_benchmark_id": "mmbench-v1.1", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Version 1.1 of MMBench, an improved bilingual benchmark for assessing multi-modal capabilities of vision-language models through multiple-choice questions in both English and Chinese, providing systematic evaluation across diverse vision-language tasks.", + "metric_id": "llm_stats.mmbench-v1.1.score", + "metric_name": "MMBench-V1.1 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmbench-v1.1", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMBench-V1.1", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.926, + "details": { + "raw_score": "0.926", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "mmbench-v1.1", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/mmbench-v1.1\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmbench-v1.1\"]", + "raw_score_id": "mmbench-v1.1::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlongbench-doc::mmlongbench-doc-qwen3.5-27b", + "evaluation_name": "llm_stats.mmlongbench-doc", + "source_data": { + "dataset_name": "MMLongBench-Doc", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/mmlongbench-doc", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlongbench-doc" + ], + "additional_details": { + "raw_benchmark_id": "mmlongbench-doc", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MMLongBench-Doc evaluates long document understanding capabilities in vision-language models.", + "metric_id": "llm_stats.mmlongbench-doc.score", + "metric_name": "MMLongBench-Doc score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlongbench-doc", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLongBench-Doc", + "raw_categories": "[\"long_context\",\"multimodal\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.602, + "details": { + "raw_score": "0.602", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "mmlongbench-doc", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/mmlongbench-doc\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlongbench-doc\"]", + "raw_score_id": "mmlongbench-doc::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu-pro::mmlu-pro-qwen3.5-27b", + "evaluation_name": "llm_stats.mmlu-pro", + "source_data": { + "dataset_name": "MMLU-Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/mmlu-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu-pro" + ], + "additional_details": { + "raw_benchmark_id": "mmlu-pro", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A more robust and challenging multi-task language understanding benchmark that extends MMLU by expanding multiple-choice options from 4 to 10, eliminating trivial questions, and focusing on reasoning-intensive tasks. Features over 12,000 curated questions across 14 domains and causes a 16-33% accuracy drop compared to original MMLU.", + "metric_id": "llm_stats.mmlu-pro.score", + "metric_name": "MMLU-Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU-Pro", + "raw_categories": "[\"language\",\"legal\",\"math\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "129" + } + }, + "score_details": { + "score": 0.861, + "details": { + "raw_score": "0.861", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "mmlu-pro", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/mmlu-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu-pro\"]", + "raw_score_id": "mmlu-pro::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu-prox::mmlu-prox-qwen3.5-27b", + "evaluation_name": "llm_stats.mmlu-prox", + "source_data": { + "dataset_name": "MMLU-ProX", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/mmlu-prox", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu-prox" + ], + "additional_details": { + "raw_benchmark_id": "mmlu-prox", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Extended version of MMLU-Pro providing additional challenging multiple-choice questions for evaluating language models across diverse academic and professional domains. Built on the foundation of the Massive Multitask Language Understanding benchmark framework.", + "metric_id": "llm_stats.mmlu-prox.score", + "metric_name": "MMLU-ProX score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu-prox", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU-ProX", + "raw_categories": "[\"language\",\"legal\",\"math\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "32" + } + }, + "score_details": { + "score": 0.822, + "details": { + "raw_score": "0.822", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "mmlu-prox", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/mmlu-prox\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu-prox\"]", + "raw_score_id": "mmlu-prox::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu-redux::mmlu-redux-qwen3.5-27b", + "evaluation_name": "llm_stats.mmlu-redux", + "source_data": { + "dataset_name": "MMLU-Redux", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/mmlu-redux", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu-redux" + ], + "additional_details": { + "raw_benchmark_id": "mmlu-redux", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "An improved version of the MMLU benchmark featuring manually re-annotated questions to identify and correct errors in the original dataset. Provides more reliable evaluation metrics for language models by addressing dataset quality issues found in the original MMLU.", + "metric_id": "llm_stats.mmlu-redux.score", + "metric_name": "MMLU-Redux score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu-redux", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU-Redux", + "raw_categories": "[\"language\",\"math\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "48" + } + }, + "score_details": { + "score": 0.932, + "details": { + "raw_score": "0.932", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "mmlu-redux", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/mmlu-redux\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu-redux\"]", + "raw_score_id": "mmlu-redux::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmmu::mmmu-qwen3.5-27b", + "evaluation_name": "llm_stats.mmmu", + "source_data": { + "dataset_name": "MMMU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/mmmu", + "https://api.llm-stats.com/leaderboard/benchmarks/mmmu" + ], + "additional_details": { + "raw_benchmark_id": "mmmu", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MMMU (Massive Multi-discipline Multimodal Understanding) is a benchmark designed to evaluate multimodal models on college-level subject knowledge and deliberate reasoning. Contains 11.5K meticulously collected multimodal questions from college exams, quizzes, and textbooks, covering six core disciplines: Art & Design, Business, Science, Health & Medicine, Humanities & Social Science, and Tech & Engineering across 30 subjects and 183 subfields.", + "metric_id": "llm_stats.mmmu.score", + "metric_name": "MMMU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmmu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMMU", + "raw_categories": "[\"multimodal\",\"reasoning\",\"general\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "63" + } + }, + "score_details": { + "score": 0.823, + "details": { + "raw_score": "0.823", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "mmmu", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/mmmu\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmmu\"]", + "raw_score_id": "mmmu::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmstar::mmstar-qwen3.5-27b", + "evaluation_name": "llm_stats.mmstar", + "source_data": { + "dataset_name": "MMStar", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/mmstar", + "https://api.llm-stats.com/leaderboard/benchmarks/mmstar" + ], + "additional_details": { + "raw_benchmark_id": "mmstar", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MMStar is an elite vision-indispensable multimodal benchmark comprising 1,500 challenge samples meticulously selected by humans to evaluate 6 core capabilities and 18 detailed axes. The benchmark addresses issues of visual content unnecessity and unintentional data leakage in existing multimodal evaluations.", + "metric_id": "llm_stats.mmstar.score", + "metric_name": "MMStar score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmstar", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMStar", + "raw_categories": "[\"multimodal\",\"reasoning\",\"general\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "22" + } + }, + "score_details": { + "score": 0.81, + "details": { + "raw_score": "0.81", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "mmstar", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/mmstar\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmstar\"]", + "raw_score_id": "mmstar::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmvu::mmvu-qwen3.5-27b", + "evaluation_name": "llm_stats.mmvu", + "source_data": { + "dataset_name": "MMVU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/mmvu", + "https://api.llm-stats.com/leaderboard/benchmarks/mmvu" + ], + "additional_details": { + "raw_benchmark_id": "mmvu", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MMVU (Multimodal Multi-disciplinary Video Understanding) is a benchmark for evaluating multimodal models on video understanding tasks across multiple disciplines, testing comprehension and reasoning capabilities on video content.", + "metric_id": "llm_stats.mmvu.score", + "metric_name": "MMVU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmvu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMVU", + "raw_categories": "[\"multimodal\",\"reasoning\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.733, + "details": { + "raw_score": "0.733", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "mmvu", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/mmvu\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmvu\"]", + "raw_score_id": "mmvu::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "multichallenge::multichallenge-qwen3.5-27b", + "evaluation_name": "llm_stats.multichallenge", + "source_data": { + "dataset_name": "Multi-Challenge", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/multichallenge", + "https://api.llm-stats.com/leaderboard/benchmarks/multichallenge" + ], + "additional_details": { + "raw_benchmark_id": "multichallenge", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MultiChallenge is a realistic multi-turn conversation evaluation benchmark that challenges frontier LLMs across four key categories: instruction retention (maintaining instructions throughout conversations), inference memory (recalling and connecting details from previous turns), reliable versioned editing (adapting to evolving instructions during collaborative editing), and self-coherence (avoiding contradictions in responses). The benchmark evaluates models on sustained, contextually complex dialogues across diverse topics including travel planning, technical documentation, and professional communication.", + "metric_id": "llm_stats.multichallenge.score", + "metric_name": "Multi-Challenge score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "multichallenge", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Multi-Challenge", + "raw_categories": "[\"reasoning\",\"communication\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "29" + } + }, + "score_details": { + "score": 0.608, + "details": { + "raw_score": "0.608", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "multichallenge", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/multichallenge\",\"https://api.llm-stats.com/leaderboard/benchmarks/multichallenge\"]", + "raw_score_id": "multichallenge::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mvbench::mvbench-qwen3.5-27b", + "evaluation_name": "llm_stats.mvbench", + "source_data": { + "dataset_name": "MVBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/mvbench", + "https://api.llm-stats.com/leaderboard/benchmarks/mvbench" + ], + "additional_details": { + "raw_benchmark_id": "mvbench", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A comprehensive multi-modal video understanding benchmark covering 20 challenging video tasks that require temporal understanding beyond single-frame analysis. Tasks span from perception to cognition, including action recognition, temporal reasoning, spatial reasoning, object interaction, scene transition, and counterfactual inference. Uses a novel static-to-dynamic method to systematically generate video tasks from existing annotations.", + "metric_id": "llm_stats.mvbench.score", + "metric_name": "MVBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mvbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MVBench", + "raw_categories": "[\"multimodal\",\"reasoning\",\"spatial_reasoning\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "17" + } + }, + "score_details": { + "score": 0.746, + "details": { + "raw_score": "0.746", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "mvbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/mvbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/mvbench\"]", + "raw_score_id": "mvbench::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "nova-63::nova-63-qwen3.5-27b", + "evaluation_name": "llm_stats.nova-63", + "source_data": { + "dataset_name": "NOVA-63", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/nova-63", + "https://api.llm-stats.com/leaderboard/benchmarks/nova-63" + ], + "additional_details": { + "raw_benchmark_id": "nova-63", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "NOVA-63 is a multilingual evaluation benchmark covering 63 languages, designed to assess LLM performance across diverse linguistic contexts and tasks.", + "metric_id": "llm_stats.nova-63.score", + "metric_name": "NOVA-63 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "nova-63", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "NOVA-63", + "raw_categories": "[\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.581, + "details": { + "raw_score": "0.581", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "nova-63", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/nova-63\",\"https://api.llm-stats.com/leaderboard/benchmarks/nova-63\"]", + "raw_score_id": "nova-63::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "nuscene::nuscene-qwen3.5-27b", + "evaluation_name": "llm_stats.nuscene", + "source_data": { + "dataset_name": "Nuscene", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/nuscene", + "https://api.llm-stats.com/leaderboard/benchmarks/nuscene" + ], + "additional_details": { + "raw_benchmark_id": "nuscene", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multimodal benchmark for scene understanding and reasoning over the nuScenes autonomous driving domain.", + "metric_id": "llm_stats.nuscene.score", + "metric_name": "Nuscene score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "nuscene", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Nuscene", + "raw_categories": "[\"multimodal\",\"reasoning\",\"spatial\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.152, + "details": { + "raw_score": "0.152", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "nuscene", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/nuscene\",\"https://api.llm-stats.com/leaderboard/benchmarks/nuscene\"]", + "raw_score_id": "nuscene::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ocrbench::ocrbench-qwen3.5-27b", + "evaluation_name": "llm_stats.ocrbench", + "source_data": { + "dataset_name": "OCRBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/ocrbench", + "https://api.llm-stats.com/leaderboard/benchmarks/ocrbench" + ], + "additional_details": { + "raw_benchmark_id": "ocrbench", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OCRBench: Comprehensive evaluation benchmark for assessing Optical Character Recognition (OCR) capabilities in Large Multimodal Models across text recognition, scene text VQA, and document understanding tasks", + "metric_id": "llm_stats.ocrbench.score", + "metric_name": "OCRBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ocrbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OCRBench", + "raw_categories": "[\"image_to_text\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "22" + } + }, + "score_details": { + "score": 0.894, + "details": { + "raw_score": "0.894", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "ocrbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/ocrbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/ocrbench\"]", + "raw_score_id": "ocrbench::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "odinw::odinw-qwen3.5-27b", + "evaluation_name": "llm_stats.odinw", + "source_data": { + "dataset_name": "ODinW", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/odinw", + "https://api.llm-stats.com/leaderboard/benchmarks/odinw" + ], + "additional_details": { + "raw_benchmark_id": "odinw", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Object Detection in the Wild (ODinW) benchmark for evaluating object detection models' task-level transfer ability across diverse real-world datasets in terms of prediction accuracy and adaptation efficiency", + "metric_id": "llm_stats.odinw.score", + "metric_name": "ODinW score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "odinw", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ODinW", + "raw_categories": "[\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.411, + "details": { + "raw_score": "0.411", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "odinw", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/odinw\",\"https://api.llm-stats.com/leaderboard/benchmarks/odinw\"]", + "raw_score_id": "odinw::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ojbench::ojbench-qwen3.5-27b", + "evaluation_name": "llm_stats.ojbench", + "source_data": { + "dataset_name": "OJBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/ojbench", + "https://api.llm-stats.com/leaderboard/benchmarks/ojbench" + ], + "additional_details": { + "raw_benchmark_id": "ojbench", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OJBench is a competition-level code benchmark designed to assess the competitive-level code reasoning abilities of large language models. It comprises 232 programming competition problems from NOI and ICPC, categorized into Easy, Medium, and Hard difficulty levels. The benchmark evaluates models' ability to solve complex competitive programming challenges using Python and C++.", + "metric_id": "llm_stats.ojbench.score", + "metric_name": "OJBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ojbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OJBench", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.401, + "details": { + "raw_score": "0.401", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "ojbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/ojbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/ojbench\"]", + "raw_score_id": "ojbench::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "omnidocbench-1.5::omnidocbench-1.5-qwen3.5-27b", + "evaluation_name": "llm_stats.omnidocbench-1.5", + "source_data": { + "dataset_name": "OmniDocBench 1.5", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/omnidocbench-1.5", + "https://api.llm-stats.com/leaderboard/benchmarks/omnidocbench-1.5" + ], + "additional_details": { + "raw_benchmark_id": "omnidocbench-1.5", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OmniDocBench 1.5 is a comprehensive benchmark for evaluating multimodal large language models on document understanding tasks, including OCR, document parsing, information extraction, and visual question answering across diverse document types. Lower Overall Edit Distance scores are better.", + "metric_id": "llm_stats.omnidocbench-1.5.score", + "metric_name": "OmniDocBench 1.5 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "omnidocbench-1.5", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OmniDocBench 1.5", + "raw_categories": "[\"multimodal\",\"reasoning\",\"structured_output\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.889, + "details": { + "raw_score": "0.889", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "omnidocbench-1.5", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/omnidocbench-1.5\",\"https://api.llm-stats.com/leaderboard/benchmarks/omnidocbench-1.5\"]", + "raw_score_id": "omnidocbench-1.5::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "osworld-verified::osworld-verified-qwen3.5-27b", + "evaluation_name": "llm_stats.osworld-verified", + "source_data": { + "dataset_name": "OSWorld-Verified", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/osworld-verified", + "https://api.llm-stats.com/leaderboard/benchmarks/osworld-verified" + ], + "additional_details": { + "raw_benchmark_id": "osworld-verified", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OSWorld-Verified is a verified subset of OSWorld, a scalable real computer environment for multimodal agents supporting task setup, execution-based evaluation, and interactive learning across Ubuntu, Windows, and macOS.", + "metric_id": "llm_stats.osworld-verified.score", + "metric_name": "OSWorld-Verified score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "osworld-verified", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OSWorld-Verified", + "raw_categories": "[\"multimodal\",\"general\",\"agents\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.562, + "details": { + "raw_score": "0.562", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "osworld-verified", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/osworld-verified\",\"https://api.llm-stats.com/leaderboard/benchmarks/osworld-verified\"]", + "raw_score_id": "osworld-verified::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "pmc-vqa::pmc-vqa-qwen3.5-27b", + "evaluation_name": "llm_stats.pmc-vqa", + "source_data": { + "dataset_name": "PMC-VQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/pmc-vqa", + "https://api.llm-stats.com/leaderboard/benchmarks/pmc-vqa" + ], + "additional_details": { + "raw_benchmark_id": "pmc-vqa", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A medical visual question answering benchmark built on biomedical literature and medical figures.", + "metric_id": "llm_stats.pmc-vqa.score", + "metric_name": "PMC-VQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "pmc-vqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PMC-VQA", + "raw_categories": "[\"multimodal\",\"reasoning\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.624, + "details": { + "raw_score": "0.624", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "pmc-vqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/pmc-vqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/pmc-vqa\"]", + "raw_score_id": "pmc-vqa::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "polymath::polymath-qwen3.5-27b", + "evaluation_name": "llm_stats.polymath", + "source_data": { + "dataset_name": "PolyMATH", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/polymath", + "https://api.llm-stats.com/leaderboard/benchmarks/polymath" + ], + "additional_details": { + "raw_benchmark_id": "polymath", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Polymath is a challenging multi-modal mathematical reasoning benchmark designed to evaluate the general cognitive reasoning abilities of Multi-modal Large Language Models (MLLMs). The benchmark comprises 5,000 manually collected high-quality images of cognitive textual and visual challenges across 10 distinct categories, including pattern recognition, spatial reasoning, and relative reasoning.", + "metric_id": "llm_stats.polymath.score", + "metric_name": "PolyMATH score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "polymath", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PolyMATH", + "raw_categories": "[\"math\",\"multimodal\",\"reasoning\",\"spatial_reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.712, + "details": { + "raw_score": "0.712", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "polymath", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/polymath\",\"https://api.llm-stats.com/leaderboard/benchmarks/polymath\"]", + "raw_score_id": "polymath::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "realworldqa::realworldqa-qwen3.5-27b", + "evaluation_name": "llm_stats.realworldqa", + "source_data": { + "dataset_name": "RealWorldQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/realworldqa", + "https://api.llm-stats.com/leaderboard/benchmarks/realworldqa" + ], + "additional_details": { + "raw_benchmark_id": "realworldqa", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "RealWorldQA is a benchmark designed to evaluate basic real-world spatial understanding capabilities of multimodal models. The initial release consists of over 700 anonymized images taken from vehicles and other real-world scenarios, each accompanied by a question and easily verifiable answer. Released by xAI as part of their Grok-1.5 Vision preview to test models' ability to understand natural scenes and spatial relationships in everyday visual contexts.", + "metric_id": "llm_stats.realworldqa.score", + "metric_name": "RealWorldQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "realworldqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "RealWorldQA", + "raw_categories": "[\"spatial_reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "25" + } + }, + "score_details": { + "score": 0.837, + "details": { + "raw_score": "0.837", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "realworldqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/realworldqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/realworldqa\"]", + "raw_score_id": "realworldqa::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "refcoco-avg::refcoco-avg-qwen3.5-27b", + "evaluation_name": "llm_stats.refcoco-avg", + "source_data": { + "dataset_name": "RefCOCO-avg", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/refcoco-avg", + "https://api.llm-stats.com/leaderboard/benchmarks/refcoco-avg" + ], + "additional_details": { + "raw_benchmark_id": "refcoco-avg", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "RefCOCO-avg measures object grounding accuracy averaged across RefCOCO, RefCOCO+, and RefCOCOg benchmarks.", + "metric_id": "llm_stats.refcoco-avg.score", + "metric_name": "RefCOCO-avg score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "refcoco-avg", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "RefCOCO-avg", + "raw_categories": "[\"spatial_reasoning\",\"grounding\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.909, + "details": { + "raw_score": "0.909", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "refcoco-avg", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/refcoco-avg\",\"https://api.llm-stats.com/leaderboard/benchmarks/refcoco-avg\"]", + "raw_score_id": "refcoco-avg::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "refspatialbench::refspatialbench-qwen3.5-27b", + "evaluation_name": "llm_stats.refspatialbench", + "source_data": { + "dataset_name": "RefSpatialBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/refspatialbench", + "https://api.llm-stats.com/leaderboard/benchmarks/refspatialbench" + ], + "additional_details": { + "raw_benchmark_id": "refspatialbench", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "RefSpatialBench evaluates spatial reference understanding and grounding.", + "metric_id": "llm_stats.refspatialbench.score", + "metric_name": "RefSpatialBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "refspatialbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "RefSpatialBench", + "raw_categories": "[\"spatial_reasoning\",\"grounding\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.677, + "details": { + "raw_score": "0.677", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "refspatialbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/refspatialbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/refspatialbench\"]", + "raw_score_id": "refspatialbench::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "screenspot-pro::screenspot-pro-qwen3.5-27b", + "evaluation_name": "llm_stats.screenspot-pro", + "source_data": { + "dataset_name": "ScreenSpot Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/screenspot-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/screenspot-pro" + ], + "additional_details": { + "raw_benchmark_id": "screenspot-pro", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ScreenSpot-Pro is a novel GUI grounding benchmark designed to rigorously evaluate the grounding capabilities of multimodal large language models (MLLMs) in professional high-resolution computing environments. The benchmark comprises 1,581 instructions across 23 applications spanning 5 industries and 3 operating systems, featuring authentic high-resolution images from professional domains with expert annotations. Unlike previous benchmarks that focus on cropped screenshots in consumer applications, ScreenSpot-Pro addresses the complexity and diversity of real-world professional software scenarios, revealing significant performance gaps in current MLLM GUI perception capabilities.", + "metric_id": "llm_stats.screenspot-pro.score", + "metric_name": "ScreenSpot Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "screenspot-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ScreenSpot Pro", + "raw_categories": "[\"multimodal\",\"spatial_reasoning\",\"grounding\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.703, + "details": { + "raw_score": "0.703", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "screenspot-pro", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/screenspot-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/screenspot-pro\"]", + "raw_score_id": "screenspot-pro::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "seal-0::seal-0-qwen3.5-27b", + "evaluation_name": "llm_stats.seal-0", + "source_data": { + "dataset_name": "Seal-0", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/seal-0", + "https://api.llm-stats.com/leaderboard/benchmarks/seal-0" + ], + "additional_details": { + "raw_benchmark_id": "seal-0", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Seal-0 is a benchmark for evaluating agentic search capabilities, testing models' ability to navigate and retrieve information using tools.", + "metric_id": "llm_stats.seal-0.score", + "metric_name": "Seal-0 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "seal-0", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Seal-0", + "raw_categories": "[\"reasoning\",\"search\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.472, + "details": { + "raw_score": "0.472", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "seal-0", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/seal-0\",\"https://api.llm-stats.com/leaderboard/benchmarks/seal-0\"]", + "raw_score_id": "seal-0::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "simplevqa::simplevqa-qwen3.5-27b", + "evaluation_name": "llm_stats.simplevqa", + "source_data": { + "dataset_name": "SimpleVQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/simplevqa", + "https://api.llm-stats.com/leaderboard/benchmarks/simplevqa" + ], + "additional_details": { + "raw_benchmark_id": "simplevqa", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SimpleVQA is a visual question answering benchmark focused on simple queries.", + "metric_id": "llm_stats.simplevqa.score", + "metric_name": "SimpleVQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "simplevqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SimpleVQA", + "raw_categories": "[\"image_to_text\",\"multimodal\",\"general\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.56, + "details": { + "raw_score": "0.56", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "simplevqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/simplevqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/simplevqa\"]", + "raw_score_id": "simplevqa::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "slakevqa::slakevqa-qwen3.5-27b", + "evaluation_name": "llm_stats.slakevqa", + "source_data": { + "dataset_name": "SlakeVQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/slakevqa", + "https://api.llm-stats.com/leaderboard/benchmarks/slakevqa" + ], + "additional_details": { + "raw_benchmark_id": "slakevqa", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A semantically-labeled knowledge-enhanced dataset for medical visual question answering. Contains 642 radiology images (CT scans, MRI scans, X-rays) covering five body parts and 14,028 bilingual English-Chinese question-answer pairs annotated by experienced physicians. Features comprehensive semantic labels and a structural medical knowledge base with both vision-only and knowledge-based questions requiring external medical knowledge reasoning.", + "metric_id": "llm_stats.slakevqa.score", + "metric_name": "SlakeVQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "slakevqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SlakeVQA", + "raw_categories": "[\"image_to_text\",\"multimodal\",\"reasoning\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.8, + "details": { + "raw_score": "0.8", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "slakevqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/slakevqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/slakevqa\"]", + "raw_score_id": "slakevqa::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "sunrgbd::sunrgbd-qwen3.5-27b", + "evaluation_name": "llm_stats.sunrgbd", + "source_data": { + "dataset_name": "SUNRGBD", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/sunrgbd", + "https://api.llm-stats.com/leaderboard/benchmarks/sunrgbd" + ], + "additional_details": { + "raw_benchmark_id": "sunrgbd", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SUNRGBD evaluates RGB-D scene understanding and 3D grounding capabilities.", + "metric_id": "llm_stats.sunrgbd.score", + "metric_name": "SUNRGBD score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "sunrgbd", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SUNRGBD", + "raw_categories": "[\"spatial_reasoning\",\"3d\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.354, + "details": { + "raw_score": "0.354", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "sunrgbd", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/sunrgbd\",\"https://api.llm-stats.com/leaderboard/benchmarks/sunrgbd\"]", + "raw_score_id": "sunrgbd::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "supergpqa::supergpqa-qwen3.5-27b", + "evaluation_name": "llm_stats.supergpqa", + "source_data": { + "dataset_name": "SuperGPQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/supergpqa", + "https://api.llm-stats.com/leaderboard/benchmarks/supergpqa" + ], + "additional_details": { + "raw_benchmark_id": "supergpqa", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SuperGPQA is a comprehensive benchmark that evaluates large language models across 285 graduate-level academic disciplines. The benchmark contains 25,957 questions covering 13 broad disciplinary areas including Engineering, Medicine, Science, and Law, with specialized fields in light industry, agriculture, and service-oriented domains. It employs a Human-LLM collaborative filtering mechanism with over 80 expert annotators to create challenging questions that assess graduate-level knowledge and reasoning capabilities.", + "metric_id": "llm_stats.supergpqa.score", + "metric_name": "SuperGPQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "supergpqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SuperGPQA", + "raw_categories": "[\"legal\",\"math\",\"physics\",\"reasoning\",\"finance\",\"general\",\"healthcare\",\"chemistry\",\"economics\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "34" + } + }, + "score_details": { + "score": 0.656, + "details": { + "raw_score": "0.656", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "supergpqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/supergpqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/supergpqa\"]", + "raw_score_id": "supergpqa::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "t2-bench::t2-bench-qwen3.5-27b", + "evaluation_name": "llm_stats.t2-bench", + "source_data": { + "dataset_name": "t2-bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/t2-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/t2-bench" + ], + "additional_details": { + "raw_benchmark_id": "t2-bench", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "t2-bench is a benchmark for evaluating agentic tool use capabilities, measuring how well models can select, sequence, and utilize tools to solve complex tasks. It tests autonomous planning and execution in multi-step scenarios.", + "metric_id": "llm_stats.t2-bench.score", + "metric_name": "t2-bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "t2-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "t2-bench", + "raw_categories": "[\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.79, + "details": { + "raw_score": "0.79", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "t2-bench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/t2-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/t2-bench\"]", + "raw_score_id": "t2-bench::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "tir-bench::tir-bench-qwen3.5-27b", + "evaluation_name": "llm_stats.tir-bench", + "source_data": { + "dataset_name": "TIR-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/tir-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/tir-bench" + ], + "additional_details": { + "raw_benchmark_id": "tir-bench", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A tool-calling and multimodal interaction benchmark for testing visual instruction following and execution reliability.", + "metric_id": "llm_stats.tir-bench.score", + "metric_name": "TIR-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "tir-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TIR-Bench", + "raw_categories": "[\"multimodal\",\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.598, + "details": { + "raw_score": "0.598", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "tir-bench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/tir-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/tir-bench\"]", + "raw_score_id": "tir-bench::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "v-star::v-star-qwen3.5-27b", + "evaluation_name": "llm_stats.v-star", + "source_data": { + "dataset_name": "V*", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/v-star", + "https://api.llm-stats.com/leaderboard/benchmarks/v-star" + ], + "additional_details": { + "raw_benchmark_id": "v-star", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A visual reasoning benchmark evaluating multimodal inference under challenging spatial and grounded tasks.", + "metric_id": "llm_stats.v-star.score", + "metric_name": "V* score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "v-star", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "V*", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.937, + "details": { + "raw_score": "0.937", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "v-star", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/v-star\",\"https://api.llm-stats.com/leaderboard/benchmarks/v-star\"]", + "raw_score_id": "v-star::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "videomme-w-o-sub.::videomme-w-o-sub.-qwen3.5-27b", + "evaluation_name": "llm_stats.videomme-w-o-sub.", + "source_data": { + "dataset_name": "VideoMME w/o sub.", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/videomme-w-o-sub.", + "https://api.llm-stats.com/leaderboard/benchmarks/videomme-w-o-sub." + ], + "additional_details": { + "raw_benchmark_id": "videomme-w-o-sub.", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Video-MME is a comprehensive evaluation benchmark for multi-modal large language models in video analysis. It features 900 videos across 6 primary visual domains with 30 subfields, ranging from 11 seconds to 1 hour in duration, with 2,700 question-answer pairs. The benchmark evaluates MLLMs' capabilities in processing sequential visual data and multi-modal content including video frames, subtitles, and audio.", + "metric_id": "llm_stats.videomme-w-o-sub..score", + "metric_name": "VideoMME w/o sub. score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "videomme-w-o-sub.", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VideoMME w/o sub.", + "raw_categories": "[\"multimodal\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.828, + "details": { + "raw_score": "0.828", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "videomme-w-o-sub.", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/videomme-w-o-sub.\",\"https://api.llm-stats.com/leaderboard/benchmarks/videomme-w-o-sub.\"]", + "raw_score_id": "videomme-w-o-sub.::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "videomme-w-sub.::videomme-w-sub.-qwen3.5-27b", + "evaluation_name": "llm_stats.videomme-w-sub.", + "source_data": { + "dataset_name": "VideoMME w sub.", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/videomme-w-sub.", + "https://api.llm-stats.com/leaderboard/benchmarks/videomme-w-sub." + ], + "additional_details": { + "raw_benchmark_id": "videomme-w-sub.", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The first-ever comprehensive evaluation benchmark of Multi-modal LLMs in Video analysis. Features 900 videos (254 hours) with 2,700 question-answer pairs covering 6 primary visual domains and 30 subfields. Evaluates temporal understanding across short (11 seconds) to long (1 hour) videos with multi-modal inputs including video frames, subtitles, and audio.", + "metric_id": "llm_stats.videomme-w-sub..score", + "metric_name": "VideoMME w sub. score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "videomme-w-sub.", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VideoMME w sub.", + "raw_categories": "[\"multimodal\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.87, + "details": { + "raw_score": "0.87", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "videomme-w-sub.", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/videomme-w-sub.\",\"https://api.llm-stats.com/leaderboard/benchmarks/videomme-w-sub.\"]", + "raw_score_id": "videomme-w-sub.::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "videommmu::videommmu-qwen3.5-27b", + "evaluation_name": "llm_stats.videommmu", + "source_data": { + "dataset_name": "VideoMMMU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/videommmu", + "https://api.llm-stats.com/leaderboard/benchmarks/videommmu" + ], + "additional_details": { + "raw_benchmark_id": "videommmu", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Video-MMMU evaluates Large Multimodal Models' ability to acquire knowledge from expert-level professional videos across six disciplines through three cognitive stages: perception, comprehension, and adaptation. Contains 300 videos and 900 human-annotated questions spanning Art, Business, Science, Medicine, Humanities, and Engineering.", + "metric_id": "llm_stats.videommmu.score", + "metric_name": "VideoMMMU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "videommmu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VideoMMMU", + "raw_categories": "[\"multimodal\",\"reasoning\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "26" + } + }, + "score_details": { + "score": 0.823, + "details": { + "raw_score": "0.823", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "videommmu", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/videommmu\",\"https://api.llm-stats.com/leaderboard/benchmarks/videommmu\"]", + "raw_score_id": "videommmu::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "vita-bench::vita-bench-qwen3.5-27b", + "evaluation_name": "llm_stats.vita-bench", + "source_data": { + "dataset_name": "VITA-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/vita-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/vita-bench" + ], + "additional_details": { + "raw_benchmark_id": "vita-bench", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VITA-Bench evaluates AI agents on real-world virtual task automation, measuring their ability to complete complex multi-step tasks in simulated environments.", + "metric_id": "llm_stats.vita-bench.score", + "metric_name": "VITA-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "vita-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VITA-Bench", + "raw_categories": "[\"reasoning\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.419, + "details": { + "raw_score": "0.419", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "vita-bench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/vita-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/vita-bench\"]", + "raw_score_id": "vita-bench::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "vlmsareblind::vlmsareblind-qwen3.5-27b", + "evaluation_name": "llm_stats.vlmsareblind", + "source_data": { + "dataset_name": "VLMsAreBlind", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/vlmsareblind", + "https://api.llm-stats.com/leaderboard/benchmarks/vlmsareblind" + ], + "additional_details": { + "raw_benchmark_id": "vlmsareblind", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A vision-language benchmark that probes blind spots and brittle reasoning in multimodal models.", + "metric_id": "llm_stats.vlmsareblind.score", + "metric_name": "VLMsAreBlind score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "vlmsareblind", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VLMsAreBlind", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.969, + "details": { + "raw_score": "0.969", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "vlmsareblind", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/vlmsareblind\",\"https://api.llm-stats.com/leaderboard/benchmarks/vlmsareblind\"]", + "raw_score_id": "vlmsareblind::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "widesearch::widesearch-qwen3.5-27b", + "evaluation_name": "llm_stats.widesearch", + "source_data": { + "dataset_name": "WideSearch", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/widesearch", + "https://api.llm-stats.com/leaderboard/benchmarks/widesearch" + ], + "additional_details": { + "raw_benchmark_id": "widesearch", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WideSearch is an agentic search benchmark that evaluates models' ability to perform broad, parallel search operations across multiple sources. It tests wide-coverage information retrieval and synthesis capabilities.", + "metric_id": "llm_stats.widesearch.score", + "metric_name": "WideSearch score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "widesearch", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WideSearch", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.611, + "details": { + "raw_score": "0.611", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "widesearch", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/widesearch\",\"https://api.llm-stats.com/leaderboard/benchmarks/widesearch\"]", + "raw_score_id": "widesearch::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "wmt24::wmt24-qwen3.5-27b", + "evaluation_name": "llm_stats.wmt24", + "source_data": { + "dataset_name": "WMT24++", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/wmt24", + "https://api.llm-stats.com/leaderboard/benchmarks/wmt24++" + ], + "additional_details": { + "raw_benchmark_id": "wmt24++", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WMT24++ is a comprehensive multilingual machine translation benchmark that expands the WMT24 dataset to cover 55 languages and dialects. It includes human-written references and post-edits across four domains (literary, news, social, and speech) to evaluate machine translation systems and large language models across diverse linguistic contexts.", + "metric_id": "llm_stats.wmt24.score", + "metric_name": "WMT24++ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "wmt24++", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WMT24++", + "raw_categories": "[\"language\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.776, + "details": { + "raw_score": "0.776", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "wmt24++", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/wmt24\",\"https://api.llm-stats.com/leaderboard/benchmarks/wmt24++\"]", + "raw_score_id": "wmt24++::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "zerobench-sub::zerobench-sub-qwen3.5-27b", + "evaluation_name": "llm_stats.zerobench-sub", + "source_data": { + "dataset_name": "ZEROBench-Sub", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/zerobench-sub", + "https://api.llm-stats.com/leaderboard/benchmarks/zerobench-sub" + ], + "additional_details": { + "raw_benchmark_id": "zerobench-sub", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ZEROBench-Sub is a subset of the ZEROBench benchmark.", + "metric_id": "llm_stats.zerobench-sub.score", + "metric_name": "ZEROBench-Sub score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "zerobench-sub", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ZEROBench-Sub", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.362, + "details": { + "raw_score": "0.362", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "zerobench-sub", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/zerobench-sub\",\"https://api.llm-stats.com/leaderboard/benchmarks/zerobench-sub\"]", + "raw_score_id": "zerobench-sub::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "zerobench::zerobench-qwen3.5-27b", + "evaluation_name": "llm_stats.zerobench", + "source_data": { + "dataset_name": "ZEROBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-27b", + "https://llm-stats.com/benchmarks/zerobench", + "https://api.llm-stats.com/leaderboard/benchmarks/zerobench" + ], + "additional_details": { + "raw_benchmark_id": "zerobench", + "raw_model_id": "qwen3.5-27b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ZEROBench is a challenging vision benchmark designed to test models on zero-shot visual understanding tasks.", + "metric_id": "llm_stats.zerobench.score", + "metric_name": "ZEROBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "zerobench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ZEROBench", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.1, + "details": { + "raw_score": "0.1", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-27b", + "raw_benchmark_id": "zerobench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-27b\",\"https://llm-stats.com/benchmarks/zerobench\",\"https://api.llm-stats.com/leaderboard/benchmarks/zerobench\"]", + "raw_score_id": "zerobench::qwen3.5-27b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/qwen/qwen3.5-2b/563a49c7-4327-4867-aef7-346768f4a0cb.json b/data/llm_stats/qwen/qwen3.5-2b/563a49c7-4327-4867-aef7-346768f4a0cb.json new file mode 100644 index 0000000000000000000000000000000000000000..023e6c7c1732143bab34574017d1d3f3b48ce1ea --- /dev/null +++ b/data/llm_stats/qwen/qwen3.5-2b/563a49c7-4327-4867-aef7-346768f4a0cb.json @@ -0,0 +1,463 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/qwen_qwen3.5-2b/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Qwen3.5-2B", + "id": "qwen/qwen3.5-2b", + "developer": "qwen", + "additional_details": { + "raw_id": "qwen3.5-2b", + "raw_name": "Qwen3.5-2B", + "raw_model_id": "qwen3.5-2b", + "raw_model_name": "Qwen3.5-2B", + "raw_organization_id": "qwen", + "raw_organization_name": "Alibaba Cloud / Qwen Team", + "raw_release_date": "2026-03-02", + "raw_announcement_date": "2026-03-02", + "raw_multimodal": "true", + "raw_param_count": "2000000000", + "raw_provider_slug": "qwen", + "raw_provider_name": "Alibaba Cloud / Qwen Team" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aa-lcr::aa-lcr-qwen3.5-2b", + "evaluation_name": "llm_stats.aa-lcr", + "source_data": { + "dataset_name": "AA-LCR", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-2b", + "https://llm-stats.com/benchmarks/aa-lcr", + "https://api.llm-stats.com/leaderboard/benchmarks/aa-lcr" + ], + "additional_details": { + "raw_benchmark_id": "aa-lcr", + "raw_model_id": "qwen3.5-2b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Agent Arena Long Context Reasoning benchmark", + "metric_id": "llm_stats.aa-lcr.score", + "metric_name": "AA-LCR score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aa-lcr", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AA-LCR", + "raw_categories": "[\"long_context\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "14" + } + }, + "score_details": { + "score": 0.256, + "details": { + "raw_score": "0.256", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-2b", + "raw_benchmark_id": "aa-lcr", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-2b\",\"https://llm-stats.com/benchmarks/aa-lcr\",\"https://api.llm-stats.com/leaderboard/benchmarks/aa-lcr\"]", + "raw_score_id": "aa-lcr::qwen3.5-2b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bfcl-v4::bfcl-v4-qwen3.5-2b", + "evaluation_name": "llm_stats.bfcl-v4", + "source_data": { + "dataset_name": "BFCL-V4", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-2b", + "https://llm-stats.com/benchmarks/bfcl-v4", + "https://api.llm-stats.com/leaderboard/benchmarks/bfcl-v4" + ], + "additional_details": { + "raw_benchmark_id": "bfcl-v4", + "raw_model_id": "qwen3.5-2b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Berkeley Function Calling Leaderboard V4 (BFCL-V4) evaluates LLMs on their ability to accurately call functions and APIs, including simple, multiple, parallel, and nested function calls across diverse programming scenarios.", + "metric_id": "llm_stats.bfcl-v4.score", + "metric_name": "BFCL-V4 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bfcl-v4", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BFCL-V4", + "raw_categories": "[\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.436, + "details": { + "raw_score": "0.436", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-2b", + "raw_benchmark_id": "bfcl-v4", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-2b\",\"https://llm-stats.com/benchmarks/bfcl-v4\",\"https://api.llm-stats.com/leaderboard/benchmarks/bfcl-v4\"]", + "raw_score_id": "bfcl-v4::qwen3.5-2b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "c-eval::c-eval-qwen3.5-2b", + "evaluation_name": "llm_stats.c-eval", + "source_data": { + "dataset_name": "C-Eval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-2b", + "https://llm-stats.com/benchmarks/c-eval", + "https://api.llm-stats.com/leaderboard/benchmarks/c-eval" + ], + "additional_details": { + "raw_benchmark_id": "c-eval", + "raw_model_id": "qwen3.5-2b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "C-Eval is a comprehensive Chinese evaluation suite designed to assess advanced knowledge and reasoning abilities of foundation models in a Chinese context. It comprises 13,948 multiple-choice questions across 52 diverse disciplines spanning humanities, science, and engineering, with four difficulty levels: middle school, high school, college, and professional. The benchmark includes C-Eval Hard, a subset of very challenging subjects requiring advanced reasoning abilities.", + "metric_id": "llm_stats.c-eval.score", + "metric_name": "C-Eval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "c-eval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "C-Eval", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.732, + "details": { + "raw_score": "0.732", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-2b", + "raw_benchmark_id": "c-eval", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-2b\",\"https://llm-stats.com/benchmarks/c-eval\",\"https://api.llm-stats.com/leaderboard/benchmarks/c-eval\"]", + "raw_score_id": "c-eval::qwen3.5-2b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "global-piqa::global-piqa-qwen3.5-2b", + "evaluation_name": "llm_stats.global-piqa", + "source_data": { + "dataset_name": "Global PIQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-2b", + "https://llm-stats.com/benchmarks/global-piqa", + "https://api.llm-stats.com/leaderboard/benchmarks/global-piqa" + ], + "additional_details": { + "raw_benchmark_id": "global-piqa", + "raw_model_id": "qwen3.5-2b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Global PIQA is a multilingual commonsense reasoning benchmark that evaluates physical interaction knowledge across 100 languages and cultures. It tests AI systems' understanding of physical world knowledge in diverse cultural contexts through multiple choice questions about everyday situations requiring physical commonsense.", + "metric_id": "llm_stats.global-piqa.score", + "metric_name": "Global PIQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "global-piqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Global PIQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.693, + "details": { + "raw_score": "0.693", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-2b", + "raw_benchmark_id": "global-piqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-2b\",\"https://llm-stats.com/benchmarks/global-piqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/global-piqa\"]", + "raw_score_id": "global-piqa::qwen3.5-2b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "longbench-v2::longbench-v2-qwen3.5-2b", + "evaluation_name": "llm_stats.longbench-v2", + "source_data": { + "dataset_name": "LongBench v2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-2b", + "https://llm-stats.com/benchmarks/longbench-v2", + "https://api.llm-stats.com/leaderboard/benchmarks/longbench-v2" + ], + "additional_details": { + "raw_benchmark_id": "longbench-v2", + "raw_model_id": "qwen3.5-2b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LongBench v2 is a benchmark designed to assess the ability of LLMs to handle long-context problems requiring deep understanding and reasoning across real-world multitasks. It consists of 503 challenging multiple-choice questions with contexts ranging from 8k to 2M words across six major task categories: single-document QA, multi-document QA, long in-context learning, long-dialogue history understanding, code repository understanding, and long structured data understanding.", + "metric_id": "llm_stats.longbench-v2.score", + "metric_name": "LongBench v2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "longbench-v2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LongBench v2", + "raw_categories": "[\"long_context\",\"reasoning\",\"structured_output\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.387, + "details": { + "raw_score": "0.387", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-2b", + "raw_benchmark_id": "longbench-v2", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-2b\",\"https://llm-stats.com/benchmarks/longbench-v2\",\"https://api.llm-stats.com/leaderboard/benchmarks/longbench-v2\"]", + "raw_score_id": "longbench-v2::qwen3.5-2b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "maxife::maxife-qwen3.5-2b", + "evaluation_name": "llm_stats.maxife", + "source_data": { + "dataset_name": "MAXIFE", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-2b", + "https://llm-stats.com/benchmarks/maxife", + "https://api.llm-stats.com/leaderboard/benchmarks/maxife" + ], + "additional_details": { + "raw_benchmark_id": "maxife", + "raw_model_id": "qwen3.5-2b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MAXIFE is a multilingual benchmark evaluating LLMs on instruction following and execution across multiple languages and cultural contexts.", + "metric_id": "llm_stats.maxife.score", + "metric_name": "MAXIFE score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "maxife", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MAXIFE", + "raw_categories": "[\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.606, + "details": { + "raw_score": "0.606", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-2b", + "raw_benchmark_id": "maxife", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-2b\",\"https://llm-stats.com/benchmarks/maxife\",\"https://api.llm-stats.com/leaderboard/benchmarks/maxife\"]", + "raw_score_id": "maxife::qwen3.5-2b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "nova-63::nova-63-qwen3.5-2b", + "evaluation_name": "llm_stats.nova-63", + "source_data": { + "dataset_name": "NOVA-63", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-2b", + "https://llm-stats.com/benchmarks/nova-63", + "https://api.llm-stats.com/leaderboard/benchmarks/nova-63" + ], + "additional_details": { + "raw_benchmark_id": "nova-63", + "raw_model_id": "qwen3.5-2b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "NOVA-63 is a multilingual evaluation benchmark covering 63 languages, designed to assess LLM performance across diverse linguistic contexts and tasks.", + "metric_id": "llm_stats.nova-63.score", + "metric_name": "NOVA-63 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "nova-63", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "NOVA-63", + "raw_categories": "[\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.464, + "details": { + "raw_score": "0.464", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-2b", + "raw_benchmark_id": "nova-63", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-2b\",\"https://llm-stats.com/benchmarks/nova-63\",\"https://api.llm-stats.com/leaderboard/benchmarks/nova-63\"]", + "raw_score_id": "nova-63::qwen3.5-2b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "wmt24::wmt24-qwen3.5-2b", + "evaluation_name": "llm_stats.wmt24", + "source_data": { + "dataset_name": "WMT24++", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-2b", + "https://llm-stats.com/benchmarks/wmt24", + "https://api.llm-stats.com/leaderboard/benchmarks/wmt24++" + ], + "additional_details": { + "raw_benchmark_id": "wmt24++", + "raw_model_id": "qwen3.5-2b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WMT24++ is a comprehensive multilingual machine translation benchmark that expands the WMT24 dataset to cover 55 languages and dialects. It includes human-written references and post-edits across four domains (literary, news, social, and speech) to evaluate machine translation systems and large language models across diverse linguistic contexts.", + "metric_id": "llm_stats.wmt24.score", + "metric_name": "WMT24++ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "wmt24++", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WMT24++", + "raw_categories": "[\"language\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.458, + "details": { + "raw_score": "0.458", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-2b", + "raw_benchmark_id": "wmt24++", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-2b\",\"https://llm-stats.com/benchmarks/wmt24\",\"https://api.llm-stats.com/leaderboard/benchmarks/wmt24++\"]", + "raw_score_id": "wmt24++::qwen3.5-2b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/qwen/qwen3.5-35b-a3b/d496c714-663a-43e7-a6ec-f4fb5f084eda.json b/data/llm_stats/qwen/qwen3.5-35b-a3b/d496c714-663a-43e7-a6ec-f4fb5f084eda.json new file mode 100644 index 0000000000000000000000000000000000000000..dc2a9a770ddefe6a64da626fafaad6eab5102713 --- /dev/null +++ b/data/llm_stats/qwen/qwen3.5-35b-a3b/d496c714-663a-43e7-a6ec-f4fb5f084eda.json @@ -0,0 +1,3791 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/qwen_qwen3.5-35b-a3b/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Qwen3.5-35B-A3B", + "id": "qwen/qwen3.5-35b-a3b", + "developer": "qwen", + "additional_details": { + "raw_id": "qwen3.5-35b-a3b", + "raw_name": "Qwen3.5-35B-A3B", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_model_name": "Qwen3.5-35B-A3B", + "raw_organization_id": "qwen", + "raw_organization_name": "Alibaba Cloud / Qwen Team", + "raw_release_date": "2026-02-24", + "raw_announcement_date": "2026-02-24", + "raw_multimodal": "true", + "raw_param_count": "35000000000", + "raw_provider_slug": "qwen", + "raw_provider_name": "Alibaba Cloud / Qwen Team" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aa-lcr::aa-lcr-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.aa-lcr", + "source_data": { + "dataset_name": "AA-LCR", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/aa-lcr", + "https://api.llm-stats.com/leaderboard/benchmarks/aa-lcr" + ], + "additional_details": { + "raw_benchmark_id": "aa-lcr", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Agent Arena Long Context Reasoning benchmark", + "metric_id": "llm_stats.aa-lcr.score", + "metric_name": "AA-LCR score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aa-lcr", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AA-LCR", + "raw_categories": "[\"long_context\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "14" + } + }, + "score_details": { + "score": 0.585, + "details": { + "raw_score": "0.585", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "aa-lcr", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/aa-lcr\",\"https://api.llm-stats.com/leaderboard/benchmarks/aa-lcr\"]", + "raw_score_id": "aa-lcr::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ai2d::ai2d-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.ai2d", + "source_data": { + "dataset_name": "AI2D", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/ai2d", + "https://api.llm-stats.com/leaderboard/benchmarks/ai2d" + ], + "additional_details": { + "raw_benchmark_id": "ai2d", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "AI2D is a dataset of 4,903 illustrative diagrams from grade school natural sciences (such as food webs, human physiology, and life cycles) with over 15,000 multiple choice questions and answers. The benchmark evaluates diagram understanding and visual reasoning capabilities, requiring models to interpret diagrammatic elements, relationships, and structure to answer questions about scientific concepts represented in visual form.", + "metric_id": "llm_stats.ai2d.score", + "metric_name": "AI2D score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ai2d", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AI2D", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "32" + } + }, + "score_details": { + "score": 0.926, + "details": { + "raw_score": "0.926", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "ai2d", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/ai2d\",\"https://api.llm-stats.com/leaderboard/benchmarks/ai2d\"]", + "raw_score_id": "ai2d::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "androidworld-sr::androidworld-sr-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.androidworld-sr", + "source_data": { + "dataset_name": "AndroidWorld_SR", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/androidworld-sr", + "https://api.llm-stats.com/leaderboard/benchmarks/androidworld-sr" + ], + "additional_details": { + "raw_benchmark_id": "androidworld-sr", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "AndroidWorld Success Rate (SR) benchmark - A dynamic benchmarking environment for autonomous agents operating on Android devices. Evaluates agents on 116 programmatic tasks across 20 real-world Android apps using multimodal inputs (screen screenshots, accessibility trees, and natural language instructions). Measures success rate of agents completing tasks like sending messages, creating calendar events, and navigating mobile interfaces. Published at ICLR 2025. Best current performance: 30.6% success rate (M3A agent) vs 80.0% human performance.", + "metric_id": "llm_stats.androidworld-sr.score", + "metric_name": "AndroidWorld_SR score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "androidworld-sr", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AndroidWorld_SR", + "raw_categories": "[\"multimodal\",\"reasoning\",\"general\",\"agents\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.711, + "details": { + "raw_score": "0.711", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "androidworld-sr", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/androidworld-sr\",\"https://api.llm-stats.com/leaderboard/benchmarks/androidworld-sr\"]", + "raw_score_id": "androidworld-sr::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "babyvision::babyvision-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.babyvision", + "source_data": { + "dataset_name": "BabyVision", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/babyvision", + "https://api.llm-stats.com/leaderboard/benchmarks/babyvision" + ], + "additional_details": { + "raw_benchmark_id": "babyvision", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A benchmark for early-stage visual reasoning and perception on child-like vision tasks.", + "metric_id": "llm_stats.babyvision.score", + "metric_name": "BabyVision score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "babyvision", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BabyVision", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.384, + "details": { + "raw_score": "0.384", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "babyvision", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/babyvision\",\"https://api.llm-stats.com/leaderboard/benchmarks/babyvision\"]", + "raw_score_id": "babyvision::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bfcl-v4::bfcl-v4-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.bfcl-v4", + "source_data": { + "dataset_name": "BFCL-V4", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/bfcl-v4", + "https://api.llm-stats.com/leaderboard/benchmarks/bfcl-v4" + ], + "additional_details": { + "raw_benchmark_id": "bfcl-v4", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Berkeley Function Calling Leaderboard V4 (BFCL-V4) evaluates LLMs on their ability to accurately call functions and APIs, including simple, multiple, parallel, and nested function calls across diverse programming scenarios.", + "metric_id": "llm_stats.bfcl-v4.score", + "metric_name": "BFCL-V4 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bfcl-v4", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BFCL-V4", + "raw_categories": "[\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.673, + "details": { + "raw_score": "0.673", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "bfcl-v4", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/bfcl-v4\",\"https://api.llm-stats.com/leaderboard/benchmarks/bfcl-v4\"]", + "raw_score_id": "bfcl-v4::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "browsecomp-zh::browsecomp-zh-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.browsecomp-zh", + "source_data": { + "dataset_name": "BrowseComp-zh", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/browsecomp-zh", + "https://api.llm-stats.com/leaderboard/benchmarks/browsecomp-zh" + ], + "additional_details": { + "raw_benchmark_id": "browsecomp-zh", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A high-difficulty benchmark purpose-built to comprehensively evaluate LLM agents on the Chinese web, consisting of 289 multi-hop questions spanning 11 diverse domains including Film & TV, Technology, Medicine, and History. Questions are reverse-engineered from short, objective, and easily verifiable answers, requiring sophisticated reasoning and information reconciliation beyond basic retrieval. The benchmark addresses linguistic, infrastructural, and censorship-related complexities in Chinese web environments.", + "metric_id": "llm_stats.browsecomp-zh.score", + "metric_name": "BrowseComp-zh score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "browsecomp-zh", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BrowseComp-zh", + "raw_categories": "[\"reasoning\",\"search\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.695, + "details": { + "raw_score": "0.695", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "browsecomp-zh", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/browsecomp-zh\",\"https://api.llm-stats.com/leaderboard/benchmarks/browsecomp-zh\"]", + "raw_score_id": "browsecomp-zh::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "c-eval::c-eval-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.c-eval", + "source_data": { + "dataset_name": "C-Eval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/c-eval", + "https://api.llm-stats.com/leaderboard/benchmarks/c-eval" + ], + "additional_details": { + "raw_benchmark_id": "c-eval", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "C-Eval is a comprehensive Chinese evaluation suite designed to assess advanced knowledge and reasoning abilities of foundation models in a Chinese context. It comprises 13,948 multiple-choice questions across 52 diverse disciplines spanning humanities, science, and engineering, with four difficulty levels: middle school, high school, college, and professional. The benchmark includes C-Eval Hard, a subset of very challenging subjects requiring advanced reasoning abilities.", + "metric_id": "llm_stats.c-eval.score", + "metric_name": "C-Eval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "c-eval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "C-Eval", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.902, + "details": { + "raw_score": "0.902", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "c-eval", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/c-eval\",\"https://api.llm-stats.com/leaderboard/benchmarks/c-eval\"]", + "raw_score_id": "c-eval::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "cc-ocr::cc-ocr-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.cc-ocr", + "source_data": { + "dataset_name": "CC-OCR", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/cc-ocr", + "https://api.llm-stats.com/leaderboard/benchmarks/cc-ocr" + ], + "additional_details": { + "raw_benchmark_id": "cc-ocr", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A comprehensive OCR benchmark for evaluating Large Multimodal Models (LMMs) in literacy. Comprises four OCR-centric tracks: multi-scene text reading, multilingual text reading, document parsing, and key information extraction. Contains 39 subsets with 7,058 fully annotated images, 41% sourced from real applications. Tests capabilities including text grounding, multi-orientation text recognition, and detecting hallucination/repetition across diverse visual challenges.", + "metric_id": "llm_stats.cc-ocr.score", + "metric_name": "CC-OCR score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "cc-ocr", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CC-OCR", + "raw_categories": "[\"multimodal\",\"structured_output\",\"text-to-image\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.807, + "details": { + "raw_score": "0.807", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "cc-ocr", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/cc-ocr\",\"https://api.llm-stats.com/leaderboard/benchmarks/cc-ocr\"]", + "raw_score_id": "cc-ocr::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "codeforces::codeforces-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.codeforces", + "source_data": { + "dataset_name": "CodeForces", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/codeforces", + "https://api.llm-stats.com/leaderboard/benchmarks/codeforces" + ], + "additional_details": { + "raw_benchmark_id": "codeforces", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A competitive programming benchmark using problems from the CodeForces platform. The benchmark evaluates code generation capabilities of LLMs on algorithmic problems with difficulty ratings ranging from 800 to 2400. Problems cover diverse algorithmic categories including dynamic programming, graph algorithms, data structures, and mathematical problems with standardized evaluation through direct platform submission.", + "metric_id": "llm_stats.codeforces.score", + "metric_name": "CodeForces score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "codeforces", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CodeForces", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.822, + "details": { + "raw_score": "0.822", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "codeforces", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/codeforces\",\"https://api.llm-stats.com/leaderboard/benchmarks/codeforces\"]", + "raw_score_id": "codeforces::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "countbench::countbench-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.countbench", + "source_data": { + "dataset_name": "CountBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/countbench", + "https://api.llm-stats.com/leaderboard/benchmarks/countbench" + ], + "additional_details": { + "raw_benchmark_id": "countbench", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CountBench evaluates object counting capabilities in visual understanding.", + "metric_id": "llm_stats.countbench.score", + "metric_name": "CountBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "countbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CountBench", + "raw_categories": "[\"reasoning\",\"spatial_reasoning\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.978, + "details": { + "raw_score": "0.978", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "countbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/countbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/countbench\"]", + "raw_score_id": "countbench::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "deep-planning::deep-planning-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.deep-planning", + "source_data": { + "dataset_name": "DeepPlanning", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/deep-planning", + "https://api.llm-stats.com/leaderboard/benchmarks/deep-planning" + ], + "additional_details": { + "raw_benchmark_id": "deep-planning", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "DeepPlanning evaluates LLMs on complex multi-step planning tasks requiring long-horizon reasoning, goal decomposition, and strategic decision-making.", + "metric_id": "llm_stats.deep-planning.score", + "metric_name": "DeepPlanning score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "deep-planning", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DeepPlanning", + "raw_categories": "[\"reasoning\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.228, + "details": { + "raw_score": "0.228", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "deep-planning", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/deep-planning\",\"https://api.llm-stats.com/leaderboard/benchmarks/deep-planning\"]", + "raw_score_id": "deep-planning::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "dynamath::dynamath-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.dynamath", + "source_data": { + "dataset_name": "DynaMath", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/dynamath", + "https://api.llm-stats.com/leaderboard/benchmarks/dynamath" + ], + "additional_details": { + "raw_benchmark_id": "dynamath", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multimodal mathematics and reasoning benchmark focused on dynamic visual problem solving.", + "metric_id": "llm_stats.dynamath.score", + "metric_name": "DynaMath score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "dynamath", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DynaMath", + "raw_categories": "[\"math\",\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.85, + "details": { + "raw_score": "0.85", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "dynamath", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/dynamath\",\"https://api.llm-stats.com/leaderboard/benchmarks/dynamath\"]", + "raw_score_id": "dynamath::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "embspatialbench::embspatialbench-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.embspatialbench", + "source_data": { + "dataset_name": "EmbSpatialBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/embspatialbench", + "https://api.llm-stats.com/leaderboard/benchmarks/embspatialbench" + ], + "additional_details": { + "raw_benchmark_id": "embspatialbench", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "EmbSpatialBench evaluates embodied spatial understanding and reasoning capabilities.", + "metric_id": "llm_stats.embspatialbench.score", + "metric_name": "EmbSpatialBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "embspatialbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "EmbSpatialBench", + "raw_categories": "[\"spatial_reasoning\",\"embodied\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.831, + "details": { + "raw_score": "0.831", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "embspatialbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/embspatialbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/embspatialbench\"]", + "raw_score_id": "embspatialbench::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "erqa::erqa-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.erqa", + "source_data": { + "dataset_name": "ERQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/erqa", + "https://api.llm-stats.com/leaderboard/benchmarks/erqa" + ], + "additional_details": { + "raw_benchmark_id": "erqa", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Embodied Reasoning Question Answering benchmark consisting of 400 multiple-choice visual questions across spatial reasoning, trajectory reasoning, action reasoning, state estimation, and multi-view reasoning for evaluating AI capabilities in physical world interactions", + "metric_id": "llm_stats.erqa.score", + "metric_name": "ERQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "erqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ERQA", + "raw_categories": "[\"reasoning\",\"spatial_reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "22" + } + }, + "score_details": { + "score": 0.648, + "details": { + "raw_score": "0.648", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "erqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/erqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/erqa\"]", + "raw_score_id": "erqa::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "fullstackbench-en::fullstackbench-en-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.fullstackbench-en", + "source_data": { + "dataset_name": "FullStackBench en", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/fullstackbench-en", + "https://api.llm-stats.com/leaderboard/benchmarks/fullstackbench-en" + ], + "additional_details": { + "raw_benchmark_id": "fullstackbench-en", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "English subset of FullStackBench for evaluating end-to-end software engineering and full-stack development capability.", + "metric_id": "llm_stats.fullstackbench-en.score", + "metric_name": "FullStackBench en score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "fullstackbench-en", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FullStackBench en", + "raw_categories": "[\"reasoning\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.581, + "details": { + "raw_score": "0.581", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "fullstackbench-en", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/fullstackbench-en\",\"https://api.llm-stats.com/leaderboard/benchmarks/fullstackbench-en\"]", + "raw_score_id": "fullstackbench-en::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "fullstackbench-zh::fullstackbench-zh-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.fullstackbench-zh", + "source_data": { + "dataset_name": "FullStackBench zh", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/fullstackbench-zh", + "https://api.llm-stats.com/leaderboard/benchmarks/fullstackbench-zh" + ], + "additional_details": { + "raw_benchmark_id": "fullstackbench-zh", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Chinese subset of FullStackBench for evaluating end-to-end software engineering and full-stack development capability.", + "metric_id": "llm_stats.fullstackbench-zh.score", + "metric_name": "FullStackBench zh score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "fullstackbench-zh", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "FullStackBench zh", + "raw_categories": "[\"reasoning\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.55, + "details": { + "raw_score": "0.55", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "fullstackbench-zh", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/fullstackbench-zh\",\"https://api.llm-stats.com/leaderboard/benchmarks/fullstackbench-zh\"]", + "raw_score_id": "fullstackbench-zh::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "global-piqa::global-piqa-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.global-piqa", + "source_data": { + "dataset_name": "Global PIQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/global-piqa", + "https://api.llm-stats.com/leaderboard/benchmarks/global-piqa" + ], + "additional_details": { + "raw_benchmark_id": "global-piqa", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Global PIQA is a multilingual commonsense reasoning benchmark that evaluates physical interaction knowledge across 100 languages and cultures. It tests AI systems' understanding of physical world knowledge in diverse cultural contexts through multiple choice questions about everyday situations requiring physical commonsense.", + "metric_id": "llm_stats.global-piqa.score", + "metric_name": "Global PIQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "global-piqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Global PIQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.866, + "details": { + "raw_score": "0.866", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "global-piqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/global-piqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/global-piqa\"]", + "raw_score_id": "global-piqa::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hallusion-bench::hallusion-bench-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.hallusion-bench", + "source_data": { + "dataset_name": "Hallusion Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/hallusion-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/hallusion-bench" + ], + "additional_details": { + "raw_benchmark_id": "hallusion-bench", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A comprehensive benchmark designed to evaluate image-context reasoning in large visual-language models (LVLMs) by challenging models with 346 images and 1,129 carefully crafted questions to assess language hallucination and visual illusion", + "metric_id": "llm_stats.hallusion-bench.score", + "metric_name": "Hallusion Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hallusion-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Hallusion Bench", + "raw_categories": "[\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.679, + "details": { + "raw_score": "0.679", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "hallusion-bench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/hallusion-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/hallusion-bench\"]", + "raw_score_id": "hallusion-bench::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hmmt-2025::hmmt-2025-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.hmmt-2025", + "source_data": { + "dataset_name": "HMMT 2025", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/hmmt-2025", + "https://api.llm-stats.com/leaderboard/benchmarks/hmmt-2025" + ], + "additional_details": { + "raw_benchmark_id": "hmmt-2025", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Harvard-MIT Mathematics Tournament 2025 - A prestigious student-organized mathematics competition for high school students featuring two tournaments (November 2025 at MIT and February 2026 at Harvard) with individual tests, team rounds, and guts rounds", + "metric_id": "llm_stats.hmmt-2025.score", + "metric_name": "HMMT 2025 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hmmt-2025", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HMMT 2025", + "raw_categories": "[\"math\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "33" + } + }, + "score_details": { + "score": 0.89, + "details": { + "raw_score": "0.89", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "hmmt-2025", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/hmmt-2025\",\"https://api.llm-stats.com/leaderboard/benchmarks/hmmt-2025\"]", + "raw_score_id": "hmmt-2025::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hmmt25::hmmt25-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.hmmt25", + "source_data": { + "dataset_name": "HMMT25", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/hmmt25", + "https://api.llm-stats.com/leaderboard/benchmarks/hmmt25" + ], + "additional_details": { + "raw_benchmark_id": "hmmt25", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Harvard-MIT Mathematics Tournament 2025 - A prestigious student-organized mathematics competition for high school students featuring two tournaments (November 2025 at MIT and February 2026 at Harvard) with individual tests, team rounds, and guts rounds", + "metric_id": "llm_stats.hmmt25.score", + "metric_name": "HMMT25 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hmmt25", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HMMT25", + "raw_categories": "[\"math\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "25" + } + }, + "score_details": { + "score": 0.892, + "details": { + "raw_score": "0.892", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "hmmt25", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/hmmt25\",\"https://api.llm-stats.com/leaderboard/benchmarks/hmmt25\"]", + "raw_score_id": "hmmt25::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hypersim::hypersim-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.hypersim", + "source_data": { + "dataset_name": "Hypersim", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/hypersim", + "https://api.llm-stats.com/leaderboard/benchmarks/hypersim" + ], + "additional_details": { + "raw_benchmark_id": "hypersim", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Hypersim evaluates 3D grounding and depth understanding in synthetic indoor scenes.", + "metric_id": "llm_stats.hypersim.score", + "metric_name": "Hypersim score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hypersim", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Hypersim", + "raw_categories": "[\"spatial_reasoning\",\"3d\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.131, + "details": { + "raw_score": "0.131", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "hypersim", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/hypersim\",\"https://api.llm-stats.com/leaderboard/benchmarks/hypersim\"]", + "raw_score_id": "hypersim::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ifbench::ifbench-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.ifbench", + "source_data": { + "dataset_name": "IFBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/ifbench", + "https://api.llm-stats.com/leaderboard/benchmarks/ifbench" + ], + "additional_details": { + "raw_benchmark_id": "ifbench", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Instruction Following Benchmark evaluating model's ability to follow complex instructions", + "metric_id": "llm_stats.ifbench.score", + "metric_name": "IFBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ifbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IFBench", + "raw_categories": "[\"instruction_following\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.702, + "details": { + "raw_score": "0.702", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "ifbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/ifbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/ifbench\"]", + "raw_score_id": "ifbench::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ifeval::ifeval-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.ifeval", + "source_data": { + "dataset_name": "IFEval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/ifeval", + "https://api.llm-stats.com/leaderboard/benchmarks/ifeval" + ], + "additional_details": { + "raw_benchmark_id": "ifeval", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Instruction-Following Evaluation (IFEval) benchmark for large language models, focusing on verifiable instructions with 25 types of instructions and around 500 prompts containing one or more verifiable constraints", + "metric_id": "llm_stats.ifeval.score", + "metric_name": "IFEval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ifeval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IFEval", + "raw_categories": "[\"instruction_following\",\"structured_output\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "65" + } + }, + "score_details": { + "score": 0.919, + "details": { + "raw_score": "0.919", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "ifeval", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/ifeval\",\"https://api.llm-stats.com/leaderboard/benchmarks/ifeval\"]", + "raw_score_id": "ifeval::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "include::include-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.include", + "source_data": { + "dataset_name": "Include", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/include", + "https://api.llm-stats.com/leaderboard/benchmarks/include" + ], + "additional_details": { + "raw_benchmark_id": "include", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Include benchmark - specific documentation not found in official sources", + "metric_id": "llm_stats.include.score", + "metric_name": "Include score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "include", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Include", + "raw_categories": "[\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "31" + } + }, + "score_details": { + "score": 0.797, + "details": { + "raw_score": "0.797", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "include", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/include\",\"https://api.llm-stats.com/leaderboard/benchmarks/include\"]", + "raw_score_id": "include::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "lingoqa::lingoqa-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.lingoqa", + "source_data": { + "dataset_name": "LingoQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/lingoqa", + "https://api.llm-stats.com/leaderboard/benchmarks/lingoqa" + ], + "additional_details": { + "raw_benchmark_id": "lingoqa", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A benchmark for multimodal spatial-language understanding and visual-linguistic question answering.", + "metric_id": "llm_stats.lingoqa.score", + "metric_name": "LingoQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "lingoqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LingoQA", + "raw_categories": "[\"language\",\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.792, + "details": { + "raw_score": "0.792", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "lingoqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/lingoqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/lingoqa\"]", + "raw_score_id": "lingoqa::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "longbench-v2::longbench-v2-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.longbench-v2", + "source_data": { + "dataset_name": "LongBench v2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/longbench-v2", + "https://api.llm-stats.com/leaderboard/benchmarks/longbench-v2" + ], + "additional_details": { + "raw_benchmark_id": "longbench-v2", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LongBench v2 is a benchmark designed to assess the ability of LLMs to handle long-context problems requiring deep understanding and reasoning across real-world multitasks. It consists of 503 challenging multiple-choice questions with contexts ranging from 8k to 2M words across six major task categories: single-document QA, multi-document QA, long in-context learning, long-dialogue history understanding, code repository understanding, and long structured data understanding.", + "metric_id": "llm_stats.longbench-v2.score", + "metric_name": "LongBench v2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "longbench-v2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LongBench v2", + "raw_categories": "[\"long_context\",\"reasoning\",\"structured_output\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.59, + "details": { + "raw_score": "0.59", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "longbench-v2", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/longbench-v2\",\"https://api.llm-stats.com/leaderboard/benchmarks/longbench-v2\"]", + "raw_score_id": "longbench-v2::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "lvbench::lvbench-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.lvbench", + "source_data": { + "dataset_name": "LVBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/lvbench", + "https://api.llm-stats.com/leaderboard/benchmarks/lvbench" + ], + "additional_details": { + "raw_benchmark_id": "lvbench", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LVBench is an extreme long video understanding benchmark designed to evaluate multimodal models on videos up to two hours in duration. It contains 6 major categories and 21 subcategories, with videos averaging five times longer than existing datasets. The benchmark addresses applications requiring comprehension of extremely long videos.", + "metric_id": "llm_stats.lvbench.score", + "metric_name": "LVBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "lvbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LVBench", + "raw_categories": "[\"long_context\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.714, + "details": { + "raw_score": "0.714", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "lvbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/lvbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/lvbench\"]", + "raw_score_id": "lvbench::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mathvision::mathvision-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.mathvision", + "source_data": { + "dataset_name": "MathVision", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/mathvision", + "https://api.llm-stats.com/leaderboard/benchmarks/mathvision" + ], + "additional_details": { + "raw_benchmark_id": "mathvision", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MATH-Vision is a dataset designed to measure multimodal mathematical reasoning capabilities. It focuses on evaluating how well models can solve mathematical problems that require both visual understanding and mathematical reasoning, bridging the gap between visual and mathematical domains.", + "metric_id": "llm_stats.mathvision.score", + "metric_name": "MathVision score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mathvision", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MathVision", + "raw_categories": "[\"math\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "31" + } + }, + "score_details": { + "score": 0.839, + "details": { + "raw_score": "0.839", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "mathvision", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/mathvision\",\"https://api.llm-stats.com/leaderboard/benchmarks/mathvision\"]", + "raw_score_id": "mathvision::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mathvista-mini::mathvista-mini-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.mathvista-mini", + "source_data": { + "dataset_name": "MathVista-Mini", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/mathvista-mini", + "https://api.llm-stats.com/leaderboard/benchmarks/mathvista-mini" + ], + "additional_details": { + "raw_benchmark_id": "mathvista-mini", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MathVista-Mini is a smaller version of the MathVista benchmark that evaluates mathematical reasoning in visual contexts. It consists of examples derived from multimodal datasets involving mathematics, combining challenges from diverse mathematical and visual tasks to assess foundation models' ability to solve problems requiring both visual understanding and mathematical reasoning.", + "metric_id": "llm_stats.mathvista-mini.score", + "metric_name": "MathVista-Mini score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mathvista-mini", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MathVista-Mini", + "raw_categories": "[\"math\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.862, + "details": { + "raw_score": "0.862", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "mathvista-mini", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/mathvista-mini\",\"https://api.llm-stats.com/leaderboard/benchmarks/mathvista-mini\"]", + "raw_score_id": "mathvista-mini::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "maxife::maxife-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.maxife", + "source_data": { + "dataset_name": "MAXIFE", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/maxife", + "https://api.llm-stats.com/leaderboard/benchmarks/maxife" + ], + "additional_details": { + "raw_benchmark_id": "maxife", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MAXIFE is a multilingual benchmark evaluating LLMs on instruction following and execution across multiple languages and cultural contexts.", + "metric_id": "llm_stats.maxife.score", + "metric_name": "MAXIFE score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "maxife", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MAXIFE", + "raw_categories": "[\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.866, + "details": { + "raw_score": "0.866", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "maxife", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/maxife\",\"https://api.llm-stats.com/leaderboard/benchmarks/maxife\"]", + "raw_score_id": "maxife::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "medxpertqa::medxpertqa-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.medxpertqa", + "source_data": { + "dataset_name": "MedXpertQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/medxpertqa", + "https://api.llm-stats.com/leaderboard/benchmarks/medxpertqa" + ], + "additional_details": { + "raw_benchmark_id": "medxpertqa", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A comprehensive benchmark to evaluate expert-level medical knowledge and advanced reasoning, featuring 4,460 questions spanning 17 specialties and 11 body systems. Includes both text-only and multimodal subsets with expert-level exam questions incorporating diverse medical images and rich clinical information.", + "metric_id": "llm_stats.medxpertqa.score", + "metric_name": "MedXpertQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "medxpertqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MedXpertQA", + "raw_categories": "[\"multimodal\",\"reasoning\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.614, + "details": { + "raw_score": "0.614", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "medxpertqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/medxpertqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/medxpertqa\"]", + "raw_score_id": "medxpertqa::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mlvu::mlvu-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.mlvu", + "source_data": { + "dataset_name": "MLVU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/mlvu", + "https://api.llm-stats.com/leaderboard/benchmarks/mlvu" + ], + "additional_details": { + "raw_benchmark_id": "mlvu", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A comprehensive benchmark for multi-task long video understanding that evaluates multimodal large language models on videos ranging from 3 minutes to 2 hours across 9 distinct tasks including reasoning, captioning, recognition, and summarization.", + "metric_id": "llm_stats.mlvu.score", + "metric_name": "MLVU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mlvu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MLVU", + "raw_categories": "[\"long_context\",\"multimodal\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.856, + "details": { + "raw_score": "0.856", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "mlvu", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/mlvu\",\"https://api.llm-stats.com/leaderboard/benchmarks/mlvu\"]", + "raw_score_id": "mlvu::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmbench-v1.1::mmbench-v1.1-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.mmbench-v1.1", + "source_data": { + "dataset_name": "MMBench-V1.1", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/mmbench-v1.1", + "https://api.llm-stats.com/leaderboard/benchmarks/mmbench-v1.1" + ], + "additional_details": { + "raw_benchmark_id": "mmbench-v1.1", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Version 1.1 of MMBench, an improved bilingual benchmark for assessing multi-modal capabilities of vision-language models through multiple-choice questions in both English and Chinese, providing systematic evaluation across diverse vision-language tasks.", + "metric_id": "llm_stats.mmbench-v1.1.score", + "metric_name": "MMBench-V1.1 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmbench-v1.1", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMBench-V1.1", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.915, + "details": { + "raw_score": "0.915", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "mmbench-v1.1", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/mmbench-v1.1\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmbench-v1.1\"]", + "raw_score_id": "mmbench-v1.1::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlongbench-doc::mmlongbench-doc-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.mmlongbench-doc", + "source_data": { + "dataset_name": "MMLongBench-Doc", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/mmlongbench-doc", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlongbench-doc" + ], + "additional_details": { + "raw_benchmark_id": "mmlongbench-doc", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MMLongBench-Doc evaluates long document understanding capabilities in vision-language models.", + "metric_id": "llm_stats.mmlongbench-doc.score", + "metric_name": "MMLongBench-Doc score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlongbench-doc", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLongBench-Doc", + "raw_categories": "[\"long_context\",\"multimodal\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.595, + "details": { + "raw_score": "0.595", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "mmlongbench-doc", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/mmlongbench-doc\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlongbench-doc\"]", + "raw_score_id": "mmlongbench-doc::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu-pro::mmlu-pro-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.mmlu-pro", + "source_data": { + "dataset_name": "MMLU-Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/mmlu-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu-pro" + ], + "additional_details": { + "raw_benchmark_id": "mmlu-pro", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A more robust and challenging multi-task language understanding benchmark that extends MMLU by expanding multiple-choice options from 4 to 10, eliminating trivial questions, and focusing on reasoning-intensive tasks. Features over 12,000 curated questions across 14 domains and causes a 16-33% accuracy drop compared to original MMLU.", + "metric_id": "llm_stats.mmlu-pro.score", + "metric_name": "MMLU-Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU-Pro", + "raw_categories": "[\"language\",\"legal\",\"math\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "129" + } + }, + "score_details": { + "score": 0.853, + "details": { + "raw_score": "0.853", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "mmlu-pro", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/mmlu-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu-pro\"]", + "raw_score_id": "mmlu-pro::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu-prox::mmlu-prox-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.mmlu-prox", + "source_data": { + "dataset_name": "MMLU-ProX", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/mmlu-prox", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu-prox" + ], + "additional_details": { + "raw_benchmark_id": "mmlu-prox", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Extended version of MMLU-Pro providing additional challenging multiple-choice questions for evaluating language models across diverse academic and professional domains. Built on the foundation of the Massive Multitask Language Understanding benchmark framework.", + "metric_id": "llm_stats.mmlu-prox.score", + "metric_name": "MMLU-ProX score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu-prox", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU-ProX", + "raw_categories": "[\"language\",\"legal\",\"math\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "32" + } + }, + "score_details": { + "score": 0.81, + "details": { + "raw_score": "0.81", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "mmlu-prox", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/mmlu-prox\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu-prox\"]", + "raw_score_id": "mmlu-prox::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu-redux::mmlu-redux-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.mmlu-redux", + "source_data": { + "dataset_name": "MMLU-Redux", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/mmlu-redux", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu-redux" + ], + "additional_details": { + "raw_benchmark_id": "mmlu-redux", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "An improved version of the MMLU benchmark featuring manually re-annotated questions to identify and correct errors in the original dataset. Provides more reliable evaluation metrics for language models by addressing dataset quality issues found in the original MMLU.", + "metric_id": "llm_stats.mmlu-redux.score", + "metric_name": "MMLU-Redux score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu-redux", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU-Redux", + "raw_categories": "[\"language\",\"math\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "48" + } + }, + "score_details": { + "score": 0.933, + "details": { + "raw_score": "0.933", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "mmlu-redux", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/mmlu-redux\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu-redux\"]", + "raw_score_id": "mmlu-redux::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmmu::mmmu-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.mmmu", + "source_data": { + "dataset_name": "MMMU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/mmmu", + "https://api.llm-stats.com/leaderboard/benchmarks/mmmu" + ], + "additional_details": { + "raw_benchmark_id": "mmmu", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MMMU (Massive Multi-discipline Multimodal Understanding) is a benchmark designed to evaluate multimodal models on college-level subject knowledge and deliberate reasoning. Contains 11.5K meticulously collected multimodal questions from college exams, quizzes, and textbooks, covering six core disciplines: Art & Design, Business, Science, Health & Medicine, Humanities & Social Science, and Tech & Engineering across 30 subjects and 183 subfields.", + "metric_id": "llm_stats.mmmu.score", + "metric_name": "MMMU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmmu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMMU", + "raw_categories": "[\"multimodal\",\"reasoning\",\"general\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "63" + } + }, + "score_details": { + "score": 0.814, + "details": { + "raw_score": "0.814", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "mmmu", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/mmmu\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmmu\"]", + "raw_score_id": "mmmu::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmstar::mmstar-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.mmstar", + "source_data": { + "dataset_name": "MMStar", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/mmstar", + "https://api.llm-stats.com/leaderboard/benchmarks/mmstar" + ], + "additional_details": { + "raw_benchmark_id": "mmstar", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MMStar is an elite vision-indispensable multimodal benchmark comprising 1,500 challenge samples meticulously selected by humans to evaluate 6 core capabilities and 18 detailed axes. The benchmark addresses issues of visual content unnecessity and unintentional data leakage in existing multimodal evaluations.", + "metric_id": "llm_stats.mmstar.score", + "metric_name": "MMStar score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmstar", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMStar", + "raw_categories": "[\"multimodal\",\"reasoning\",\"general\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "22" + } + }, + "score_details": { + "score": 0.819, + "details": { + "raw_score": "0.819", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "mmstar", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/mmstar\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmstar\"]", + "raw_score_id": "mmstar::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmvu::mmvu-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.mmvu", + "source_data": { + "dataset_name": "MMVU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/mmvu", + "https://api.llm-stats.com/leaderboard/benchmarks/mmvu" + ], + "additional_details": { + "raw_benchmark_id": "mmvu", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MMVU (Multimodal Multi-disciplinary Video Understanding) is a benchmark for evaluating multimodal models on video understanding tasks across multiple disciplines, testing comprehension and reasoning capabilities on video content.", + "metric_id": "llm_stats.mmvu.score", + "metric_name": "MMVU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmvu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMVU", + "raw_categories": "[\"multimodal\",\"reasoning\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.723, + "details": { + "raw_score": "0.723", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "mmvu", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/mmvu\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmvu\"]", + "raw_score_id": "mmvu::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "multichallenge::multichallenge-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.multichallenge", + "source_data": { + "dataset_name": "Multi-Challenge", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/multichallenge", + "https://api.llm-stats.com/leaderboard/benchmarks/multichallenge" + ], + "additional_details": { + "raw_benchmark_id": "multichallenge", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MultiChallenge is a realistic multi-turn conversation evaluation benchmark that challenges frontier LLMs across four key categories: instruction retention (maintaining instructions throughout conversations), inference memory (recalling and connecting details from previous turns), reliable versioned editing (adapting to evolving instructions during collaborative editing), and self-coherence (avoiding contradictions in responses). The benchmark evaluates models on sustained, contextually complex dialogues across diverse topics including travel planning, technical documentation, and professional communication.", + "metric_id": "llm_stats.multichallenge.score", + "metric_name": "Multi-Challenge score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "multichallenge", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Multi-Challenge", + "raw_categories": "[\"reasoning\",\"communication\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "29" + } + }, + "score_details": { + "score": 0.6, + "details": { + "raw_score": "0.6", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "multichallenge", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/multichallenge\",\"https://api.llm-stats.com/leaderboard/benchmarks/multichallenge\"]", + "raw_score_id": "multichallenge::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mvbench::mvbench-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.mvbench", + "source_data": { + "dataset_name": "MVBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/mvbench", + "https://api.llm-stats.com/leaderboard/benchmarks/mvbench" + ], + "additional_details": { + "raw_benchmark_id": "mvbench", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A comprehensive multi-modal video understanding benchmark covering 20 challenging video tasks that require temporal understanding beyond single-frame analysis. Tasks span from perception to cognition, including action recognition, temporal reasoning, spatial reasoning, object interaction, scene transition, and counterfactual inference. Uses a novel static-to-dynamic method to systematically generate video tasks from existing annotations.", + "metric_id": "llm_stats.mvbench.score", + "metric_name": "MVBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mvbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MVBench", + "raw_categories": "[\"multimodal\",\"reasoning\",\"spatial_reasoning\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "17" + } + }, + "score_details": { + "score": 0.748, + "details": { + "raw_score": "0.748", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "mvbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/mvbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/mvbench\"]", + "raw_score_id": "mvbench::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "nova-63::nova-63-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.nova-63", + "source_data": { + "dataset_name": "NOVA-63", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/nova-63", + "https://api.llm-stats.com/leaderboard/benchmarks/nova-63" + ], + "additional_details": { + "raw_benchmark_id": "nova-63", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "NOVA-63 is a multilingual evaluation benchmark covering 63 languages, designed to assess LLM performance across diverse linguistic contexts and tasks.", + "metric_id": "llm_stats.nova-63.score", + "metric_name": "NOVA-63 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "nova-63", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "NOVA-63", + "raw_categories": "[\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.571, + "details": { + "raw_score": "0.571", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "nova-63", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/nova-63\",\"https://api.llm-stats.com/leaderboard/benchmarks/nova-63\"]", + "raw_score_id": "nova-63::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "nuscene::nuscene-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.nuscene", + "source_data": { + "dataset_name": "Nuscene", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/nuscene", + "https://api.llm-stats.com/leaderboard/benchmarks/nuscene" + ], + "additional_details": { + "raw_benchmark_id": "nuscene", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multimodal benchmark for scene understanding and reasoning over the nuScenes autonomous driving domain.", + "metric_id": "llm_stats.nuscene.score", + "metric_name": "Nuscene score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "nuscene", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Nuscene", + "raw_categories": "[\"multimodal\",\"reasoning\",\"spatial\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.146, + "details": { + "raw_score": "0.146", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "nuscene", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/nuscene\",\"https://api.llm-stats.com/leaderboard/benchmarks/nuscene\"]", + "raw_score_id": "nuscene::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ocrbench::ocrbench-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.ocrbench", + "source_data": { + "dataset_name": "OCRBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/ocrbench", + "https://api.llm-stats.com/leaderboard/benchmarks/ocrbench" + ], + "additional_details": { + "raw_benchmark_id": "ocrbench", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OCRBench: Comprehensive evaluation benchmark for assessing Optical Character Recognition (OCR) capabilities in Large Multimodal Models across text recognition, scene text VQA, and document understanding tasks", + "metric_id": "llm_stats.ocrbench.score", + "metric_name": "OCRBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ocrbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OCRBench", + "raw_categories": "[\"image_to_text\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "22" + } + }, + "score_details": { + "score": 0.91, + "details": { + "raw_score": "0.91", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "ocrbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/ocrbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/ocrbench\"]", + "raw_score_id": "ocrbench::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "odinw::odinw-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.odinw", + "source_data": { + "dataset_name": "ODinW", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/odinw", + "https://api.llm-stats.com/leaderboard/benchmarks/odinw" + ], + "additional_details": { + "raw_benchmark_id": "odinw", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Object Detection in the Wild (ODinW) benchmark for evaluating object detection models' task-level transfer ability across diverse real-world datasets in terms of prediction accuracy and adaptation efficiency", + "metric_id": "llm_stats.odinw.score", + "metric_name": "ODinW score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "odinw", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ODinW", + "raw_categories": "[\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.426, + "details": { + "raw_score": "0.426", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "odinw", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/odinw\",\"https://api.llm-stats.com/leaderboard/benchmarks/odinw\"]", + "raw_score_id": "odinw::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ojbench::ojbench-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.ojbench", + "source_data": { + "dataset_name": "OJBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/ojbench", + "https://api.llm-stats.com/leaderboard/benchmarks/ojbench" + ], + "additional_details": { + "raw_benchmark_id": "ojbench", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OJBench is a competition-level code benchmark designed to assess the competitive-level code reasoning abilities of large language models. It comprises 232 programming competition problems from NOI and ICPC, categorized into Easy, Medium, and Hard difficulty levels. The benchmark evaluates models' ability to solve complex competitive programming challenges using Python and C++.", + "metric_id": "llm_stats.ojbench.score", + "metric_name": "OJBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ojbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OJBench", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.36, + "details": { + "raw_score": "0.36", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "ojbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/ojbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/ojbench\"]", + "raw_score_id": "ojbench::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "omnidocbench-1.5::omnidocbench-1.5-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.omnidocbench-1.5", + "source_data": { + "dataset_name": "OmniDocBench 1.5", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/omnidocbench-1.5", + "https://api.llm-stats.com/leaderboard/benchmarks/omnidocbench-1.5" + ], + "additional_details": { + "raw_benchmark_id": "omnidocbench-1.5", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OmniDocBench 1.5 is a comprehensive benchmark for evaluating multimodal large language models on document understanding tasks, including OCR, document parsing, information extraction, and visual question answering across diverse document types. Lower Overall Edit Distance scores are better.", + "metric_id": "llm_stats.omnidocbench-1.5.score", + "metric_name": "OmniDocBench 1.5 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "omnidocbench-1.5", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OmniDocBench 1.5", + "raw_categories": "[\"multimodal\",\"reasoning\",\"structured_output\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.893, + "details": { + "raw_score": "0.893", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "omnidocbench-1.5", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/omnidocbench-1.5\",\"https://api.llm-stats.com/leaderboard/benchmarks/omnidocbench-1.5\"]", + "raw_score_id": "omnidocbench-1.5::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "osworld-verified::osworld-verified-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.osworld-verified", + "source_data": { + "dataset_name": "OSWorld-Verified", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/osworld-verified", + "https://api.llm-stats.com/leaderboard/benchmarks/osworld-verified" + ], + "additional_details": { + "raw_benchmark_id": "osworld-verified", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OSWorld-Verified is a verified subset of OSWorld, a scalable real computer environment for multimodal agents supporting task setup, execution-based evaluation, and interactive learning across Ubuntu, Windows, and macOS.", + "metric_id": "llm_stats.osworld-verified.score", + "metric_name": "OSWorld-Verified score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "osworld-verified", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OSWorld-Verified", + "raw_categories": "[\"multimodal\",\"general\",\"agents\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.545, + "details": { + "raw_score": "0.545", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "osworld-verified", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/osworld-verified\",\"https://api.llm-stats.com/leaderboard/benchmarks/osworld-verified\"]", + "raw_score_id": "osworld-verified::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "pmc-vqa::pmc-vqa-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.pmc-vqa", + "source_data": { + "dataset_name": "PMC-VQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/pmc-vqa", + "https://api.llm-stats.com/leaderboard/benchmarks/pmc-vqa" + ], + "additional_details": { + "raw_benchmark_id": "pmc-vqa", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A medical visual question answering benchmark built on biomedical literature and medical figures.", + "metric_id": "llm_stats.pmc-vqa.score", + "metric_name": "PMC-VQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "pmc-vqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PMC-VQA", + "raw_categories": "[\"multimodal\",\"reasoning\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.62, + "details": { + "raw_score": "0.62", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "pmc-vqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/pmc-vqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/pmc-vqa\"]", + "raw_score_id": "pmc-vqa::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "polymath::polymath-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.polymath", + "source_data": { + "dataset_name": "PolyMATH", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/polymath", + "https://api.llm-stats.com/leaderboard/benchmarks/polymath" + ], + "additional_details": { + "raw_benchmark_id": "polymath", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Polymath is a challenging multi-modal mathematical reasoning benchmark designed to evaluate the general cognitive reasoning abilities of Multi-modal Large Language Models (MLLMs). The benchmark comprises 5,000 manually collected high-quality images of cognitive textual and visual challenges across 10 distinct categories, including pattern recognition, spatial reasoning, and relative reasoning.", + "metric_id": "llm_stats.polymath.score", + "metric_name": "PolyMATH score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "polymath", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PolyMATH", + "raw_categories": "[\"math\",\"multimodal\",\"reasoning\",\"spatial_reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.644, + "details": { + "raw_score": "0.644", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "polymath", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/polymath\",\"https://api.llm-stats.com/leaderboard/benchmarks/polymath\"]", + "raw_score_id": "polymath::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "realworldqa::realworldqa-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.realworldqa", + "source_data": { + "dataset_name": "RealWorldQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/realworldqa", + "https://api.llm-stats.com/leaderboard/benchmarks/realworldqa" + ], + "additional_details": { + "raw_benchmark_id": "realworldqa", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "RealWorldQA is a benchmark designed to evaluate basic real-world spatial understanding capabilities of multimodal models. The initial release consists of over 700 anonymized images taken from vehicles and other real-world scenarios, each accompanied by a question and easily verifiable answer. Released by xAI as part of their Grok-1.5 Vision preview to test models' ability to understand natural scenes and spatial relationships in everyday visual contexts.", + "metric_id": "llm_stats.realworldqa.score", + "metric_name": "RealWorldQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "realworldqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "RealWorldQA", + "raw_categories": "[\"spatial_reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "25" + } + }, + "score_details": { + "score": 0.841, + "details": { + "raw_score": "0.841", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "realworldqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/realworldqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/realworldqa\"]", + "raw_score_id": "realworldqa::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "refcoco-avg::refcoco-avg-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.refcoco-avg", + "source_data": { + "dataset_name": "RefCOCO-avg", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/refcoco-avg", + "https://api.llm-stats.com/leaderboard/benchmarks/refcoco-avg" + ], + "additional_details": { + "raw_benchmark_id": "refcoco-avg", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "RefCOCO-avg measures object grounding accuracy averaged across RefCOCO, RefCOCO+, and RefCOCOg benchmarks.", + "metric_id": "llm_stats.refcoco-avg.score", + "metric_name": "RefCOCO-avg score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "refcoco-avg", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "RefCOCO-avg", + "raw_categories": "[\"spatial_reasoning\",\"grounding\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.892, + "details": { + "raw_score": "0.892", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "refcoco-avg", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/refcoco-avg\",\"https://api.llm-stats.com/leaderboard/benchmarks/refcoco-avg\"]", + "raw_score_id": "refcoco-avg::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "refspatialbench::refspatialbench-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.refspatialbench", + "source_data": { + "dataset_name": "RefSpatialBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/refspatialbench", + "https://api.llm-stats.com/leaderboard/benchmarks/refspatialbench" + ], + "additional_details": { + "raw_benchmark_id": "refspatialbench", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "RefSpatialBench evaluates spatial reference understanding and grounding.", + "metric_id": "llm_stats.refspatialbench.score", + "metric_name": "RefSpatialBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "refspatialbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "RefSpatialBench", + "raw_categories": "[\"spatial_reasoning\",\"grounding\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.635, + "details": { + "raw_score": "0.635", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "refspatialbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/refspatialbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/refspatialbench\"]", + "raw_score_id": "refspatialbench::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "screenspot-pro::screenspot-pro-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.screenspot-pro", + "source_data": { + "dataset_name": "ScreenSpot Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/screenspot-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/screenspot-pro" + ], + "additional_details": { + "raw_benchmark_id": "screenspot-pro", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ScreenSpot-Pro is a novel GUI grounding benchmark designed to rigorously evaluate the grounding capabilities of multimodal large language models (MLLMs) in professional high-resolution computing environments. The benchmark comprises 1,581 instructions across 23 applications spanning 5 industries and 3 operating systems, featuring authentic high-resolution images from professional domains with expert annotations. Unlike previous benchmarks that focus on cropped screenshots in consumer applications, ScreenSpot-Pro addresses the complexity and diversity of real-world professional software scenarios, revealing significant performance gaps in current MLLM GUI perception capabilities.", + "metric_id": "llm_stats.screenspot-pro.score", + "metric_name": "ScreenSpot Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "screenspot-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ScreenSpot Pro", + "raw_categories": "[\"multimodal\",\"spatial_reasoning\",\"grounding\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.686, + "details": { + "raw_score": "0.686", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "screenspot-pro", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/screenspot-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/screenspot-pro\"]", + "raw_score_id": "screenspot-pro::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "seal-0::seal-0-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.seal-0", + "source_data": { + "dataset_name": "Seal-0", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/seal-0", + "https://api.llm-stats.com/leaderboard/benchmarks/seal-0" + ], + "additional_details": { + "raw_benchmark_id": "seal-0", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Seal-0 is a benchmark for evaluating agentic search capabilities, testing models' ability to navigate and retrieve information using tools.", + "metric_id": "llm_stats.seal-0.score", + "metric_name": "Seal-0 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "seal-0", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Seal-0", + "raw_categories": "[\"reasoning\",\"search\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.414, + "details": { + "raw_score": "0.414", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "seal-0", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/seal-0\",\"https://api.llm-stats.com/leaderboard/benchmarks/seal-0\"]", + "raw_score_id": "seal-0::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "simplevqa::simplevqa-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.simplevqa", + "source_data": { + "dataset_name": "SimpleVQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/simplevqa", + "https://api.llm-stats.com/leaderboard/benchmarks/simplevqa" + ], + "additional_details": { + "raw_benchmark_id": "simplevqa", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SimpleVQA is a visual question answering benchmark focused on simple queries.", + "metric_id": "llm_stats.simplevqa.score", + "metric_name": "SimpleVQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "simplevqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SimpleVQA", + "raw_categories": "[\"image_to_text\",\"multimodal\",\"general\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.583, + "details": { + "raw_score": "0.583", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "simplevqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/simplevqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/simplevqa\"]", + "raw_score_id": "simplevqa::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "slakevqa::slakevqa-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.slakevqa", + "source_data": { + "dataset_name": "SlakeVQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/slakevqa", + "https://api.llm-stats.com/leaderboard/benchmarks/slakevqa" + ], + "additional_details": { + "raw_benchmark_id": "slakevqa", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A semantically-labeled knowledge-enhanced dataset for medical visual question answering. Contains 642 radiology images (CT scans, MRI scans, X-rays) covering five body parts and 14,028 bilingual English-Chinese question-answer pairs annotated by experienced physicians. Features comprehensive semantic labels and a structural medical knowledge base with both vision-only and knowledge-based questions requiring external medical knowledge reasoning.", + "metric_id": "llm_stats.slakevqa.score", + "metric_name": "SlakeVQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "slakevqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SlakeVQA", + "raw_categories": "[\"image_to_text\",\"multimodal\",\"reasoning\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.787, + "details": { + "raw_score": "0.787", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "slakevqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/slakevqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/slakevqa\"]", + "raw_score_id": "slakevqa::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "sunrgbd::sunrgbd-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.sunrgbd", + "source_data": { + "dataset_name": "SUNRGBD", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/sunrgbd", + "https://api.llm-stats.com/leaderboard/benchmarks/sunrgbd" + ], + "additional_details": { + "raw_benchmark_id": "sunrgbd", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SUNRGBD evaluates RGB-D scene understanding and 3D grounding capabilities.", + "metric_id": "llm_stats.sunrgbd.score", + "metric_name": "SUNRGBD score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "sunrgbd", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SUNRGBD", + "raw_categories": "[\"spatial_reasoning\",\"3d\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.334, + "details": { + "raw_score": "0.334", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "sunrgbd", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/sunrgbd\",\"https://api.llm-stats.com/leaderboard/benchmarks/sunrgbd\"]", + "raw_score_id": "sunrgbd::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "supergpqa::supergpqa-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.supergpqa", + "source_data": { + "dataset_name": "SuperGPQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/supergpqa", + "https://api.llm-stats.com/leaderboard/benchmarks/supergpqa" + ], + "additional_details": { + "raw_benchmark_id": "supergpqa", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SuperGPQA is a comprehensive benchmark that evaluates large language models across 285 graduate-level academic disciplines. The benchmark contains 25,957 questions covering 13 broad disciplinary areas including Engineering, Medicine, Science, and Law, with specialized fields in light industry, agriculture, and service-oriented domains. It employs a Human-LLM collaborative filtering mechanism with over 80 expert annotators to create challenging questions that assess graduate-level knowledge and reasoning capabilities.", + "metric_id": "llm_stats.supergpqa.score", + "metric_name": "SuperGPQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "supergpqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SuperGPQA", + "raw_categories": "[\"legal\",\"math\",\"physics\",\"reasoning\",\"finance\",\"general\",\"healthcare\",\"chemistry\",\"economics\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "34" + } + }, + "score_details": { + "score": 0.634, + "details": { + "raw_score": "0.634", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "supergpqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/supergpqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/supergpqa\"]", + "raw_score_id": "supergpqa::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "t2-bench::t2-bench-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.t2-bench", + "source_data": { + "dataset_name": "t2-bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/t2-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/t2-bench" + ], + "additional_details": { + "raw_benchmark_id": "t2-bench", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "t2-bench is a benchmark for evaluating agentic tool use capabilities, measuring how well models can select, sequence, and utilize tools to solve complex tasks. It tests autonomous planning and execution in multi-step scenarios.", + "metric_id": "llm_stats.t2-bench.score", + "metric_name": "t2-bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "t2-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "t2-bench", + "raw_categories": "[\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.812, + "details": { + "raw_score": "0.812", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "t2-bench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/t2-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/t2-bench\"]", + "raw_score_id": "t2-bench::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "tir-bench::tir-bench-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.tir-bench", + "source_data": { + "dataset_name": "TIR-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/tir-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/tir-bench" + ], + "additional_details": { + "raw_benchmark_id": "tir-bench", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A tool-calling and multimodal interaction benchmark for testing visual instruction following and execution reliability.", + "metric_id": "llm_stats.tir-bench.score", + "metric_name": "TIR-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "tir-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TIR-Bench", + "raw_categories": "[\"multimodal\",\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.555, + "details": { + "raw_score": "0.555", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "tir-bench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/tir-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/tir-bench\"]", + "raw_score_id": "tir-bench::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "v-star::v-star-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.v-star", + "source_data": { + "dataset_name": "V*", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/v-star", + "https://api.llm-stats.com/leaderboard/benchmarks/v-star" + ], + "additional_details": { + "raw_benchmark_id": "v-star", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A visual reasoning benchmark evaluating multimodal inference under challenging spatial and grounded tasks.", + "metric_id": "llm_stats.v-star.score", + "metric_name": "V* score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "v-star", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "V*", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.927, + "details": { + "raw_score": "0.927", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "v-star", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/v-star\",\"https://api.llm-stats.com/leaderboard/benchmarks/v-star\"]", + "raw_score_id": "v-star::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "videomme-w-o-sub.::videomme-w-o-sub.-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.videomme-w-o-sub.", + "source_data": { + "dataset_name": "VideoMME w/o sub.", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/videomme-w-o-sub.", + "https://api.llm-stats.com/leaderboard/benchmarks/videomme-w-o-sub." + ], + "additional_details": { + "raw_benchmark_id": "videomme-w-o-sub.", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Video-MME is a comprehensive evaluation benchmark for multi-modal large language models in video analysis. It features 900 videos across 6 primary visual domains with 30 subfields, ranging from 11 seconds to 1 hour in duration, with 2,700 question-answer pairs. The benchmark evaluates MLLMs' capabilities in processing sequential visual data and multi-modal content including video frames, subtitles, and audio.", + "metric_id": "llm_stats.videomme-w-o-sub..score", + "metric_name": "VideoMME w/o sub. score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "videomme-w-o-sub.", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VideoMME w/o sub.", + "raw_categories": "[\"multimodal\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.825, + "details": { + "raw_score": "0.825", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "videomme-w-o-sub.", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/videomme-w-o-sub.\",\"https://api.llm-stats.com/leaderboard/benchmarks/videomme-w-o-sub.\"]", + "raw_score_id": "videomme-w-o-sub.::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "videomme-w-sub.::videomme-w-sub.-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.videomme-w-sub.", + "source_data": { + "dataset_name": "VideoMME w sub.", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/videomme-w-sub.", + "https://api.llm-stats.com/leaderboard/benchmarks/videomme-w-sub." + ], + "additional_details": { + "raw_benchmark_id": "videomme-w-sub.", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The first-ever comprehensive evaluation benchmark of Multi-modal LLMs in Video analysis. Features 900 videos (254 hours) with 2,700 question-answer pairs covering 6 primary visual domains and 30 subfields. Evaluates temporal understanding across short (11 seconds) to long (1 hour) videos with multi-modal inputs including video frames, subtitles, and audio.", + "metric_id": "llm_stats.videomme-w-sub..score", + "metric_name": "VideoMME w sub. score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "videomme-w-sub.", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VideoMME w sub.", + "raw_categories": "[\"multimodal\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.866, + "details": { + "raw_score": "0.866", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "videomme-w-sub.", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/videomme-w-sub.\",\"https://api.llm-stats.com/leaderboard/benchmarks/videomme-w-sub.\"]", + "raw_score_id": "videomme-w-sub.::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "videommmu::videommmu-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.videommmu", + "source_data": { + "dataset_name": "VideoMMMU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/videommmu", + "https://api.llm-stats.com/leaderboard/benchmarks/videommmu" + ], + "additional_details": { + "raw_benchmark_id": "videommmu", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Video-MMMU evaluates Large Multimodal Models' ability to acquire knowledge from expert-level professional videos across six disciplines through three cognitive stages: perception, comprehension, and adaptation. Contains 300 videos and 900 human-annotated questions spanning Art, Business, Science, Medicine, Humanities, and Engineering.", + "metric_id": "llm_stats.videommmu.score", + "metric_name": "VideoMMMU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "videommmu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VideoMMMU", + "raw_categories": "[\"multimodal\",\"reasoning\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "26" + } + }, + "score_details": { + "score": 0.804, + "details": { + "raw_score": "0.804", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "videommmu", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/videommmu\",\"https://api.llm-stats.com/leaderboard/benchmarks/videommmu\"]", + "raw_score_id": "videommmu::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "vita-bench::vita-bench-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.vita-bench", + "source_data": { + "dataset_name": "VITA-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/vita-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/vita-bench" + ], + "additional_details": { + "raw_benchmark_id": "vita-bench", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VITA-Bench evaluates AI agents on real-world virtual task automation, measuring their ability to complete complex multi-step tasks in simulated environments.", + "metric_id": "llm_stats.vita-bench.score", + "metric_name": "VITA-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "vita-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VITA-Bench", + "raw_categories": "[\"reasoning\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.319, + "details": { + "raw_score": "0.319", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "vita-bench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/vita-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/vita-bench\"]", + "raw_score_id": "vita-bench::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "vlmsareblind::vlmsareblind-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.vlmsareblind", + "source_data": { + "dataset_name": "VLMsAreBlind", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/vlmsareblind", + "https://api.llm-stats.com/leaderboard/benchmarks/vlmsareblind" + ], + "additional_details": { + "raw_benchmark_id": "vlmsareblind", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A vision-language benchmark that probes blind spots and brittle reasoning in multimodal models.", + "metric_id": "llm_stats.vlmsareblind.score", + "metric_name": "VLMsAreBlind score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "vlmsareblind", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VLMsAreBlind", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.97, + "details": { + "raw_score": "0.97", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "vlmsareblind", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/vlmsareblind\",\"https://api.llm-stats.com/leaderboard/benchmarks/vlmsareblind\"]", + "raw_score_id": "vlmsareblind::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "widesearch::widesearch-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.widesearch", + "source_data": { + "dataset_name": "WideSearch", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/widesearch", + "https://api.llm-stats.com/leaderboard/benchmarks/widesearch" + ], + "additional_details": { + "raw_benchmark_id": "widesearch", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WideSearch is an agentic search benchmark that evaluates models' ability to perform broad, parallel search operations across multiple sources. It tests wide-coverage information retrieval and synthesis capabilities.", + "metric_id": "llm_stats.widesearch.score", + "metric_name": "WideSearch score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "widesearch", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WideSearch", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.571, + "details": { + "raw_score": "0.571", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "widesearch", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/widesearch\",\"https://api.llm-stats.com/leaderboard/benchmarks/widesearch\"]", + "raw_score_id": "widesearch::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "wmt24::wmt24-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.wmt24", + "source_data": { + "dataset_name": "WMT24++", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/wmt24", + "https://api.llm-stats.com/leaderboard/benchmarks/wmt24++" + ], + "additional_details": { + "raw_benchmark_id": "wmt24++", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WMT24++ is a comprehensive multilingual machine translation benchmark that expands the WMT24 dataset to cover 55 languages and dialects. It includes human-written references and post-edits across four domains (literary, news, social, and speech) to evaluate machine translation systems and large language models across diverse linguistic contexts.", + "metric_id": "llm_stats.wmt24.score", + "metric_name": "WMT24++ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "wmt24++", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WMT24++", + "raw_categories": "[\"language\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.763, + "details": { + "raw_score": "0.763", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "wmt24++", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/wmt24\",\"https://api.llm-stats.com/leaderboard/benchmarks/wmt24++\"]", + "raw_score_id": "wmt24++::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "zerobench-sub::zerobench-sub-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.zerobench-sub", + "source_data": { + "dataset_name": "ZEROBench-Sub", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/zerobench-sub", + "https://api.llm-stats.com/leaderboard/benchmarks/zerobench-sub" + ], + "additional_details": { + "raw_benchmark_id": "zerobench-sub", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ZEROBench-Sub is a subset of the ZEROBench benchmark.", + "metric_id": "llm_stats.zerobench-sub.score", + "metric_name": "ZEROBench-Sub score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "zerobench-sub", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ZEROBench-Sub", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.341, + "details": { + "raw_score": "0.341", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "zerobench-sub", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/zerobench-sub\",\"https://api.llm-stats.com/leaderboard/benchmarks/zerobench-sub\"]", + "raw_score_id": "zerobench-sub::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "zerobench::zerobench-qwen3.5-35b-a3b", + "evaluation_name": "llm_stats.zerobench", + "source_data": { + "dataset_name": "ZEROBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-35b-a3b", + "https://llm-stats.com/benchmarks/zerobench", + "https://api.llm-stats.com/leaderboard/benchmarks/zerobench" + ], + "additional_details": { + "raw_benchmark_id": "zerobench", + "raw_model_id": "qwen3.5-35b-a3b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ZEROBench is a challenging vision benchmark designed to test models on zero-shot visual understanding tasks.", + "metric_id": "llm_stats.zerobench.score", + "metric_name": "ZEROBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "zerobench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ZEROBench", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.08, + "details": { + "raw_score": "0.08", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-35b-a3b", + "raw_benchmark_id": "zerobench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-35b-a3b\",\"https://llm-stats.com/benchmarks/zerobench\",\"https://api.llm-stats.com/leaderboard/benchmarks/zerobench\"]", + "raw_score_id": "zerobench::qwen3.5-35b-a3b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/qwen/qwen3.5-397b-a17b/cded27a4-5568-4bf6-85ad-46693ce31f9b.json b/data/llm_stats/qwen/qwen3.5-397b-a17b/cded27a4-5568-4bf6-85ad-46693ce31f9b.json new file mode 100644 index 0000000000000000000000000000000000000000..4639449ce2c67c57919953c67c81888bce305b7f --- /dev/null +++ b/data/llm_stats/qwen/qwen3.5-397b-a17b/cded27a4-5568-4bf6-85ad-46693ce31f9b.json @@ -0,0 +1,1763 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/qwen_qwen3.5-397b-a17b/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Qwen3.5-397B-A17B", + "id": "qwen/qwen3.5-397b-a17b", + "developer": "qwen", + "additional_details": { + "raw_id": "qwen3.5-397b-a17b", + "raw_name": "Qwen3.5-397B-A17B", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_model_name": "Qwen3.5-397B-A17B", + "raw_organization_id": "qwen", + "raw_organization_name": "Alibaba Cloud / Qwen Team", + "raw_release_date": "2026-02-16", + "raw_announcement_date": "2026-02-16", + "raw_multimodal": "true", + "raw_param_count": "397000000000", + "raw_provider_slug": "qwen", + "raw_provider_name": "Alibaba Cloud / Qwen Team" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aa-lcr::aa-lcr-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.aa-lcr", + "source_data": { + "dataset_name": "AA-LCR", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/aa-lcr", + "https://api.llm-stats.com/leaderboard/benchmarks/aa-lcr" + ], + "additional_details": { + "raw_benchmark_id": "aa-lcr", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Agent Arena Long Context Reasoning benchmark", + "metric_id": "llm_stats.aa-lcr.score", + "metric_name": "AA-LCR score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aa-lcr", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AA-LCR", + "raw_categories": "[\"long_context\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "14" + } + }, + "score_details": { + "score": 0.687, + "details": { + "raw_score": "0.687", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "aa-lcr", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/aa-lcr\",\"https://api.llm-stats.com/leaderboard/benchmarks/aa-lcr\"]", + "raw_score_id": "aa-lcr::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "aime-2026::aime-2026-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.aime-2026", + "source_data": { + "dataset_name": "AIME 2026", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/aime-2026", + "https://api.llm-stats.com/leaderboard/benchmarks/aime-2026" + ], + "additional_details": { + "raw_benchmark_id": "aime-2026", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "All 30 problems from the 2026 American Invitational Mathematics Examination (AIME I and AIME II), testing olympiad-level mathematical reasoning with integer answers from 000-999. Used as an AI benchmark to evaluate large language models' ability to solve complex mathematical problems requiring multi-step logical deductions and structured symbolic reasoning.", + "metric_id": "llm_stats.aime-2026.score", + "metric_name": "AIME 2026 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aime-2026", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AIME 2026", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "17" + } + }, + "score_details": { + "score": 0.913, + "details": { + "raw_score": "0.913", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "aime-2026", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/aime-2026\",\"https://api.llm-stats.com/leaderboard/benchmarks/aime-2026\"]", + "raw_score_id": "aime-2026::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bfcl-v4::bfcl-v4-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.bfcl-v4", + "source_data": { + "dataset_name": "BFCL-V4", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/bfcl-v4", + "https://api.llm-stats.com/leaderboard/benchmarks/bfcl-v4" + ], + "additional_details": { + "raw_benchmark_id": "bfcl-v4", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Berkeley Function Calling Leaderboard V4 (BFCL-V4) evaluates LLMs on their ability to accurately call functions and APIs, including simple, multiple, parallel, and nested function calls across diverse programming scenarios.", + "metric_id": "llm_stats.bfcl-v4.score", + "metric_name": "BFCL-V4 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bfcl-v4", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BFCL-V4", + "raw_categories": "[\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.729, + "details": { + "raw_score": "0.729", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "bfcl-v4", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/bfcl-v4\",\"https://api.llm-stats.com/leaderboard/benchmarks/bfcl-v4\"]", + "raw_score_id": "bfcl-v4::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "browsecomp-zh::browsecomp-zh-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.browsecomp-zh", + "source_data": { + "dataset_name": "BrowseComp-zh", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/browsecomp-zh", + "https://api.llm-stats.com/leaderboard/benchmarks/browsecomp-zh" + ], + "additional_details": { + "raw_benchmark_id": "browsecomp-zh", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A high-difficulty benchmark purpose-built to comprehensively evaluate LLM agents on the Chinese web, consisting of 289 multi-hop questions spanning 11 diverse domains including Film & TV, Technology, Medicine, and History. Questions are reverse-engineered from short, objective, and easily verifiable answers, requiring sophisticated reasoning and information reconciliation beyond basic retrieval. The benchmark addresses linguistic, infrastructural, and censorship-related complexities in Chinese web environments.", + "metric_id": "llm_stats.browsecomp-zh.score", + "metric_name": "BrowseComp-zh score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "browsecomp-zh", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BrowseComp-zh", + "raw_categories": "[\"reasoning\",\"search\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.703, + "details": { + "raw_score": "0.703", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "browsecomp-zh", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/browsecomp-zh\",\"https://api.llm-stats.com/leaderboard/benchmarks/browsecomp-zh\"]", + "raw_score_id": "browsecomp-zh::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "c-eval::c-eval-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.c-eval", + "source_data": { + "dataset_name": "C-Eval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/c-eval", + "https://api.llm-stats.com/leaderboard/benchmarks/c-eval" + ], + "additional_details": { + "raw_benchmark_id": "c-eval", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "C-Eval is a comprehensive Chinese evaluation suite designed to assess advanced knowledge and reasoning abilities of foundation models in a Chinese context. It comprises 13,948 multiple-choice questions across 52 diverse disciplines spanning humanities, science, and engineering, with four difficulty levels: middle school, high school, college, and professional. The benchmark includes C-Eval Hard, a subset of very challenging subjects requiring advanced reasoning abilities.", + "metric_id": "llm_stats.c-eval.score", + "metric_name": "C-Eval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "c-eval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "C-Eval", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.93, + "details": { + "raw_score": "0.93", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "c-eval", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/c-eval\",\"https://api.llm-stats.com/leaderboard/benchmarks/c-eval\"]", + "raw_score_id": "c-eval::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "deep-planning::deep-planning-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.deep-planning", + "source_data": { + "dataset_name": "DeepPlanning", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/deep-planning", + "https://api.llm-stats.com/leaderboard/benchmarks/deep-planning" + ], + "additional_details": { + "raw_benchmark_id": "deep-planning", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "DeepPlanning evaluates LLMs on complex multi-step planning tasks requiring long-horizon reasoning, goal decomposition, and strategic decision-making.", + "metric_id": "llm_stats.deep-planning.score", + "metric_name": "DeepPlanning score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "deep-planning", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DeepPlanning", + "raw_categories": "[\"reasoning\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.343, + "details": { + "raw_score": "0.343", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "deep-planning", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/deep-planning\",\"https://api.llm-stats.com/leaderboard/benchmarks/deep-planning\"]", + "raw_score_id": "deep-planning::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "global-piqa::global-piqa-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.global-piqa", + "source_data": { + "dataset_name": "Global PIQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/global-piqa", + "https://api.llm-stats.com/leaderboard/benchmarks/global-piqa" + ], + "additional_details": { + "raw_benchmark_id": "global-piqa", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Global PIQA is a multilingual commonsense reasoning benchmark that evaluates physical interaction knowledge across 100 languages and cultures. It tests AI systems' understanding of physical world knowledge in diverse cultural contexts through multiple choice questions about everyday situations requiring physical commonsense.", + "metric_id": "llm_stats.global-piqa.score", + "metric_name": "Global PIQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "global-piqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Global PIQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.898, + "details": { + "raw_score": "0.898", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "global-piqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/global-piqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/global-piqa\"]", + "raw_score_id": "global-piqa::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hmmt-2025::hmmt-2025-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.hmmt-2025", + "source_data": { + "dataset_name": "HMMT 2025", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/hmmt-2025", + "https://api.llm-stats.com/leaderboard/benchmarks/hmmt-2025" + ], + "additional_details": { + "raw_benchmark_id": "hmmt-2025", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Harvard-MIT Mathematics Tournament 2025 - A prestigious student-organized mathematics competition for high school students featuring two tournaments (November 2025 at MIT and February 2026 at Harvard) with individual tests, team rounds, and guts rounds", + "metric_id": "llm_stats.hmmt-2025.score", + "metric_name": "HMMT 2025 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hmmt-2025", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HMMT 2025", + "raw_categories": "[\"math\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "33" + } + }, + "score_details": { + "score": 0.948, + "details": { + "raw_score": "0.948", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "hmmt-2025", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/hmmt-2025\",\"https://api.llm-stats.com/leaderboard/benchmarks/hmmt-2025\"]", + "raw_score_id": "hmmt-2025::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hmmt25::hmmt25-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.hmmt25", + "source_data": { + "dataset_name": "HMMT25", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/hmmt25", + "https://api.llm-stats.com/leaderboard/benchmarks/hmmt25" + ], + "additional_details": { + "raw_benchmark_id": "hmmt25", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Harvard-MIT Mathematics Tournament 2025 - A prestigious student-organized mathematics competition for high school students featuring two tournaments (November 2025 at MIT and February 2026 at Harvard) with individual tests, team rounds, and guts rounds", + "metric_id": "llm_stats.hmmt25.score", + "metric_name": "HMMT25 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hmmt25", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HMMT25", + "raw_categories": "[\"math\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "25" + } + }, + "score_details": { + "score": 0.927, + "details": { + "raw_score": "0.927", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "hmmt25", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/hmmt25\",\"https://api.llm-stats.com/leaderboard/benchmarks/hmmt25\"]", + "raw_score_id": "hmmt25::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ifbench::ifbench-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.ifbench", + "source_data": { + "dataset_name": "IFBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/ifbench", + "https://api.llm-stats.com/leaderboard/benchmarks/ifbench" + ], + "additional_details": { + "raw_benchmark_id": "ifbench", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Instruction Following Benchmark evaluating model's ability to follow complex instructions", + "metric_id": "llm_stats.ifbench.score", + "metric_name": "IFBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ifbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IFBench", + "raw_categories": "[\"instruction_following\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.765, + "details": { + "raw_score": "0.765", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "ifbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/ifbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/ifbench\"]", + "raw_score_id": "ifbench::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ifeval::ifeval-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.ifeval", + "source_data": { + "dataset_name": "IFEval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/ifeval", + "https://api.llm-stats.com/leaderboard/benchmarks/ifeval" + ], + "additional_details": { + "raw_benchmark_id": "ifeval", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Instruction-Following Evaluation (IFEval) benchmark for large language models, focusing on verifiable instructions with 25 types of instructions and around 500 prompts containing one or more verifiable constraints", + "metric_id": "llm_stats.ifeval.score", + "metric_name": "IFEval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ifeval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IFEval", + "raw_categories": "[\"instruction_following\",\"structured_output\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "65" + } + }, + "score_details": { + "score": 0.926, + "details": { + "raw_score": "0.926", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "ifeval", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/ifeval\",\"https://api.llm-stats.com/leaderboard/benchmarks/ifeval\"]", + "raw_score_id": "ifeval::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "imo-answerbench::imo-answerbench-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.imo-answerbench", + "source_data": { + "dataset_name": "IMO-AnswerBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/imo-answerbench", + "https://api.llm-stats.com/leaderboard/benchmarks/imo-answerbench" + ], + "additional_details": { + "raw_benchmark_id": "imo-answerbench", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "IMO-AnswerBench is a benchmark for evaluating mathematical reasoning capabilities on International Mathematical Olympiad (IMO) problems, focusing on answer generation and verification.", + "metric_id": "llm_stats.imo-answerbench.score", + "metric_name": "IMO-AnswerBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "imo-answerbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IMO-AnswerBench", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.809, + "details": { + "raw_score": "0.809", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "imo-answerbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/imo-answerbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/imo-answerbench\"]", + "raw_score_id": "imo-answerbench::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "include::include-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.include", + "source_data": { + "dataset_name": "Include", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/include", + "https://api.llm-stats.com/leaderboard/benchmarks/include" + ], + "additional_details": { + "raw_benchmark_id": "include", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Include benchmark - specific documentation not found in official sources", + "metric_id": "llm_stats.include.score", + "metric_name": "Include score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "include", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Include", + "raw_categories": "[\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "31" + } + }, + "score_details": { + "score": 0.856, + "details": { + "raw_score": "0.856", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "include", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/include\",\"https://api.llm-stats.com/leaderboard/benchmarks/include\"]", + "raw_score_id": "include::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "livecodebench-v6::livecodebench-v6-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.livecodebench-v6", + "source_data": { + "dataset_name": "LiveCodeBench v6", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/livecodebench-v6", + "https://api.llm-stats.com/leaderboard/benchmarks/livecodebench-v6" + ], + "additional_details": { + "raw_benchmark_id": "livecodebench-v6", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LiveCodeBench is a holistic and contamination-free evaluation benchmark for large language models for code. It continuously collects new problems from programming contests (LeetCode, AtCoder, CodeForces) and evaluates four different scenarios: code generation, self-repair, code execution, and test output prediction. Problems are annotated with release dates to enable evaluation on unseen problems released after a model's training cutoff.", + "metric_id": "llm_stats.livecodebench-v6.score", + "metric_name": "LiveCodeBench v6 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "livecodebench-v6", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LiveCodeBench v6", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "53" + } + }, + "score_details": { + "score": 0.836, + "details": { + "raw_score": "0.836", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "livecodebench-v6", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/livecodebench-v6\",\"https://api.llm-stats.com/leaderboard/benchmarks/livecodebench-v6\"]", + "raw_score_id": "livecodebench-v6::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "longbench-v2::longbench-v2-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.longbench-v2", + "source_data": { + "dataset_name": "LongBench v2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/longbench-v2", + "https://api.llm-stats.com/leaderboard/benchmarks/longbench-v2" + ], + "additional_details": { + "raw_benchmark_id": "longbench-v2", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LongBench v2 is a benchmark designed to assess the ability of LLMs to handle long-context problems requiring deep understanding and reasoning across real-world multitasks. It consists of 503 challenging multiple-choice questions with contexts ranging from 8k to 2M words across six major task categories: single-document QA, multi-document QA, long in-context learning, long-dialogue history understanding, code repository understanding, and long structured data understanding.", + "metric_id": "llm_stats.longbench-v2.score", + "metric_name": "LongBench v2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "longbench-v2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LongBench v2", + "raw_categories": "[\"long_context\",\"reasoning\",\"structured_output\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.632, + "details": { + "raw_score": "0.632", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "longbench-v2", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/longbench-v2\",\"https://api.llm-stats.com/leaderboard/benchmarks/longbench-v2\"]", + "raw_score_id": "longbench-v2::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "maxife::maxife-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.maxife", + "source_data": { + "dataset_name": "MAXIFE", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/maxife", + "https://api.llm-stats.com/leaderboard/benchmarks/maxife" + ], + "additional_details": { + "raw_benchmark_id": "maxife", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MAXIFE is a multilingual benchmark evaluating LLMs on instruction following and execution across multiple languages and cultural contexts.", + "metric_id": "llm_stats.maxife.score", + "metric_name": "MAXIFE score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "maxife", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MAXIFE", + "raw_categories": "[\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.882, + "details": { + "raw_score": "0.882", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "maxife", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/maxife\",\"https://api.llm-stats.com/leaderboard/benchmarks/maxife\"]", + "raw_score_id": "maxife::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mcp-mark::mcp-mark-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.mcp-mark", + "source_data": { + "dataset_name": "MCP-Mark", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/mcp-mark", + "https://api.llm-stats.com/leaderboard/benchmarks/mcp-mark" + ], + "additional_details": { + "raw_benchmark_id": "mcp-mark", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MCP-Mark evaluates LLMs on their ability to use Model Context Protocol (MCP) tools effectively, testing tool discovery, selection, invocation, and result interpretation across diverse MCP server scenarios.", + "metric_id": "llm_stats.mcp-mark.score", + "metric_name": "MCP-Mark score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mcp-mark", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MCP-Mark", + "raw_categories": "[\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.461, + "details": { + "raw_score": "0.461", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "mcp-mark", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/mcp-mark\",\"https://api.llm-stats.com/leaderboard/benchmarks/mcp-mark\"]", + "raw_score_id": "mcp-mark::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu-pro::mmlu-pro-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.mmlu-pro", + "source_data": { + "dataset_name": "MMLU-Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/mmlu-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu-pro" + ], + "additional_details": { + "raw_benchmark_id": "mmlu-pro", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A more robust and challenging multi-task language understanding benchmark that extends MMLU by expanding multiple-choice options from 4 to 10, eliminating trivial questions, and focusing on reasoning-intensive tasks. Features over 12,000 curated questions across 14 domains and causes a 16-33% accuracy drop compared to original MMLU.", + "metric_id": "llm_stats.mmlu-pro.score", + "metric_name": "MMLU-Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU-Pro", + "raw_categories": "[\"language\",\"legal\",\"math\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "129" + } + }, + "score_details": { + "score": 0.878, + "details": { + "raw_score": "0.878", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "mmlu-pro", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/mmlu-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu-pro\"]", + "raw_score_id": "mmlu-pro::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu-prox::mmlu-prox-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.mmlu-prox", + "source_data": { + "dataset_name": "MMLU-ProX", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/mmlu-prox", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu-prox" + ], + "additional_details": { + "raw_benchmark_id": "mmlu-prox", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Extended version of MMLU-Pro providing additional challenging multiple-choice questions for evaluating language models across diverse academic and professional domains. Built on the foundation of the Massive Multitask Language Understanding benchmark framework.", + "metric_id": "llm_stats.mmlu-prox.score", + "metric_name": "MMLU-ProX score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu-prox", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU-ProX", + "raw_categories": "[\"language\",\"legal\",\"math\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "32" + } + }, + "score_details": { + "score": 0.847, + "details": { + "raw_score": "0.847", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "mmlu-prox", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/mmlu-prox\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu-prox\"]", + "raw_score_id": "mmlu-prox::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu-redux::mmlu-redux-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.mmlu-redux", + "source_data": { + "dataset_name": "MMLU-Redux", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/mmlu-redux", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu-redux" + ], + "additional_details": { + "raw_benchmark_id": "mmlu-redux", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "An improved version of the MMLU benchmark featuring manually re-annotated questions to identify and correct errors in the original dataset. Provides more reliable evaluation metrics for language models by addressing dataset quality issues found in the original MMLU.", + "metric_id": "llm_stats.mmlu-redux.score", + "metric_name": "MMLU-Redux score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu-redux", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU-Redux", + "raw_categories": "[\"language\",\"math\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "48" + } + }, + "score_details": { + "score": 0.949, + "details": { + "raw_score": "0.949", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "mmlu-redux", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/mmlu-redux\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu-redux\"]", + "raw_score_id": "mmlu-redux::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmmlu::mmmlu-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.mmmlu", + "source_data": { + "dataset_name": "MMMLU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/mmmlu", + "https://api.llm-stats.com/leaderboard/benchmarks/mmmlu" + ], + "additional_details": { + "raw_benchmark_id": "mmmlu", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Multilingual Massive Multitask Language Understanding dataset released by OpenAI, featuring professionally translated MMLU test questions across 14 languages including Arabic, Bengali, German, Spanish, French, Hindi, Indonesian, Italian, Japanese, Korean, Portuguese, Swahili, Yoruba, and Chinese. Contains approximately 15,908 multiple-choice questions per language covering 57 subjects.", + "metric_id": "llm_stats.mmmlu.score", + "metric_name": "MMMLU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmmlu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMMLU", + "raw_categories": "[\"language\",\"math\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "49" + } + }, + "score_details": { + "score": 0.885, + "details": { + "raw_score": "0.885", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "mmmlu", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/mmmlu\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmmlu\"]", + "raw_score_id": "mmmlu::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "multichallenge::multichallenge-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.multichallenge", + "source_data": { + "dataset_name": "Multi-Challenge", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/multichallenge", + "https://api.llm-stats.com/leaderboard/benchmarks/multichallenge" + ], + "additional_details": { + "raw_benchmark_id": "multichallenge", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MultiChallenge is a realistic multi-turn conversation evaluation benchmark that challenges frontier LLMs across four key categories: instruction retention (maintaining instructions throughout conversations), inference memory (recalling and connecting details from previous turns), reliable versioned editing (adapting to evolving instructions during collaborative editing), and self-coherence (avoiding contradictions in responses). The benchmark evaluates models on sustained, contextually complex dialogues across diverse topics including travel planning, technical documentation, and professional communication.", + "metric_id": "llm_stats.multichallenge.score", + "metric_name": "Multi-Challenge score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "multichallenge", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Multi-Challenge", + "raw_categories": "[\"reasoning\",\"communication\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "29" + } + }, + "score_details": { + "score": 0.676, + "details": { + "raw_score": "0.676", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "multichallenge", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/multichallenge\",\"https://api.llm-stats.com/leaderboard/benchmarks/multichallenge\"]", + "raw_score_id": "multichallenge::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "nova-63::nova-63-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.nova-63", + "source_data": { + "dataset_name": "NOVA-63", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/nova-63", + "https://api.llm-stats.com/leaderboard/benchmarks/nova-63" + ], + "additional_details": { + "raw_benchmark_id": "nova-63", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "NOVA-63 is a multilingual evaluation benchmark covering 63 languages, designed to assess LLM performance across diverse linguistic contexts and tasks.", + "metric_id": "llm_stats.nova-63.score", + "metric_name": "NOVA-63 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "nova-63", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "NOVA-63", + "raw_categories": "[\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.591, + "details": { + "raw_score": "0.591", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "nova-63", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/nova-63\",\"https://api.llm-stats.com/leaderboard/benchmarks/nova-63\"]", + "raw_score_id": "nova-63::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "polymath::polymath-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.polymath", + "source_data": { + "dataset_name": "PolyMATH", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/polymath", + "https://api.llm-stats.com/leaderboard/benchmarks/polymath" + ], + "additional_details": { + "raw_benchmark_id": "polymath", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Polymath is a challenging multi-modal mathematical reasoning benchmark designed to evaluate the general cognitive reasoning abilities of Multi-modal Large Language Models (MLLMs). The benchmark comprises 5,000 manually collected high-quality images of cognitive textual and visual challenges across 10 distinct categories, including pattern recognition, spatial reasoning, and relative reasoning.", + "metric_id": "llm_stats.polymath.score", + "metric_name": "PolyMATH score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "polymath", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PolyMATH", + "raw_categories": "[\"math\",\"multimodal\",\"reasoning\",\"spatial_reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.733, + "details": { + "raw_score": "0.733", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "polymath", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/polymath\",\"https://api.llm-stats.com/leaderboard/benchmarks/polymath\"]", + "raw_score_id": "polymath::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "seal-0::seal-0-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.seal-0", + "source_data": { + "dataset_name": "Seal-0", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/seal-0", + "https://api.llm-stats.com/leaderboard/benchmarks/seal-0" + ], + "additional_details": { + "raw_benchmark_id": "seal-0", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Seal-0 is a benchmark for evaluating agentic search capabilities, testing models' ability to navigate and retrieve information using tools.", + "metric_id": "llm_stats.seal-0.score", + "metric_name": "Seal-0 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "seal-0", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Seal-0", + "raw_categories": "[\"reasoning\",\"search\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.469, + "details": { + "raw_score": "0.469", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "seal-0", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/seal-0\",\"https://api.llm-stats.com/leaderboard/benchmarks/seal-0\"]", + "raw_score_id": "seal-0::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "seccodebench::seccodebench-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.seccodebench", + "source_data": { + "dataset_name": "SecCodeBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/seccodebench", + "https://api.llm-stats.com/leaderboard/benchmarks/seccodebench" + ], + "additional_details": { + "raw_benchmark_id": "seccodebench", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SecCodeBench evaluates LLM coding agents on secure code generation and vulnerability detection, testing the ability to produce code that is both functional and free from security vulnerabilities.", + "metric_id": "llm_stats.seccodebench.score", + "metric_name": "SecCodeBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "seccodebench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SecCodeBench", + "raw_categories": "[\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.683, + "details": { + "raw_score": "0.683", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "seccodebench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/seccodebench\",\"https://api.llm-stats.com/leaderboard/benchmarks/seccodebench\"]", + "raw_score_id": "seccodebench::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "supergpqa::supergpqa-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.supergpqa", + "source_data": { + "dataset_name": "SuperGPQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/supergpqa", + "https://api.llm-stats.com/leaderboard/benchmarks/supergpqa" + ], + "additional_details": { + "raw_benchmark_id": "supergpqa", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SuperGPQA is a comprehensive benchmark that evaluates large language models across 285 graduate-level academic disciplines. The benchmark contains 25,957 questions covering 13 broad disciplinary areas including Engineering, Medicine, Science, and Law, with specialized fields in light industry, agriculture, and service-oriented domains. It employs a Human-LLM collaborative filtering mechanism with over 80 expert annotators to create challenging questions that assess graduate-level knowledge and reasoning capabilities.", + "metric_id": "llm_stats.supergpqa.score", + "metric_name": "SuperGPQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "supergpqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SuperGPQA", + "raw_categories": "[\"legal\",\"math\",\"physics\",\"reasoning\",\"finance\",\"general\",\"healthcare\",\"chemistry\",\"economics\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "34" + } + }, + "score_details": { + "score": 0.704, + "details": { + "raw_score": "0.704", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "supergpqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/supergpqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/supergpqa\"]", + "raw_score_id": "supergpqa::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-multilingual::swe-bench-multilingual-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.swe-bench-multilingual", + "source_data": { + "dataset_name": "SWE-bench Multilingual", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/swe-bench-multilingual", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-multilingual" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-multilingual", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multilingual benchmark for issue resolving in software engineering that covers Java, TypeScript, JavaScript, Go, Rust, C, and C++. Contains 1,632 high-quality instances carefully annotated from 2,456 candidates by 68 expert annotators, designed to evaluate Large Language Models across diverse software ecosystems beyond Python.", + "metric_id": "llm_stats.swe-bench-multilingual.score", + "metric_name": "SWE-bench Multilingual score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-multilingual", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-bench Multilingual", + "raw_categories": "[\"reasoning\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "33" + } + }, + "score_details": { + "score": 0.693, + "details": { + "raw_score": "0.693", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "swe-bench-multilingual", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/swe-bench-multilingual\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-multilingual\"]", + "raw_score_id": "swe-bench-multilingual::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "t2-bench::t2-bench-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.t2-bench", + "source_data": { + "dataset_name": "t2-bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/t2-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/t2-bench" + ], + "additional_details": { + "raw_benchmark_id": "t2-bench", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "t2-bench is a benchmark for evaluating agentic tool use capabilities, measuring how well models can select, sequence, and utilize tools to solve complex tasks. It tests autonomous planning and execution in multi-step scenarios.", + "metric_id": "llm_stats.t2-bench.score", + "metric_name": "t2-bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "t2-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "t2-bench", + "raw_categories": "[\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.867, + "details": { + "raw_score": "0.867", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "t2-bench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/t2-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/t2-bench\"]", + "raw_score_id": "t2-bench::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "toolathlon::toolathlon-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.toolathlon", + "source_data": { + "dataset_name": "Toolathlon", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/toolathlon", + "https://api.llm-stats.com/leaderboard/benchmarks/toolathlon" + ], + "additional_details": { + "raw_benchmark_id": "toolathlon", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Tool Decathlon is a comprehensive benchmark for evaluating AI agents' ability to use multiple tools across diverse task categories. It measures proficiency in tool selection, sequencing, and execution across ten different tool-use scenarios.", + "metric_id": "llm_stats.toolathlon.score", + "metric_name": "Toolathlon score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "toolathlon", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Toolathlon", + "raw_categories": "[\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "24" + } + }, + "score_details": { + "score": 0.383, + "details": { + "raw_score": "0.383", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "toolathlon", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/toolathlon\",\"https://api.llm-stats.com/leaderboard/benchmarks/toolathlon\"]", + "raw_score_id": "toolathlon::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "vita-bench::vita-bench-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.vita-bench", + "source_data": { + "dataset_name": "VITA-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/vita-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/vita-bench" + ], + "additional_details": { + "raw_benchmark_id": "vita-bench", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VITA-Bench evaluates AI agents on real-world virtual task automation, measuring their ability to complete complex multi-step tasks in simulated environments.", + "metric_id": "llm_stats.vita-bench.score", + "metric_name": "VITA-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "vita-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VITA-Bench", + "raw_categories": "[\"reasoning\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.497, + "details": { + "raw_score": "0.497", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "vita-bench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/vita-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/vita-bench\"]", + "raw_score_id": "vita-bench::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "widesearch::widesearch-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.widesearch", + "source_data": { + "dataset_name": "WideSearch", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/widesearch", + "https://api.llm-stats.com/leaderboard/benchmarks/widesearch" + ], + "additional_details": { + "raw_benchmark_id": "widesearch", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WideSearch is an agentic search benchmark that evaluates models' ability to perform broad, parallel search operations across multiple sources. It tests wide-coverage information retrieval and synthesis capabilities.", + "metric_id": "llm_stats.widesearch.score", + "metric_name": "WideSearch score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "widesearch", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WideSearch", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.74, + "details": { + "raw_score": "0.74", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "widesearch", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/widesearch\",\"https://api.llm-stats.com/leaderboard/benchmarks/widesearch\"]", + "raw_score_id": "widesearch::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "wmt24::wmt24-qwen3.5-397b-a17b", + "evaluation_name": "llm_stats.wmt24", + "source_data": { + "dataset_name": "WMT24++", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-397b-a17b", + "https://llm-stats.com/benchmarks/wmt24", + "https://api.llm-stats.com/leaderboard/benchmarks/wmt24++" + ], + "additional_details": { + "raw_benchmark_id": "wmt24++", + "raw_model_id": "qwen3.5-397b-a17b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WMT24++ is a comprehensive multilingual machine translation benchmark that expands the WMT24 dataset to cover 55 languages and dialects. It includes human-written references and post-edits across four domains (literary, news, social, and speech) to evaluate machine translation systems and large language models across diverse linguistic contexts.", + "metric_id": "llm_stats.wmt24.score", + "metric_name": "WMT24++ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "wmt24++", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WMT24++", + "raw_categories": "[\"language\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.789, + "details": { + "raw_score": "0.789", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-397b-a17b", + "raw_benchmark_id": "wmt24++", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-397b-a17b\",\"https://llm-stats.com/benchmarks/wmt24\",\"https://api.llm-stats.com/leaderboard/benchmarks/wmt24++\"]", + "raw_score_id": "wmt24++::qwen3.5-397b-a17b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/qwen/qwen3.5-4b/41e52f18-094e-4929-b8fd-09738be52fef.json b/data/llm_stats/qwen/qwen3.5-4b/41e52f18-094e-4929-b8fd-09738be52fef.json new file mode 100644 index 0000000000000000000000000000000000000000..07612a1363346df4fdf5b98267dae9fac3ae0d4a --- /dev/null +++ b/data/llm_stats/qwen/qwen3.5-4b/41e52f18-094e-4929-b8fd-09738be52fef.json @@ -0,0 +1,879 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/qwen_qwen3.5-4b/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Qwen3.5-4B", + "id": "qwen/qwen3.5-4b", + "developer": "qwen", + "additional_details": { + "raw_id": "qwen3.5-4b", + "raw_name": "Qwen3.5-4B", + "raw_model_id": "qwen3.5-4b", + "raw_model_name": "Qwen3.5-4B", + "raw_organization_id": "qwen", + "raw_organization_name": "Alibaba Cloud / Qwen Team", + "raw_release_date": "2026-03-02", + "raw_announcement_date": "2026-03-02", + "raw_multimodal": "true", + "raw_param_count": "4000000000", + "raw_provider_slug": "qwen", + "raw_provider_name": "Alibaba Cloud / Qwen Team" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aa-lcr::aa-lcr-qwen3.5-4b", + "evaluation_name": "llm_stats.aa-lcr", + "source_data": { + "dataset_name": "AA-LCR", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-4b", + "https://llm-stats.com/benchmarks/aa-lcr", + "https://api.llm-stats.com/leaderboard/benchmarks/aa-lcr" + ], + "additional_details": { + "raw_benchmark_id": "aa-lcr", + "raw_model_id": "qwen3.5-4b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Agent Arena Long Context Reasoning benchmark", + "metric_id": "llm_stats.aa-lcr.score", + "metric_name": "AA-LCR score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aa-lcr", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AA-LCR", + "raw_categories": "[\"long_context\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "14" + } + }, + "score_details": { + "score": 0.57, + "details": { + "raw_score": "0.57", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-4b", + "raw_benchmark_id": "aa-lcr", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-4b\",\"https://llm-stats.com/benchmarks/aa-lcr\",\"https://api.llm-stats.com/leaderboard/benchmarks/aa-lcr\"]", + "raw_score_id": "aa-lcr::qwen3.5-4b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bfcl-v4::bfcl-v4-qwen3.5-4b", + "evaluation_name": "llm_stats.bfcl-v4", + "source_data": { + "dataset_name": "BFCL-V4", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-4b", + "https://llm-stats.com/benchmarks/bfcl-v4", + "https://api.llm-stats.com/leaderboard/benchmarks/bfcl-v4" + ], + "additional_details": { + "raw_benchmark_id": "bfcl-v4", + "raw_model_id": "qwen3.5-4b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Berkeley Function Calling Leaderboard V4 (BFCL-V4) evaluates LLMs on their ability to accurately call functions and APIs, including simple, multiple, parallel, and nested function calls across diverse programming scenarios.", + "metric_id": "llm_stats.bfcl-v4.score", + "metric_name": "BFCL-V4 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bfcl-v4", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BFCL-V4", + "raw_categories": "[\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.503, + "details": { + "raw_score": "0.503", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-4b", + "raw_benchmark_id": "bfcl-v4", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-4b\",\"https://llm-stats.com/benchmarks/bfcl-v4\",\"https://api.llm-stats.com/leaderboard/benchmarks/bfcl-v4\"]", + "raw_score_id": "bfcl-v4::qwen3.5-4b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "c-eval::c-eval-qwen3.5-4b", + "evaluation_name": "llm_stats.c-eval", + "source_data": { + "dataset_name": "C-Eval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-4b", + "https://llm-stats.com/benchmarks/c-eval", + "https://api.llm-stats.com/leaderboard/benchmarks/c-eval" + ], + "additional_details": { + "raw_benchmark_id": "c-eval", + "raw_model_id": "qwen3.5-4b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "C-Eval is a comprehensive Chinese evaluation suite designed to assess advanced knowledge and reasoning abilities of foundation models in a Chinese context. It comprises 13,948 multiple-choice questions across 52 diverse disciplines spanning humanities, science, and engineering, with four difficulty levels: middle school, high school, college, and professional. The benchmark includes C-Eval Hard, a subset of very challenging subjects requiring advanced reasoning abilities.", + "metric_id": "llm_stats.c-eval.score", + "metric_name": "C-Eval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "c-eval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "C-Eval", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.851, + "details": { + "raw_score": "0.851", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-4b", + "raw_benchmark_id": "c-eval", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-4b\",\"https://llm-stats.com/benchmarks/c-eval\",\"https://api.llm-stats.com/leaderboard/benchmarks/c-eval\"]", + "raw_score_id": "c-eval::qwen3.5-4b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "deep-planning::deep-planning-qwen3.5-4b", + "evaluation_name": "llm_stats.deep-planning", + "source_data": { + "dataset_name": "DeepPlanning", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-4b", + "https://llm-stats.com/benchmarks/deep-planning", + "https://api.llm-stats.com/leaderboard/benchmarks/deep-planning" + ], + "additional_details": { + "raw_benchmark_id": "deep-planning", + "raw_model_id": "qwen3.5-4b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "DeepPlanning evaluates LLMs on complex multi-step planning tasks requiring long-horizon reasoning, goal decomposition, and strategic decision-making.", + "metric_id": "llm_stats.deep-planning.score", + "metric_name": "DeepPlanning score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "deep-planning", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DeepPlanning", + "raw_categories": "[\"reasoning\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.176, + "details": { + "raw_score": "0.176", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-4b", + "raw_benchmark_id": "deep-planning", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-4b\",\"https://llm-stats.com/benchmarks/deep-planning\",\"https://api.llm-stats.com/leaderboard/benchmarks/deep-planning\"]", + "raw_score_id": "deep-planning::qwen3.5-4b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "global-piqa::global-piqa-qwen3.5-4b", + "evaluation_name": "llm_stats.global-piqa", + "source_data": { + "dataset_name": "Global PIQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-4b", + "https://llm-stats.com/benchmarks/global-piqa", + "https://api.llm-stats.com/leaderboard/benchmarks/global-piqa" + ], + "additional_details": { + "raw_benchmark_id": "global-piqa", + "raw_model_id": "qwen3.5-4b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Global PIQA is a multilingual commonsense reasoning benchmark that evaluates physical interaction knowledge across 100 languages and cultures. It tests AI systems' understanding of physical world knowledge in diverse cultural contexts through multiple choice questions about everyday situations requiring physical commonsense.", + "metric_id": "llm_stats.global-piqa.score", + "metric_name": "Global PIQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "global-piqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Global PIQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.789, + "details": { + "raw_score": "0.789", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-4b", + "raw_benchmark_id": "global-piqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-4b\",\"https://llm-stats.com/benchmarks/global-piqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/global-piqa\"]", + "raw_score_id": "global-piqa::qwen3.5-4b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hmmt25::hmmt25-qwen3.5-4b", + "evaluation_name": "llm_stats.hmmt25", + "source_data": { + "dataset_name": "HMMT25", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-4b", + "https://llm-stats.com/benchmarks/hmmt25", + "https://api.llm-stats.com/leaderboard/benchmarks/hmmt25" + ], + "additional_details": { + "raw_benchmark_id": "hmmt25", + "raw_model_id": "qwen3.5-4b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Harvard-MIT Mathematics Tournament 2025 - A prestigious student-organized mathematics competition for high school students featuring two tournaments (November 2025 at MIT and February 2026 at Harvard) with individual tests, team rounds, and guts rounds", + "metric_id": "llm_stats.hmmt25.score", + "metric_name": "HMMT25 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hmmt25", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HMMT25", + "raw_categories": "[\"math\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "25" + } + }, + "score_details": { + "score": 0.768, + "details": { + "raw_score": "0.768", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-4b", + "raw_benchmark_id": "hmmt25", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-4b\",\"https://llm-stats.com/benchmarks/hmmt25\",\"https://api.llm-stats.com/leaderboard/benchmarks/hmmt25\"]", + "raw_score_id": "hmmt25::qwen3.5-4b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ifeval::ifeval-qwen3.5-4b", + "evaluation_name": "llm_stats.ifeval", + "source_data": { + "dataset_name": "IFEval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-4b", + "https://llm-stats.com/benchmarks/ifeval", + "https://api.llm-stats.com/leaderboard/benchmarks/ifeval" + ], + "additional_details": { + "raw_benchmark_id": "ifeval", + "raw_model_id": "qwen3.5-4b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Instruction-Following Evaluation (IFEval) benchmark for large language models, focusing on verifiable instructions with 25 types of instructions and around 500 prompts containing one or more verifiable constraints", + "metric_id": "llm_stats.ifeval.score", + "metric_name": "IFEval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ifeval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IFEval", + "raw_categories": "[\"instruction_following\",\"structured_output\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "65" + } + }, + "score_details": { + "score": 0.898, + "details": { + "raw_score": "0.898", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-4b", + "raw_benchmark_id": "ifeval", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-4b\",\"https://llm-stats.com/benchmarks/ifeval\",\"https://api.llm-stats.com/leaderboard/benchmarks/ifeval\"]", + "raw_score_id": "ifeval::qwen3.5-4b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "longbench-v2::longbench-v2-qwen3.5-4b", + "evaluation_name": "llm_stats.longbench-v2", + "source_data": { + "dataset_name": "LongBench v2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-4b", + "https://llm-stats.com/benchmarks/longbench-v2", + "https://api.llm-stats.com/leaderboard/benchmarks/longbench-v2" + ], + "additional_details": { + "raw_benchmark_id": "longbench-v2", + "raw_model_id": "qwen3.5-4b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LongBench v2 is a benchmark designed to assess the ability of LLMs to handle long-context problems requiring deep understanding and reasoning across real-world multitasks. It consists of 503 challenging multiple-choice questions with contexts ranging from 8k to 2M words across six major task categories: single-document QA, multi-document QA, long in-context learning, long-dialogue history understanding, code repository understanding, and long structured data understanding.", + "metric_id": "llm_stats.longbench-v2.score", + "metric_name": "LongBench v2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "longbench-v2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LongBench v2", + "raw_categories": "[\"long_context\",\"reasoning\",\"structured_output\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.5, + "details": { + "raw_score": "0.5", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-4b", + "raw_benchmark_id": "longbench-v2", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-4b\",\"https://llm-stats.com/benchmarks/longbench-v2\",\"https://api.llm-stats.com/leaderboard/benchmarks/longbench-v2\"]", + "raw_score_id": "longbench-v2::qwen3.5-4b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "maxife::maxife-qwen3.5-4b", + "evaluation_name": "llm_stats.maxife", + "source_data": { + "dataset_name": "MAXIFE", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-4b", + "https://llm-stats.com/benchmarks/maxife", + "https://api.llm-stats.com/leaderboard/benchmarks/maxife" + ], + "additional_details": { + "raw_benchmark_id": "maxife", + "raw_model_id": "qwen3.5-4b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MAXIFE is a multilingual benchmark evaluating LLMs on instruction following and execution across multiple languages and cultural contexts.", + "metric_id": "llm_stats.maxife.score", + "metric_name": "MAXIFE score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "maxife", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MAXIFE", + "raw_categories": "[\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.78, + "details": { + "raw_score": "0.78", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-4b", + "raw_benchmark_id": "maxife", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-4b\",\"https://llm-stats.com/benchmarks/maxife\",\"https://api.llm-stats.com/leaderboard/benchmarks/maxife\"]", + "raw_score_id": "maxife::qwen3.5-4b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu-prox::mmlu-prox-qwen3.5-4b", + "evaluation_name": "llm_stats.mmlu-prox", + "source_data": { + "dataset_name": "MMLU-ProX", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-4b", + "https://llm-stats.com/benchmarks/mmlu-prox", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu-prox" + ], + "additional_details": { + "raw_benchmark_id": "mmlu-prox", + "raw_model_id": "qwen3.5-4b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Extended version of MMLU-Pro providing additional challenging multiple-choice questions for evaluating language models across diverse academic and professional domains. Built on the foundation of the Massive Multitask Language Understanding benchmark framework.", + "metric_id": "llm_stats.mmlu-prox.score", + "metric_name": "MMLU-ProX score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu-prox", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU-ProX", + "raw_categories": "[\"language\",\"legal\",\"math\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "32" + } + }, + "score_details": { + "score": 0.715, + "details": { + "raw_score": "0.715", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-4b", + "raw_benchmark_id": "mmlu-prox", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-4b\",\"https://llm-stats.com/benchmarks/mmlu-prox\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu-prox\"]", + "raw_score_id": "mmlu-prox::qwen3.5-4b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "multichallenge::multichallenge-qwen3.5-4b", + "evaluation_name": "llm_stats.multichallenge", + "source_data": { + "dataset_name": "Multi-Challenge", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-4b", + "https://llm-stats.com/benchmarks/multichallenge", + "https://api.llm-stats.com/leaderboard/benchmarks/multichallenge" + ], + "additional_details": { + "raw_benchmark_id": "multichallenge", + "raw_model_id": "qwen3.5-4b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MultiChallenge is a realistic multi-turn conversation evaluation benchmark that challenges frontier LLMs across four key categories: instruction retention (maintaining instructions throughout conversations), inference memory (recalling and connecting details from previous turns), reliable versioned editing (adapting to evolving instructions during collaborative editing), and self-coherence (avoiding contradictions in responses). The benchmark evaluates models on sustained, contextually complex dialogues across diverse topics including travel planning, technical documentation, and professional communication.", + "metric_id": "llm_stats.multichallenge.score", + "metric_name": "Multi-Challenge score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "multichallenge", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Multi-Challenge", + "raw_categories": "[\"reasoning\",\"communication\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "29" + } + }, + "score_details": { + "score": 0.49, + "details": { + "raw_score": "0.49", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-4b", + "raw_benchmark_id": "multichallenge", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-4b\",\"https://llm-stats.com/benchmarks/multichallenge\",\"https://api.llm-stats.com/leaderboard/benchmarks/multichallenge\"]", + "raw_score_id": "multichallenge::qwen3.5-4b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "nova-63::nova-63-qwen3.5-4b", + "evaluation_name": "llm_stats.nova-63", + "source_data": { + "dataset_name": "NOVA-63", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-4b", + "https://llm-stats.com/benchmarks/nova-63", + "https://api.llm-stats.com/leaderboard/benchmarks/nova-63" + ], + "additional_details": { + "raw_benchmark_id": "nova-63", + "raw_model_id": "qwen3.5-4b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "NOVA-63 is a multilingual evaluation benchmark covering 63 languages, designed to assess LLM performance across diverse linguistic contexts and tasks.", + "metric_id": "llm_stats.nova-63.score", + "metric_name": "NOVA-63 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "nova-63", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "NOVA-63", + "raw_categories": "[\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.543, + "details": { + "raw_score": "0.543", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-4b", + "raw_benchmark_id": "nova-63", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-4b\",\"https://llm-stats.com/benchmarks/nova-63\",\"https://api.llm-stats.com/leaderboard/benchmarks/nova-63\"]", + "raw_score_id": "nova-63::qwen3.5-4b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "polymath::polymath-qwen3.5-4b", + "evaluation_name": "llm_stats.polymath", + "source_data": { + "dataset_name": "PolyMATH", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-4b", + "https://llm-stats.com/benchmarks/polymath", + "https://api.llm-stats.com/leaderboard/benchmarks/polymath" + ], + "additional_details": { + "raw_benchmark_id": "polymath", + "raw_model_id": "qwen3.5-4b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Polymath is a challenging multi-modal mathematical reasoning benchmark designed to evaluate the general cognitive reasoning abilities of Multi-modal Large Language Models (MLLMs). The benchmark comprises 5,000 manually collected high-quality images of cognitive textual and visual challenges across 10 distinct categories, including pattern recognition, spatial reasoning, and relative reasoning.", + "metric_id": "llm_stats.polymath.score", + "metric_name": "PolyMATH score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "polymath", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PolyMATH", + "raw_categories": "[\"math\",\"multimodal\",\"reasoning\",\"spatial_reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.511, + "details": { + "raw_score": "0.511", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-4b", + "raw_benchmark_id": "polymath", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-4b\",\"https://llm-stats.com/benchmarks/polymath\",\"https://api.llm-stats.com/leaderboard/benchmarks/polymath\"]", + "raw_score_id": "polymath::qwen3.5-4b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "t2-bench::t2-bench-qwen3.5-4b", + "evaluation_name": "llm_stats.t2-bench", + "source_data": { + "dataset_name": "t2-bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-4b", + "https://llm-stats.com/benchmarks/t2-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/t2-bench" + ], + "additional_details": { + "raw_benchmark_id": "t2-bench", + "raw_model_id": "qwen3.5-4b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "t2-bench is a benchmark for evaluating agentic tool use capabilities, measuring how well models can select, sequence, and utilize tools to solve complex tasks. It tests autonomous planning and execution in multi-step scenarios.", + "metric_id": "llm_stats.t2-bench.score", + "metric_name": "t2-bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "t2-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "t2-bench", + "raw_categories": "[\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.799, + "details": { + "raw_score": "0.799", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-4b", + "raw_benchmark_id": "t2-bench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-4b\",\"https://llm-stats.com/benchmarks/t2-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/t2-bench\"]", + "raw_score_id": "t2-bench::qwen3.5-4b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "vita-bench::vita-bench-qwen3.5-4b", + "evaluation_name": "llm_stats.vita-bench", + "source_data": { + "dataset_name": "VITA-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-4b", + "https://llm-stats.com/benchmarks/vita-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/vita-bench" + ], + "additional_details": { + "raw_benchmark_id": "vita-bench", + "raw_model_id": "qwen3.5-4b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VITA-Bench evaluates AI agents on real-world virtual task automation, measuring their ability to complete complex multi-step tasks in simulated environments.", + "metric_id": "llm_stats.vita-bench.score", + "metric_name": "VITA-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "vita-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VITA-Bench", + "raw_categories": "[\"reasoning\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.22, + "details": { + "raw_score": "0.22", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-4b", + "raw_benchmark_id": "vita-bench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-4b\",\"https://llm-stats.com/benchmarks/vita-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/vita-bench\"]", + "raw_score_id": "vita-bench::qwen3.5-4b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "wmt24::wmt24-qwen3.5-4b", + "evaluation_name": "llm_stats.wmt24", + "source_data": { + "dataset_name": "WMT24++", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-4b", + "https://llm-stats.com/benchmarks/wmt24", + "https://api.llm-stats.com/leaderboard/benchmarks/wmt24++" + ], + "additional_details": { + "raw_benchmark_id": "wmt24++", + "raw_model_id": "qwen3.5-4b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WMT24++ is a comprehensive multilingual machine translation benchmark that expands the WMT24 dataset to cover 55 languages and dialects. It includes human-written references and post-edits across four domains (literary, news, social, and speech) to evaluate machine translation systems and large language models across diverse linguistic contexts.", + "metric_id": "llm_stats.wmt24.score", + "metric_name": "WMT24++ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "wmt24++", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WMT24++", + "raw_categories": "[\"language\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.666, + "details": { + "raw_score": "0.666", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-4b", + "raw_benchmark_id": "wmt24++", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-4b\",\"https://llm-stats.com/benchmarks/wmt24\",\"https://api.llm-stats.com/leaderboard/benchmarks/wmt24++\"]", + "raw_score_id": "wmt24++::qwen3.5-4b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/qwen/qwen3.5-9b/250a94e1-a383-4639-9812-1949409dbf25.json b/data/llm_stats/qwen/qwen3.5-9b/250a94e1-a383-4639-9812-1949409dbf25.json new file mode 100644 index 0000000000000000000000000000000000000000..b7c0d64057222673eb0320e5b376a0cdfa01ad3e --- /dev/null +++ b/data/llm_stats/qwen/qwen3.5-9b/250a94e1-a383-4639-9812-1949409dbf25.json @@ -0,0 +1,983 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/qwen_qwen3.5-9b/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "Qwen3.5-9B", + "id": "qwen/qwen3.5-9b", + "developer": "qwen", + "additional_details": { + "raw_id": "qwen3.5-9b", + "raw_name": "Qwen3.5-9B", + "raw_model_id": "qwen3.5-9b", + "raw_model_name": "Qwen3.5-9B", + "raw_organization_id": "qwen", + "raw_organization_name": "Alibaba Cloud / Qwen Team", + "raw_release_date": "2026-03-02", + "raw_announcement_date": "2026-03-02", + "raw_multimodal": "true", + "raw_param_count": "9000000000", + "raw_provider_slug": "qwen", + "raw_provider_name": "Alibaba Cloud / Qwen Team" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aa-lcr::aa-lcr-qwen3.5-9b", + "evaluation_name": "llm_stats.aa-lcr", + "source_data": { + "dataset_name": "AA-LCR", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-9b", + "https://llm-stats.com/benchmarks/aa-lcr", + "https://api.llm-stats.com/leaderboard/benchmarks/aa-lcr" + ], + "additional_details": { + "raw_benchmark_id": "aa-lcr", + "raw_model_id": "qwen3.5-9b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Agent Arena Long Context Reasoning benchmark", + "metric_id": "llm_stats.aa-lcr.score", + "metric_name": "AA-LCR score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aa-lcr", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AA-LCR", + "raw_categories": "[\"long_context\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "14" + } + }, + "score_details": { + "score": 0.63, + "details": { + "raw_score": "0.63", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-9b", + "raw_benchmark_id": "aa-lcr", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-9b\",\"https://llm-stats.com/benchmarks/aa-lcr\",\"https://api.llm-stats.com/leaderboard/benchmarks/aa-lcr\"]", + "raw_score_id": "aa-lcr::qwen3.5-9b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bfcl-v4::bfcl-v4-qwen3.5-9b", + "evaluation_name": "llm_stats.bfcl-v4", + "source_data": { + "dataset_name": "BFCL-V4", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-9b", + "https://llm-stats.com/benchmarks/bfcl-v4", + "https://api.llm-stats.com/leaderboard/benchmarks/bfcl-v4" + ], + "additional_details": { + "raw_benchmark_id": "bfcl-v4", + "raw_model_id": "qwen3.5-9b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Berkeley Function Calling Leaderboard V4 (BFCL-V4) evaluates LLMs on their ability to accurately call functions and APIs, including simple, multiple, parallel, and nested function calls across diverse programming scenarios.", + "metric_id": "llm_stats.bfcl-v4.score", + "metric_name": "BFCL-V4 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bfcl-v4", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BFCL-V4", + "raw_categories": "[\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.661, + "details": { + "raw_score": "0.661", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-9b", + "raw_benchmark_id": "bfcl-v4", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-9b\",\"https://llm-stats.com/benchmarks/bfcl-v4\",\"https://api.llm-stats.com/leaderboard/benchmarks/bfcl-v4\"]", + "raw_score_id": "bfcl-v4::qwen3.5-9b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "c-eval::c-eval-qwen3.5-9b", + "evaluation_name": "llm_stats.c-eval", + "source_data": { + "dataset_name": "C-Eval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-9b", + "https://llm-stats.com/benchmarks/c-eval", + "https://api.llm-stats.com/leaderboard/benchmarks/c-eval" + ], + "additional_details": { + "raw_benchmark_id": "c-eval", + "raw_model_id": "qwen3.5-9b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "C-Eval is a comprehensive Chinese evaluation suite designed to assess advanced knowledge and reasoning abilities of foundation models in a Chinese context. It comprises 13,948 multiple-choice questions across 52 diverse disciplines spanning humanities, science, and engineering, with four difficulty levels: middle school, high school, college, and professional. The benchmark includes C-Eval Hard, a subset of very challenging subjects requiring advanced reasoning abilities.", + "metric_id": "llm_stats.c-eval.score", + "metric_name": "C-Eval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "c-eval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "C-Eval", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.882, + "details": { + "raw_score": "0.882", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-9b", + "raw_benchmark_id": "c-eval", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-9b\",\"https://llm-stats.com/benchmarks/c-eval\",\"https://api.llm-stats.com/leaderboard/benchmarks/c-eval\"]", + "raw_score_id": "c-eval::qwen3.5-9b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "deep-planning::deep-planning-qwen3.5-9b", + "evaluation_name": "llm_stats.deep-planning", + "source_data": { + "dataset_name": "DeepPlanning", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-9b", + "https://llm-stats.com/benchmarks/deep-planning", + "https://api.llm-stats.com/leaderboard/benchmarks/deep-planning" + ], + "additional_details": { + "raw_benchmark_id": "deep-planning", + "raw_model_id": "qwen3.5-9b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "DeepPlanning evaluates LLMs on complex multi-step planning tasks requiring long-horizon reasoning, goal decomposition, and strategic decision-making.", + "metric_id": "llm_stats.deep-planning.score", + "metric_name": "DeepPlanning score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "deep-planning", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DeepPlanning", + "raw_categories": "[\"reasoning\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.18, + "details": { + "raw_score": "0.18", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-9b", + "raw_benchmark_id": "deep-planning", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-9b\",\"https://llm-stats.com/benchmarks/deep-planning\",\"https://api.llm-stats.com/leaderboard/benchmarks/deep-planning\"]", + "raw_score_id": "deep-planning::qwen3.5-9b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "global-piqa::global-piqa-qwen3.5-9b", + "evaluation_name": "llm_stats.global-piqa", + "source_data": { + "dataset_name": "Global PIQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-9b", + "https://llm-stats.com/benchmarks/global-piqa", + "https://api.llm-stats.com/leaderboard/benchmarks/global-piqa" + ], + "additional_details": { + "raw_benchmark_id": "global-piqa", + "raw_model_id": "qwen3.5-9b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Global PIQA is a multilingual commonsense reasoning benchmark that evaluates physical interaction knowledge across 100 languages and cultures. It tests AI systems' understanding of physical world knowledge in diverse cultural contexts through multiple choice questions about everyday situations requiring physical commonsense.", + "metric_id": "llm_stats.global-piqa.score", + "metric_name": "Global PIQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "global-piqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Global PIQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.832, + "details": { + "raw_score": "0.832", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-9b", + "raw_benchmark_id": "global-piqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-9b\",\"https://llm-stats.com/benchmarks/global-piqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/global-piqa\"]", + "raw_score_id": "global-piqa::qwen3.5-9b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hmmt25::hmmt25-qwen3.5-9b", + "evaluation_name": "llm_stats.hmmt25", + "source_data": { + "dataset_name": "HMMT25", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-9b", + "https://llm-stats.com/benchmarks/hmmt25", + "https://api.llm-stats.com/leaderboard/benchmarks/hmmt25" + ], + "additional_details": { + "raw_benchmark_id": "hmmt25", + "raw_model_id": "qwen3.5-9b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Harvard-MIT Mathematics Tournament 2025 - A prestigious student-organized mathematics competition for high school students featuring two tournaments (November 2025 at MIT and February 2026 at Harvard) with individual tests, team rounds, and guts rounds", + "metric_id": "llm_stats.hmmt25.score", + "metric_name": "HMMT25 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hmmt25", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HMMT25", + "raw_categories": "[\"math\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "25" + } + }, + "score_details": { + "score": 0.829, + "details": { + "raw_score": "0.829", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-9b", + "raw_benchmark_id": "hmmt25", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-9b\",\"https://llm-stats.com/benchmarks/hmmt25\",\"https://api.llm-stats.com/leaderboard/benchmarks/hmmt25\"]", + "raw_score_id": "hmmt25::qwen3.5-9b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ifeval::ifeval-qwen3.5-9b", + "evaluation_name": "llm_stats.ifeval", + "source_data": { + "dataset_name": "IFEval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-9b", + "https://llm-stats.com/benchmarks/ifeval", + "https://api.llm-stats.com/leaderboard/benchmarks/ifeval" + ], + "additional_details": { + "raw_benchmark_id": "ifeval", + "raw_model_id": "qwen3.5-9b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Instruction-Following Evaluation (IFEval) benchmark for large language models, focusing on verifiable instructions with 25 types of instructions and around 500 prompts containing one or more verifiable constraints", + "metric_id": "llm_stats.ifeval.score", + "metric_name": "IFEval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ifeval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IFEval", + "raw_categories": "[\"instruction_following\",\"structured_output\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "65" + } + }, + "score_details": { + "score": 0.915, + "details": { + "raw_score": "0.915", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-9b", + "raw_benchmark_id": "ifeval", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-9b\",\"https://llm-stats.com/benchmarks/ifeval\",\"https://api.llm-stats.com/leaderboard/benchmarks/ifeval\"]", + "raw_score_id": "ifeval::qwen3.5-9b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "include::include-qwen3.5-9b", + "evaluation_name": "llm_stats.include", + "source_data": { + "dataset_name": "Include", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-9b", + "https://llm-stats.com/benchmarks/include", + "https://api.llm-stats.com/leaderboard/benchmarks/include" + ], + "additional_details": { + "raw_benchmark_id": "include", + "raw_model_id": "qwen3.5-9b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Include benchmark - specific documentation not found in official sources", + "metric_id": "llm_stats.include.score", + "metric_name": "Include score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "include", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Include", + "raw_categories": "[\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "31" + } + }, + "score_details": { + "score": 0.756, + "details": { + "raw_score": "0.756", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-9b", + "raw_benchmark_id": "include", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-9b\",\"https://llm-stats.com/benchmarks/include\",\"https://api.llm-stats.com/leaderboard/benchmarks/include\"]", + "raw_score_id": "include::qwen3.5-9b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "longbench-v2::longbench-v2-qwen3.5-9b", + "evaluation_name": "llm_stats.longbench-v2", + "source_data": { + "dataset_name": "LongBench v2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-9b", + "https://llm-stats.com/benchmarks/longbench-v2", + "https://api.llm-stats.com/leaderboard/benchmarks/longbench-v2" + ], + "additional_details": { + "raw_benchmark_id": "longbench-v2", + "raw_model_id": "qwen3.5-9b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LongBench v2 is a benchmark designed to assess the ability of LLMs to handle long-context problems requiring deep understanding and reasoning across real-world multitasks. It consists of 503 challenging multiple-choice questions with contexts ranging from 8k to 2M words across six major task categories: single-document QA, multi-document QA, long in-context learning, long-dialogue history understanding, code repository understanding, and long structured data understanding.", + "metric_id": "llm_stats.longbench-v2.score", + "metric_name": "LongBench v2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "longbench-v2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LongBench v2", + "raw_categories": "[\"long_context\",\"reasoning\",\"structured_output\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.552, + "details": { + "raw_score": "0.552", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-9b", + "raw_benchmark_id": "longbench-v2", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-9b\",\"https://llm-stats.com/benchmarks/longbench-v2\",\"https://api.llm-stats.com/leaderboard/benchmarks/longbench-v2\"]", + "raw_score_id": "longbench-v2::qwen3.5-9b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "maxife::maxife-qwen3.5-9b", + "evaluation_name": "llm_stats.maxife", + "source_data": { + "dataset_name": "MAXIFE", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-9b", + "https://llm-stats.com/benchmarks/maxife", + "https://api.llm-stats.com/leaderboard/benchmarks/maxife" + ], + "additional_details": { + "raw_benchmark_id": "maxife", + "raw_model_id": "qwen3.5-9b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MAXIFE is a multilingual benchmark evaluating LLMs on instruction following and execution across multiple languages and cultural contexts.", + "metric_id": "llm_stats.maxife.score", + "metric_name": "MAXIFE score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "maxife", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MAXIFE", + "raw_categories": "[\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.834, + "details": { + "raw_score": "0.834", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-9b", + "raw_benchmark_id": "maxife", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-9b\",\"https://llm-stats.com/benchmarks/maxife\",\"https://api.llm-stats.com/leaderboard/benchmarks/maxife\"]", + "raw_score_id": "maxife::qwen3.5-9b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu-prox::mmlu-prox-qwen3.5-9b", + "evaluation_name": "llm_stats.mmlu-prox", + "source_data": { + "dataset_name": "MMLU-ProX", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-9b", + "https://llm-stats.com/benchmarks/mmlu-prox", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu-prox" + ], + "additional_details": { + "raw_benchmark_id": "mmlu-prox", + "raw_model_id": "qwen3.5-9b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Extended version of MMLU-Pro providing additional challenging multiple-choice questions for evaluating language models across diverse academic and professional domains. Built on the foundation of the Massive Multitask Language Understanding benchmark framework.", + "metric_id": "llm_stats.mmlu-prox.score", + "metric_name": "MMLU-ProX score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu-prox", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU-ProX", + "raw_categories": "[\"language\",\"legal\",\"math\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "32" + } + }, + "score_details": { + "score": 0.763, + "details": { + "raw_score": "0.763", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-9b", + "raw_benchmark_id": "mmlu-prox", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-9b\",\"https://llm-stats.com/benchmarks/mmlu-prox\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu-prox\"]", + "raw_score_id": "mmlu-prox::qwen3.5-9b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "multichallenge::multichallenge-qwen3.5-9b", + "evaluation_name": "llm_stats.multichallenge", + "source_data": { + "dataset_name": "Multi-Challenge", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-9b", + "https://llm-stats.com/benchmarks/multichallenge", + "https://api.llm-stats.com/leaderboard/benchmarks/multichallenge" + ], + "additional_details": { + "raw_benchmark_id": "multichallenge", + "raw_model_id": "qwen3.5-9b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MultiChallenge is a realistic multi-turn conversation evaluation benchmark that challenges frontier LLMs across four key categories: instruction retention (maintaining instructions throughout conversations), inference memory (recalling and connecting details from previous turns), reliable versioned editing (adapting to evolving instructions during collaborative editing), and self-coherence (avoiding contradictions in responses). The benchmark evaluates models on sustained, contextually complex dialogues across diverse topics including travel planning, technical documentation, and professional communication.", + "metric_id": "llm_stats.multichallenge.score", + "metric_name": "Multi-Challenge score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "multichallenge", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Multi-Challenge", + "raw_categories": "[\"reasoning\",\"communication\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "29" + } + }, + "score_details": { + "score": 0.545, + "details": { + "raw_score": "0.545", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-9b", + "raw_benchmark_id": "multichallenge", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-9b\",\"https://llm-stats.com/benchmarks/multichallenge\",\"https://api.llm-stats.com/leaderboard/benchmarks/multichallenge\"]", + "raw_score_id": "multichallenge::qwen3.5-9b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "nova-63::nova-63-qwen3.5-9b", + "evaluation_name": "llm_stats.nova-63", + "source_data": { + "dataset_name": "NOVA-63", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-9b", + "https://llm-stats.com/benchmarks/nova-63", + "https://api.llm-stats.com/leaderboard/benchmarks/nova-63" + ], + "additional_details": { + "raw_benchmark_id": "nova-63", + "raw_model_id": "qwen3.5-9b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "NOVA-63 is a multilingual evaluation benchmark covering 63 languages, designed to assess LLM performance across diverse linguistic contexts and tasks.", + "metric_id": "llm_stats.nova-63.score", + "metric_name": "NOVA-63 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "nova-63", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "NOVA-63", + "raw_categories": "[\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.559, + "details": { + "raw_score": "0.559", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-9b", + "raw_benchmark_id": "nova-63", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-9b\",\"https://llm-stats.com/benchmarks/nova-63\",\"https://api.llm-stats.com/leaderboard/benchmarks/nova-63\"]", + "raw_score_id": "nova-63::qwen3.5-9b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "polymath::polymath-qwen3.5-9b", + "evaluation_name": "llm_stats.polymath", + "source_data": { + "dataset_name": "PolyMATH", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-9b", + "https://llm-stats.com/benchmarks/polymath", + "https://api.llm-stats.com/leaderboard/benchmarks/polymath" + ], + "additional_details": { + "raw_benchmark_id": "polymath", + "raw_model_id": "qwen3.5-9b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Polymath is a challenging multi-modal mathematical reasoning benchmark designed to evaluate the general cognitive reasoning abilities of Multi-modal Large Language Models (MLLMs). The benchmark comprises 5,000 manually collected high-quality images of cognitive textual and visual challenges across 10 distinct categories, including pattern recognition, spatial reasoning, and relative reasoning.", + "metric_id": "llm_stats.polymath.score", + "metric_name": "PolyMATH score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "polymath", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PolyMATH", + "raw_categories": "[\"math\",\"multimodal\",\"reasoning\",\"spatial_reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.573, + "details": { + "raw_score": "0.573", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-9b", + "raw_benchmark_id": "polymath", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-9b\",\"https://llm-stats.com/benchmarks/polymath\",\"https://api.llm-stats.com/leaderboard/benchmarks/polymath\"]", + "raw_score_id": "polymath::qwen3.5-9b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "supergpqa::supergpqa-qwen3.5-9b", + "evaluation_name": "llm_stats.supergpqa", + "source_data": { + "dataset_name": "SuperGPQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-9b", + "https://llm-stats.com/benchmarks/supergpqa", + "https://api.llm-stats.com/leaderboard/benchmarks/supergpqa" + ], + "additional_details": { + "raw_benchmark_id": "supergpqa", + "raw_model_id": "qwen3.5-9b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SuperGPQA is a comprehensive benchmark that evaluates large language models across 285 graduate-level academic disciplines. The benchmark contains 25,957 questions covering 13 broad disciplinary areas including Engineering, Medicine, Science, and Law, with specialized fields in light industry, agriculture, and service-oriented domains. It employs a Human-LLM collaborative filtering mechanism with over 80 expert annotators to create challenging questions that assess graduate-level knowledge and reasoning capabilities.", + "metric_id": "llm_stats.supergpqa.score", + "metric_name": "SuperGPQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "supergpqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SuperGPQA", + "raw_categories": "[\"legal\",\"math\",\"physics\",\"reasoning\",\"finance\",\"general\",\"healthcare\",\"chemistry\",\"economics\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "34" + } + }, + "score_details": { + "score": 0.582, + "details": { + "raw_score": "0.582", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-9b", + "raw_benchmark_id": "supergpqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-9b\",\"https://llm-stats.com/benchmarks/supergpqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/supergpqa\"]", + "raw_score_id": "supergpqa::qwen3.5-9b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "t2-bench::t2-bench-qwen3.5-9b", + "evaluation_name": "llm_stats.t2-bench", + "source_data": { + "dataset_name": "t2-bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-9b", + "https://llm-stats.com/benchmarks/t2-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/t2-bench" + ], + "additional_details": { + "raw_benchmark_id": "t2-bench", + "raw_model_id": "qwen3.5-9b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "t2-bench is a benchmark for evaluating agentic tool use capabilities, measuring how well models can select, sequence, and utilize tools to solve complex tasks. It tests autonomous planning and execution in multi-step scenarios.", + "metric_id": "llm_stats.t2-bench.score", + "metric_name": "t2-bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "t2-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "t2-bench", + "raw_categories": "[\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.791, + "details": { + "raw_score": "0.791", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-9b", + "raw_benchmark_id": "t2-bench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-9b\",\"https://llm-stats.com/benchmarks/t2-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/t2-bench\"]", + "raw_score_id": "t2-bench::qwen3.5-9b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "vita-bench::vita-bench-qwen3.5-9b", + "evaluation_name": "llm_stats.vita-bench", + "source_data": { + "dataset_name": "VITA-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-9b", + "https://llm-stats.com/benchmarks/vita-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/vita-bench" + ], + "additional_details": { + "raw_benchmark_id": "vita-bench", + "raw_model_id": "qwen3.5-9b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VITA-Bench evaluates AI agents on real-world virtual task automation, measuring their ability to complete complex multi-step tasks in simulated environments.", + "metric_id": "llm_stats.vita-bench.score", + "metric_name": "VITA-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "vita-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VITA-Bench", + "raw_categories": "[\"reasoning\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.298, + "details": { + "raw_score": "0.298", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-9b", + "raw_benchmark_id": "vita-bench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-9b\",\"https://llm-stats.com/benchmarks/vita-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/vita-bench\"]", + "raw_score_id": "vita-bench::qwen3.5-9b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "wmt24::wmt24-qwen3.5-9b", + "evaluation_name": "llm_stats.wmt24", + "source_data": { + "dataset_name": "WMT24++", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.5-9b", + "https://llm-stats.com/benchmarks/wmt24", + "https://api.llm-stats.com/leaderboard/benchmarks/wmt24++" + ], + "additional_details": { + "raw_benchmark_id": "wmt24++", + "raw_model_id": "qwen3.5-9b", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WMT24++ is a comprehensive multilingual machine translation benchmark that expands the WMT24 dataset to cover 55 languages and dialects. It includes human-written references and post-edits across four domains (literary, news, social, and speech) to evaluate machine translation systems and large language models across diverse linguistic contexts.", + "metric_id": "llm_stats.wmt24.score", + "metric_name": "WMT24++ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "wmt24++", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WMT24++", + "raw_categories": "[\"language\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.726, + "details": { + "raw_score": "0.726", + "raw_score_field": "score", + "raw_model_id": "qwen3.5-9b", + "raw_benchmark_id": "wmt24++", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.5-9b\",\"https://llm-stats.com/benchmarks/wmt24\",\"https://api.llm-stats.com/leaderboard/benchmarks/wmt24++\"]", + "raw_score_id": "wmt24++::qwen3.5-9b", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/qwen/qwen3.6-plus/51f9c782-d9c5-43b3-a48f-f7861f9439e9.json b/data/llm_stats/qwen/qwen3.6-plus/51f9c782-d9c5-43b3-a48f-f7861f9439e9.json new file mode 100644 index 0000000000000000000000000000000000000000..9fa5873d03fa4b057dc83eebb9fb901a2c6ad6ca --- /dev/null +++ b/data/llm_stats/qwen/qwen3.6-plus/51f9c782-d9c5-43b3-a48f-f7861f9439e9.json @@ -0,0 +1,3220 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/qwen_qwen3.6-plus/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.6-plus", + "id": "qwen/qwen3.6-plus", + "developer": "qwen", + "additional_details": { + "raw_id": "qwen3.6-plus", + "raw_model_name": "qwen3.6-plus", + "raw_display_name": "Qwen3.6 Plus", + "raw_organization_id": "qwen", + "raw_organization_name": "Alibaba Cloud / Qwen Team", + "raw_context_length": "1000000", + "raw_input_modalities": "[\"image\",\"text\",\"video\"]", + "raw_output_modalities": "[\"text\"]", + "raw_input_price": "0.5", + "raw_output_price": "3.0", + "raw_model_id": "qwen3.6-plus", + "raw_provider_slug": "qwen", + "raw_provider_name": "ZeroEval" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aa-lcr::aa-lcr-qwen3.6-plus", + "evaluation_name": "llm_stats.aa-lcr", + "source_data": { + "dataset_name": "AA-LCR", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/aa-lcr", + "https://api.llm-stats.com/leaderboard/benchmarks/aa-lcr" + ], + "additional_details": { + "raw_benchmark_id": "aa-lcr", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Agent Arena Long Context Reasoning benchmark", + "metric_id": "llm_stats.aa-lcr.score", + "metric_name": "AA-LCR score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aa-lcr", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AA-LCR", + "raw_categories": "[\"long_context\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "14" + } + }, + "score_details": { + "score": 0.683, + "details": { + "raw_score": "0.683", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "aa-lcr", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/aa-lcr\",\"https://api.llm-stats.com/leaderboard/benchmarks/aa-lcr\"]", + "raw_score_id": "aa-lcr::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ai2d::ai2d-qwen3.6-plus", + "evaluation_name": "llm_stats.ai2d", + "source_data": { + "dataset_name": "AI2D", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/ai2d", + "https://api.llm-stats.com/leaderboard/benchmarks/ai2d" + ], + "additional_details": { + "raw_benchmark_id": "ai2d", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "AI2D is a dataset of 4,903 illustrative diagrams from grade school natural sciences (such as food webs, human physiology, and life cycles) with over 15,000 multiple choice questions and answers. The benchmark evaluates diagram understanding and visual reasoning capabilities, requiring models to interpret diagrammatic elements, relationships, and structure to answer questions about scientific concepts represented in visual form.", + "metric_id": "llm_stats.ai2d.score", + "metric_name": "AI2D score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ai2d", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AI2D", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "32" + } + }, + "score_details": { + "score": 0.944, + "details": { + "raw_score": "0.944", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "ai2d", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/ai2d\",\"https://api.llm-stats.com/leaderboard/benchmarks/ai2d\"]", + "raw_score_id": "ai2d::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "aime-2026::aime-2026-qwen3.6-plus", + "evaluation_name": "llm_stats.aime-2026", + "source_data": { + "dataset_name": "AIME 2026", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/aime-2026", + "https://api.llm-stats.com/leaderboard/benchmarks/aime-2026" + ], + "additional_details": { + "raw_benchmark_id": "aime-2026", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "All 30 problems from the 2026 American Invitational Mathematics Examination (AIME I and AIME II), testing olympiad-level mathematical reasoning with integer answers from 000-999. Used as an AI benchmark to evaluate large language models' ability to solve complex mathematical problems requiring multi-step logical deductions and structured symbolic reasoning.", + "metric_id": "llm_stats.aime-2026.score", + "metric_name": "AIME 2026 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aime-2026", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AIME 2026", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "17" + } + }, + "score_details": { + "score": 0.953, + "details": { + "raw_score": "0.953", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "aime-2026", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/aime-2026\",\"https://api.llm-stats.com/leaderboard/benchmarks/aime-2026\"]", + "raw_score_id": "aime-2026::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "c-eval::c-eval-qwen3.6-plus", + "evaluation_name": "llm_stats.c-eval", + "source_data": { + "dataset_name": "C-Eval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/c-eval", + "https://api.llm-stats.com/leaderboard/benchmarks/c-eval" + ], + "additional_details": { + "raw_benchmark_id": "c-eval", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "C-Eval is a comprehensive Chinese evaluation suite designed to assess advanced knowledge and reasoning abilities of foundation models in a Chinese context. It comprises 13,948 multiple-choice questions across 52 diverse disciplines spanning humanities, science, and engineering, with four difficulty levels: middle school, high school, college, and professional. The benchmark includes C-Eval Hard, a subset of very challenging subjects requiring advanced reasoning abilities.", + "metric_id": "llm_stats.c-eval.score", + "metric_name": "C-Eval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "c-eval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "C-Eval", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.933, + "details": { + "raw_score": "0.933", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "c-eval", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/c-eval\",\"https://api.llm-stats.com/leaderboard/benchmarks/c-eval\"]", + "raw_score_id": "c-eval::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "cc-ocr::cc-ocr-qwen3.6-plus", + "evaluation_name": "llm_stats.cc-ocr", + "source_data": { + "dataset_name": "CC-OCR", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/cc-ocr", + "https://api.llm-stats.com/leaderboard/benchmarks/cc-ocr" + ], + "additional_details": { + "raw_benchmark_id": "cc-ocr", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A comprehensive OCR benchmark for evaluating Large Multimodal Models (LMMs) in literacy. Comprises four OCR-centric tracks: multi-scene text reading, multilingual text reading, document parsing, and key information extraction. Contains 39 subsets with 7,058 fully annotated images, 41% sourced from real applications. Tests capabilities including text grounding, multi-orientation text recognition, and detecting hallucination/repetition across diverse visual challenges.", + "metric_id": "llm_stats.cc-ocr.score", + "metric_name": "CC-OCR score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "cc-ocr", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CC-OCR", + "raw_categories": "[\"multimodal\",\"structured_output\",\"text-to-image\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.834, + "details": { + "raw_score": "0.834", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "cc-ocr", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/cc-ocr\",\"https://api.llm-stats.com/leaderboard/benchmarks/cc-ocr\"]", + "raw_score_id": "cc-ocr::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "charxiv-r::charxiv-r-qwen3.6-plus", + "evaluation_name": "llm_stats.charxiv-r", + "source_data": { + "dataset_name": "CharXiv-R", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/charxiv-r", + "https://api.llm-stats.com/leaderboard/benchmarks/charxiv-r" + ], + "additional_details": { + "raw_benchmark_id": "charxiv-r", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CharXiv-R is the reasoning component of the CharXiv benchmark, focusing on complex reasoning questions that require synthesizing information across visual chart elements. It evaluates multimodal large language models on their ability to understand and reason about scientific charts from arXiv papers through various reasoning tasks.", + "metric_id": "llm_stats.charxiv-r.score", + "metric_name": "CharXiv-R score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "charxiv-r", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CharXiv-R", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "43" + } + }, + "score_details": { + "score": 0.815, + "details": { + "raw_score": "0.815", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "charxiv-r", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/charxiv-r\",\"https://api.llm-stats.com/leaderboard/benchmarks/charxiv-r\"]", + "raw_score_id": "charxiv-r::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "claw-eval::claw-eval-qwen3.6-plus", + "evaluation_name": "llm_stats.claw-eval", + "source_data": { + "dataset_name": "Claw-Eval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/claw-eval", + "https://api.llm-stats.com/leaderboard/benchmarks/claw-eval" + ], + "additional_details": { + "raw_benchmark_id": "claw-eval", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Claw-Eval tests real-world agentic task completion across complex multi-step scenarios, evaluating a model's ability to use tools, navigate environments, and complete end-to-end tasks autonomously.", + "metric_id": "llm_stats.claw-eval.score", + "metric_name": "Claw-Eval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "claw-eval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Claw-Eval", + "raw_categories": "[\"agents\",\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.587, + "details": { + "raw_score": "0.587", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "claw-eval", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/claw-eval\",\"https://api.llm-stats.com/leaderboard/benchmarks/claw-eval\"]", + "raw_score_id": "claw-eval::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "countbench::countbench-qwen3.6-plus", + "evaluation_name": "llm_stats.countbench", + "source_data": { + "dataset_name": "CountBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/countbench", + "https://api.llm-stats.com/leaderboard/benchmarks/countbench" + ], + "additional_details": { + "raw_benchmark_id": "countbench", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CountBench evaluates object counting capabilities in visual understanding.", + "metric_id": "llm_stats.countbench.score", + "metric_name": "CountBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "countbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CountBench", + "raw_categories": "[\"reasoning\",\"spatial_reasoning\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.976, + "details": { + "raw_score": "0.976", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "countbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/countbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/countbench\"]", + "raw_score_id": "countbench::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "deep-planning::deep-planning-qwen3.6-plus", + "evaluation_name": "llm_stats.deep-planning", + "source_data": { + "dataset_name": "DeepPlanning", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/deep-planning", + "https://api.llm-stats.com/leaderboard/benchmarks/deep-planning" + ], + "additional_details": { + "raw_benchmark_id": "deep-planning", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "DeepPlanning evaluates LLMs on complex multi-step planning tasks requiring long-horizon reasoning, goal decomposition, and strategic decision-making.", + "metric_id": "llm_stats.deep-planning.score", + "metric_name": "DeepPlanning score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "deep-planning", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DeepPlanning", + "raw_categories": "[\"reasoning\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.415, + "details": { + "raw_score": "0.415", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "deep-planning", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/deep-planning\",\"https://api.llm-stats.com/leaderboard/benchmarks/deep-planning\"]", + "raw_score_id": "deep-planning::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "dynamath::dynamath-qwen3.6-plus", + "evaluation_name": "llm_stats.dynamath", + "source_data": { + "dataset_name": "DynaMath", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/dynamath", + "https://api.llm-stats.com/leaderboard/benchmarks/dynamath" + ], + "additional_details": { + "raw_benchmark_id": "dynamath", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multimodal mathematics and reasoning benchmark focused on dynamic visual problem solving.", + "metric_id": "llm_stats.dynamath.score", + "metric_name": "DynaMath score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "dynamath", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DynaMath", + "raw_categories": "[\"math\",\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.88, + "details": { + "raw_score": "0.88", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "dynamath", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/dynamath\",\"https://api.llm-stats.com/leaderboard/benchmarks/dynamath\"]", + "raw_score_id": "dynamath::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "erqa::erqa-qwen3.6-plus", + "evaluation_name": "llm_stats.erqa", + "source_data": { + "dataset_name": "ERQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/erqa", + "https://api.llm-stats.com/leaderboard/benchmarks/erqa" + ], + "additional_details": { + "raw_benchmark_id": "erqa", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Embodied Reasoning Question Answering benchmark consisting of 400 multiple-choice visual questions across spatial reasoning, trajectory reasoning, action reasoning, state estimation, and multi-view reasoning for evaluating AI capabilities in physical world interactions", + "metric_id": "llm_stats.erqa.score", + "metric_name": "ERQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "erqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ERQA", + "raw_categories": "[\"reasoning\",\"spatial_reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "22" + } + }, + "score_details": { + "score": 0.657, + "details": { + "raw_score": "0.657", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "erqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/erqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/erqa\"]", + "raw_score_id": "erqa::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "global-piqa::global-piqa-qwen3.6-plus", + "evaluation_name": "llm_stats.global-piqa", + "source_data": { + "dataset_name": "Global PIQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/global-piqa", + "https://api.llm-stats.com/leaderboard/benchmarks/global-piqa" + ], + "additional_details": { + "raw_benchmark_id": "global-piqa", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Global PIQA is a multilingual commonsense reasoning benchmark that evaluates physical interaction knowledge across 100 languages and cultures. It tests AI systems' understanding of physical world knowledge in diverse cultural contexts through multiple choice questions about everyday situations requiring physical commonsense.", + "metric_id": "llm_stats.global-piqa.score", + "metric_name": "Global PIQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "global-piqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Global PIQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.898, + "details": { + "raw_score": "0.898", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "global-piqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/global-piqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/global-piqa\"]", + "raw_score_id": "global-piqa::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gpqa::gpqa-qwen3.6-plus", + "evaluation_name": "llm_stats.gpqa", + "source_data": { + "dataset_name": "GPQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/gpqa", + "https://api.llm-stats.com/leaderboard/benchmarks/gpqa" + ], + "additional_details": { + "raw_benchmark_id": "gpqa", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A challenging dataset of 448 multiple-choice questions written by domain experts in biology, physics, and chemistry. Questions are Google-proof and extremely difficult, with PhD experts reaching 65% accuracy.", + "metric_id": "llm_stats.gpqa.score", + "metric_name": "GPQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gpqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GPQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\",\"biology\",\"chemistry\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "226" + } + }, + "score_details": { + "score": 0.904, + "details": { + "raw_score": "0.904", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "gpqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/gpqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/gpqa\"]", + "raw_score_id": "gpqa::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hmmt-2025::hmmt-2025-qwen3.6-plus", + "evaluation_name": "llm_stats.hmmt-2025", + "source_data": { + "dataset_name": "HMMT 2025", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/hmmt-2025", + "https://api.llm-stats.com/leaderboard/benchmarks/hmmt-2025" + ], + "additional_details": { + "raw_benchmark_id": "hmmt-2025", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Harvard-MIT Mathematics Tournament 2025 - A prestigious student-organized mathematics competition for high school students featuring two tournaments (November 2025 at MIT and February 2026 at Harvard) with individual tests, team rounds, and guts rounds", + "metric_id": "llm_stats.hmmt-2025.score", + "metric_name": "HMMT 2025 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hmmt-2025", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HMMT 2025", + "raw_categories": "[\"math\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "33" + } + }, + "score_details": { + "score": 0.967, + "details": { + "raw_score": "0.967", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "hmmt-2025", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/hmmt-2025\",\"https://api.llm-stats.com/leaderboard/benchmarks/hmmt-2025\"]", + "raw_score_id": "hmmt-2025::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hmmt-feb-26::hmmt-feb-26-qwen3.6-plus", + "evaluation_name": "llm_stats.hmmt-feb-26", + "source_data": { + "dataset_name": "HMMT Feb 26", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/hmmt-feb-26", + "https://api.llm-stats.com/leaderboard/benchmarks/hmmt-feb-26" + ], + "additional_details": { + "raw_benchmark_id": "hmmt-feb-26", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "HMMT February 2026 is a math competition benchmark based on problems from the Harvard-MIT Mathematics Tournament, testing advanced mathematical problem-solving and reasoning.", + "metric_id": "llm_stats.hmmt-feb-26.score", + "metric_name": "HMMT Feb 26 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hmmt-feb-26", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HMMT Feb 26", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.878, + "details": { + "raw_score": "0.878", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "hmmt-feb-26", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/hmmt-feb-26\",\"https://api.llm-stats.com/leaderboard/benchmarks/hmmt-feb-26\"]", + "raw_score_id": "hmmt-feb-26::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hmmt25::hmmt25-qwen3.6-plus", + "evaluation_name": "llm_stats.hmmt25", + "source_data": { + "dataset_name": "HMMT25", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/hmmt25", + "https://api.llm-stats.com/leaderboard/benchmarks/hmmt25" + ], + "additional_details": { + "raw_benchmark_id": "hmmt25", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Harvard-MIT Mathematics Tournament 2025 - A prestigious student-organized mathematics competition for high school students featuring two tournaments (November 2025 at MIT and February 2026 at Harvard) with individual tests, team rounds, and guts rounds", + "metric_id": "llm_stats.hmmt25.score", + "metric_name": "HMMT25 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hmmt25", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HMMT25", + "raw_categories": "[\"math\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "25" + } + }, + "score_details": { + "score": 0.946, + "details": { + "raw_score": "0.946", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "hmmt25", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/hmmt25\",\"https://api.llm-stats.com/leaderboard/benchmarks/hmmt25\"]", + "raw_score_id": "hmmt25::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ifbench::ifbench-qwen3.6-plus", + "evaluation_name": "llm_stats.ifbench", + "source_data": { + "dataset_name": "IFBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/ifbench", + "https://api.llm-stats.com/leaderboard/benchmarks/ifbench" + ], + "additional_details": { + "raw_benchmark_id": "ifbench", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Instruction Following Benchmark evaluating model's ability to follow complex instructions", + "metric_id": "llm_stats.ifbench.score", + "metric_name": "IFBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ifbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IFBench", + "raw_categories": "[\"instruction_following\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.742, + "details": { + "raw_score": "0.742", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "ifbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/ifbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/ifbench\"]", + "raw_score_id": "ifbench::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ifeval::ifeval-qwen3.6-plus", + "evaluation_name": "llm_stats.ifeval", + "source_data": { + "dataset_name": "IFEval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/ifeval", + "https://api.llm-stats.com/leaderboard/benchmarks/ifeval" + ], + "additional_details": { + "raw_benchmark_id": "ifeval", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Instruction-Following Evaluation (IFEval) benchmark for large language models, focusing on verifiable instructions with 25 types of instructions and around 500 prompts containing one or more verifiable constraints", + "metric_id": "llm_stats.ifeval.score", + "metric_name": "IFEval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ifeval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IFEval", + "raw_categories": "[\"instruction_following\",\"structured_output\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "65" + } + }, + "score_details": { + "score": 0.943, + "details": { + "raw_score": "0.943", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "ifeval", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/ifeval\",\"https://api.llm-stats.com/leaderboard/benchmarks/ifeval\"]", + "raw_score_id": "ifeval::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "imo-answerbench::imo-answerbench-qwen3.6-plus", + "evaluation_name": "llm_stats.imo-answerbench", + "source_data": { + "dataset_name": "IMO-AnswerBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/imo-answerbench", + "https://api.llm-stats.com/leaderboard/benchmarks/imo-answerbench" + ], + "additional_details": { + "raw_benchmark_id": "imo-answerbench", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "IMO-AnswerBench is a benchmark for evaluating mathematical reasoning capabilities on International Mathematical Olympiad (IMO) problems, focusing on answer generation and verification.", + "metric_id": "llm_stats.imo-answerbench.score", + "metric_name": "IMO-AnswerBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "imo-answerbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IMO-AnswerBench", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.838, + "details": { + "raw_score": "0.838", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "imo-answerbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/imo-answerbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/imo-answerbench\"]", + "raw_score_id": "imo-answerbench::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "include::include-qwen3.6-plus", + "evaluation_name": "llm_stats.include", + "source_data": { + "dataset_name": "Include", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/include", + "https://api.llm-stats.com/leaderboard/benchmarks/include" + ], + "additional_details": { + "raw_benchmark_id": "include", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Include benchmark - specific documentation not found in official sources", + "metric_id": "llm_stats.include.score", + "metric_name": "Include score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "include", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Include", + "raw_categories": "[\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "31" + } + }, + "score_details": { + "score": 0.851, + "details": { + "raw_score": "0.851", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "include", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/include\",\"https://api.llm-stats.com/leaderboard/benchmarks/include\"]", + "raw_score_id": "include::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "livecodebench-v6::livecodebench-v6-qwen3.6-plus", + "evaluation_name": "llm_stats.livecodebench-v6", + "source_data": { + "dataset_name": "LiveCodeBench v6", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/livecodebench-v6", + "https://api.llm-stats.com/leaderboard/benchmarks/livecodebench-v6" + ], + "additional_details": { + "raw_benchmark_id": "livecodebench-v6", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LiveCodeBench is a holistic and contamination-free evaluation benchmark for large language models for code. It continuously collects new problems from programming contests (LeetCode, AtCoder, CodeForces) and evaluates four different scenarios: code generation, self-repair, code execution, and test output prediction. Problems are annotated with release dates to enable evaluation on unseen problems released after a model's training cutoff.", + "metric_id": "llm_stats.livecodebench-v6.score", + "metric_name": "LiveCodeBench v6 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "livecodebench-v6", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LiveCodeBench v6", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "53" + } + }, + "score_details": { + "score": 0.871, + "details": { + "raw_score": "0.871", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "livecodebench-v6", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/livecodebench-v6\",\"https://api.llm-stats.com/leaderboard/benchmarks/livecodebench-v6\"]", + "raw_score_id": "livecodebench-v6::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "longbench-v2::longbench-v2-qwen3.6-plus", + "evaluation_name": "llm_stats.longbench-v2", + "source_data": { + "dataset_name": "LongBench v2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/longbench-v2", + "https://api.llm-stats.com/leaderboard/benchmarks/longbench-v2" + ], + "additional_details": { + "raw_benchmark_id": "longbench-v2", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LongBench v2 is a benchmark designed to assess the ability of LLMs to handle long-context problems requiring deep understanding and reasoning across real-world multitasks. It consists of 503 challenging multiple-choice questions with contexts ranging from 8k to 2M words across six major task categories: single-document QA, multi-document QA, long in-context learning, long-dialogue history understanding, code repository understanding, and long structured data understanding.", + "metric_id": "llm_stats.longbench-v2.score", + "metric_name": "LongBench v2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "longbench-v2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LongBench v2", + "raw_categories": "[\"long_context\",\"reasoning\",\"structured_output\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.62, + "details": { + "raw_score": "0.62", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "longbench-v2", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/longbench-v2\",\"https://api.llm-stats.com/leaderboard/benchmarks/longbench-v2\"]", + "raw_score_id": "longbench-v2::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mathvision::mathvision-qwen3.6-plus", + "evaluation_name": "llm_stats.mathvision", + "source_data": { + "dataset_name": "MathVision", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/mathvision", + "https://api.llm-stats.com/leaderboard/benchmarks/mathvision" + ], + "additional_details": { + "raw_benchmark_id": "mathvision", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MATH-Vision is a dataset designed to measure multimodal mathematical reasoning capabilities. It focuses on evaluating how well models can solve mathematical problems that require both visual understanding and mathematical reasoning, bridging the gap between visual and mathematical domains.", + "metric_id": "llm_stats.mathvision.score", + "metric_name": "MathVision score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mathvision", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MathVision", + "raw_categories": "[\"math\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "31" + } + }, + "score_details": { + "score": 0.88, + "details": { + "raw_score": "0.88", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "mathvision", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/mathvision\",\"https://api.llm-stats.com/leaderboard/benchmarks/mathvision\"]", + "raw_score_id": "mathvision::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "maxife::maxife-qwen3.6-plus", + "evaluation_name": "llm_stats.maxife", + "source_data": { + "dataset_name": "MAXIFE", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/maxife", + "https://api.llm-stats.com/leaderboard/benchmarks/maxife" + ], + "additional_details": { + "raw_benchmark_id": "maxife", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MAXIFE is a multilingual benchmark evaluating LLMs on instruction following and execution across multiple languages and cultural contexts.", + "metric_id": "llm_stats.maxife.score", + "metric_name": "MAXIFE score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "maxife", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MAXIFE", + "raw_categories": "[\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.882, + "details": { + "raw_score": "0.882", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "maxife", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/maxife\",\"https://api.llm-stats.com/leaderboard/benchmarks/maxife\"]", + "raw_score_id": "maxife::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mcp-atlas::mcp-atlas-qwen3.6-plus", + "evaluation_name": "llm_stats.mcp-atlas", + "source_data": { + "dataset_name": "MCP Atlas", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/mcp-atlas", + "https://api.llm-stats.com/leaderboard/benchmarks/mcp-atlas" + ], + "additional_details": { + "raw_benchmark_id": "mcp-atlas", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MCP Atlas is a benchmark for evaluating AI models on scaled tool use capabilities, measuring how well models can coordinate and utilize multiple tools across complex multi-step tasks.", + "metric_id": "llm_stats.mcp-atlas.score", + "metric_name": "MCP Atlas score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mcp-atlas", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MCP Atlas", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.741, + "details": { + "raw_score": "0.741", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "mcp-atlas", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/mcp-atlas\",\"https://api.llm-stats.com/leaderboard/benchmarks/mcp-atlas\"]", + "raw_score_id": "mcp-atlas::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mcp-mark::mcp-mark-qwen3.6-plus", + "evaluation_name": "llm_stats.mcp-mark", + "source_data": { + "dataset_name": "MCP-Mark", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/mcp-mark", + "https://api.llm-stats.com/leaderboard/benchmarks/mcp-mark" + ], + "additional_details": { + "raw_benchmark_id": "mcp-mark", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MCP-Mark evaluates LLMs on their ability to use Model Context Protocol (MCP) tools effectively, testing tool discovery, selection, invocation, and result interpretation across diverse MCP server scenarios.", + "metric_id": "llm_stats.mcp-mark.score", + "metric_name": "MCP-Mark score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mcp-mark", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MCP-Mark", + "raw_categories": "[\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.482, + "details": { + "raw_score": "0.482", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "mcp-mark", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/mcp-mark\",\"https://api.llm-stats.com/leaderboard/benchmarks/mcp-mark\"]", + "raw_score_id": "mcp-mark::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mlvu::mlvu-qwen3.6-plus", + "evaluation_name": "llm_stats.mlvu", + "source_data": { + "dataset_name": "MLVU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/mlvu", + "https://api.llm-stats.com/leaderboard/benchmarks/mlvu" + ], + "additional_details": { + "raw_benchmark_id": "mlvu", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A comprehensive benchmark for multi-task long video understanding that evaluates multimodal large language models on videos ranging from 3 minutes to 2 hours across 9 distinct tasks including reasoning, captioning, recognition, and summarization.", + "metric_id": "llm_stats.mlvu.score", + "metric_name": "MLVU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mlvu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MLVU", + "raw_categories": "[\"long_context\",\"multimodal\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.867, + "details": { + "raw_score": "0.867", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "mlvu", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/mlvu\",\"https://api.llm-stats.com/leaderboard/benchmarks/mlvu\"]", + "raw_score_id": "mlvu::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlongbench-doc::mmlongbench-doc-qwen3.6-plus", + "evaluation_name": "llm_stats.mmlongbench-doc", + "source_data": { + "dataset_name": "MMLongBench-Doc", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/mmlongbench-doc", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlongbench-doc" + ], + "additional_details": { + "raw_benchmark_id": "mmlongbench-doc", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MMLongBench-Doc evaluates long document understanding capabilities in vision-language models.", + "metric_id": "llm_stats.mmlongbench-doc.score", + "metric_name": "MMLongBench-Doc score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlongbench-doc", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLongBench-Doc", + "raw_categories": "[\"long_context\",\"multimodal\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.62, + "details": { + "raw_score": "0.62", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "mmlongbench-doc", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/mmlongbench-doc\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlongbench-doc\"]", + "raw_score_id": "mmlongbench-doc::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu-pro::mmlu-pro-qwen3.6-plus", + "evaluation_name": "llm_stats.mmlu-pro", + "source_data": { + "dataset_name": "MMLU-Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/mmlu-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu-pro" + ], + "additional_details": { + "raw_benchmark_id": "mmlu-pro", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A more robust and challenging multi-task language understanding benchmark that extends MMLU by expanding multiple-choice options from 4 to 10, eliminating trivial questions, and focusing on reasoning-intensive tasks. Features over 12,000 curated questions across 14 domains and causes a 16-33% accuracy drop compared to original MMLU.", + "metric_id": "llm_stats.mmlu-pro.score", + "metric_name": "MMLU-Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU-Pro", + "raw_categories": "[\"language\",\"legal\",\"math\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "129" + } + }, + "score_details": { + "score": 0.885, + "details": { + "raw_score": "0.885", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "mmlu-pro", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/mmlu-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu-pro\"]", + "raw_score_id": "mmlu-pro::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu-prox::mmlu-prox-qwen3.6-plus", + "evaluation_name": "llm_stats.mmlu-prox", + "source_data": { + "dataset_name": "MMLU-ProX", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/mmlu-prox", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu-prox" + ], + "additional_details": { + "raw_benchmark_id": "mmlu-prox", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Extended version of MMLU-Pro providing additional challenging multiple-choice questions for evaluating language models across diverse academic and professional domains. Built on the foundation of the Massive Multitask Language Understanding benchmark framework.", + "metric_id": "llm_stats.mmlu-prox.score", + "metric_name": "MMLU-ProX score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu-prox", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU-ProX", + "raw_categories": "[\"language\",\"legal\",\"math\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "32" + } + }, + "score_details": { + "score": 0.847, + "details": { + "raw_score": "0.847", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "mmlu-prox", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/mmlu-prox\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu-prox\"]", + "raw_score_id": "mmlu-prox::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu-redux::mmlu-redux-qwen3.6-plus", + "evaluation_name": "llm_stats.mmlu-redux", + "source_data": { + "dataset_name": "MMLU-Redux", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/mmlu-redux", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu-redux" + ], + "additional_details": { + "raw_benchmark_id": "mmlu-redux", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "An improved version of the MMLU benchmark featuring manually re-annotated questions to identify and correct errors in the original dataset. Provides more reliable evaluation metrics for language models by addressing dataset quality issues found in the original MMLU.", + "metric_id": "llm_stats.mmlu-redux.score", + "metric_name": "MMLU-Redux score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu-redux", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU-Redux", + "raw_categories": "[\"language\",\"math\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "48" + } + }, + "score_details": { + "score": 0.945, + "details": { + "raw_score": "0.945", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "mmlu-redux", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/mmlu-redux\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu-redux\"]", + "raw_score_id": "mmlu-redux::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmmlu::mmmlu-qwen3.6-plus", + "evaluation_name": "llm_stats.mmmlu", + "source_data": { + "dataset_name": "MMMLU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/mmmlu", + "https://api.llm-stats.com/leaderboard/benchmarks/mmmlu" + ], + "additional_details": { + "raw_benchmark_id": "mmmlu", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Multilingual Massive Multitask Language Understanding dataset released by OpenAI, featuring professionally translated MMLU test questions across 14 languages including Arabic, Bengali, German, Spanish, French, Hindi, Indonesian, Italian, Japanese, Korean, Portuguese, Swahili, Yoruba, and Chinese. Contains approximately 15,908 multiple-choice questions per language covering 57 subjects.", + "metric_id": "llm_stats.mmmlu.score", + "metric_name": "MMMLU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmmlu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMMLU", + "raw_categories": "[\"language\",\"math\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "49" + } + }, + "score_details": { + "score": 0.895, + "details": { + "raw_score": "0.895", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "mmmlu", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/mmmlu\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmmlu\"]", + "raw_score_id": "mmmlu::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmmu-pro::mmmu-pro-qwen3.6-plus", + "evaluation_name": "llm_stats.mmmu-pro", + "source_data": { + "dataset_name": "MMMU-Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/mmmu-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/mmmu-pro" + ], + "additional_details": { + "raw_benchmark_id": "mmmu-pro", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A more robust multi-discipline multimodal understanding benchmark that enhances MMMU through a three-step process: filtering text-only answerable questions, augmenting candidate options, and introducing vision-only input settings. Achieves significantly lower model performance (16.8-26.9%) compared to original MMMU, providing more rigorous evaluation that closely mimics real-world scenarios.", + "metric_id": "llm_stats.mmmu-pro.score", + "metric_name": "MMMU-Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmmu-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMMU-Pro", + "raw_categories": "[\"multimodal\",\"reasoning\",\"general\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "60" + } + }, + "score_details": { + "score": 0.788, + "details": { + "raw_score": "0.788", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "mmmu-pro", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/mmmu-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmmu-pro\"]", + "raw_score_id": "mmmu-pro::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmmu::mmmu-qwen3.6-plus", + "evaluation_name": "llm_stats.mmmu", + "source_data": { + "dataset_name": "MMMU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/mmmu", + "https://api.llm-stats.com/leaderboard/benchmarks/mmmu" + ], + "additional_details": { + "raw_benchmark_id": "mmmu", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MMMU (Massive Multi-discipline Multimodal Understanding) is a benchmark designed to evaluate multimodal models on college-level subject knowledge and deliberate reasoning. Contains 11.5K meticulously collected multimodal questions from college exams, quizzes, and textbooks, covering six core disciplines: Art & Design, Business, Science, Health & Medicine, Humanities & Social Science, and Tech & Engineering across 30 subjects and 183 subfields.", + "metric_id": "llm_stats.mmmu.score", + "metric_name": "MMMU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmmu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMMU", + "raw_categories": "[\"multimodal\",\"reasoning\",\"general\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "63" + } + }, + "score_details": { + "score": 0.86, + "details": { + "raw_score": "0.86", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "mmmu", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/mmmu\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmmu\"]", + "raw_score_id": "mmmu::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmstar::mmstar-qwen3.6-plus", + "evaluation_name": "llm_stats.mmstar", + "source_data": { + "dataset_name": "MMStar", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/mmstar", + "https://api.llm-stats.com/leaderboard/benchmarks/mmstar" + ], + "additional_details": { + "raw_benchmark_id": "mmstar", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MMStar is an elite vision-indispensable multimodal benchmark comprising 1,500 challenge samples meticulously selected by humans to evaluate 6 core capabilities and 18 detailed axes. The benchmark addresses issues of visual content unnecessity and unintentional data leakage in existing multimodal evaluations.", + "metric_id": "llm_stats.mmstar.score", + "metric_name": "MMStar score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmstar", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMStar", + "raw_categories": "[\"multimodal\",\"reasoning\",\"general\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "22" + } + }, + "score_details": { + "score": 0.833, + "details": { + "raw_score": "0.833", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "mmstar", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/mmstar\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmstar\"]", + "raw_score_id": "mmstar::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "nl2repo::nl2repo-qwen3.6-plus", + "evaluation_name": "llm_stats.nl2repo", + "source_data": { + "dataset_name": "NL2Repo", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/nl2repo", + "https://api.llm-stats.com/leaderboard/benchmarks/nl2repo" + ], + "additional_details": { + "raw_benchmark_id": "nl2repo", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "NL2Repo evaluates long-horizon coding capabilities including repository-level understanding, where models must generate or modify code across entire repositories from natural language specifications.", + "metric_id": "llm_stats.nl2repo.score", + "metric_name": "NL2Repo score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "nl2repo", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "NL2Repo", + "raw_categories": "[\"agents\",\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.379, + "details": { + "raw_score": "0.379", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "nl2repo", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/nl2repo\",\"https://api.llm-stats.com/leaderboard/benchmarks/nl2repo\"]", + "raw_score_id": "nl2repo::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "nova-63::nova-63-qwen3.6-plus", + "evaluation_name": "llm_stats.nova-63", + "source_data": { + "dataset_name": "NOVA-63", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/nova-63", + "https://api.llm-stats.com/leaderboard/benchmarks/nova-63" + ], + "additional_details": { + "raw_benchmark_id": "nova-63", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "NOVA-63 is a multilingual evaluation benchmark covering 63 languages, designed to assess LLM performance across diverse linguistic contexts and tasks.", + "metric_id": "llm_stats.nova-63.score", + "metric_name": "NOVA-63 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "nova-63", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "NOVA-63", + "raw_categories": "[\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.579, + "details": { + "raw_score": "0.579", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "nova-63", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/nova-63\",\"https://api.llm-stats.com/leaderboard/benchmarks/nova-63\"]", + "raw_score_id": "nova-63::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "odinw::odinw-qwen3.6-plus", + "evaluation_name": "llm_stats.odinw", + "source_data": { + "dataset_name": "ODinW", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/odinw", + "https://api.llm-stats.com/leaderboard/benchmarks/odinw" + ], + "additional_details": { + "raw_benchmark_id": "odinw", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Object Detection in the Wild (ODinW) benchmark for evaluating object detection models' task-level transfer ability across diverse real-world datasets in terms of prediction accuracy and adaptation efficiency", + "metric_id": "llm_stats.odinw.score", + "metric_name": "ODinW score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "odinw", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ODinW", + "raw_categories": "[\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.518, + "details": { + "raw_score": "0.518", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "odinw", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/odinw\",\"https://api.llm-stats.com/leaderboard/benchmarks/odinw\"]", + "raw_score_id": "odinw::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "omnidocbench-1.5::omnidocbench-1.5-qwen3.6-plus", + "evaluation_name": "llm_stats.omnidocbench-1.5", + "source_data": { + "dataset_name": "OmniDocBench 1.5", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/omnidocbench-1.5", + "https://api.llm-stats.com/leaderboard/benchmarks/omnidocbench-1.5" + ], + "additional_details": { + "raw_benchmark_id": "omnidocbench-1.5", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OmniDocBench 1.5 is a comprehensive benchmark for evaluating multimodal large language models on document understanding tasks, including OCR, document parsing, information extraction, and visual question answering across diverse document types. Lower Overall Edit Distance scores are better.", + "metric_id": "llm_stats.omnidocbench-1.5.score", + "metric_name": "OmniDocBench 1.5 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "omnidocbench-1.5", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OmniDocBench 1.5", + "raw_categories": "[\"multimodal\",\"reasoning\",\"structured_output\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.912, + "details": { + "raw_score": "0.912", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "omnidocbench-1.5", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/omnidocbench-1.5\",\"https://api.llm-stats.com/leaderboard/benchmarks/omnidocbench-1.5\"]", + "raw_score_id": "omnidocbench-1.5::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "osworld-verified::osworld-verified-qwen3.6-plus", + "evaluation_name": "llm_stats.osworld-verified", + "source_data": { + "dataset_name": "OSWorld-Verified", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/osworld-verified", + "https://api.llm-stats.com/leaderboard/benchmarks/osworld-verified" + ], + "additional_details": { + "raw_benchmark_id": "osworld-verified", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OSWorld-Verified is a verified subset of OSWorld, a scalable real computer environment for multimodal agents supporting task setup, execution-based evaluation, and interactive learning across Ubuntu, Windows, and macOS.", + "metric_id": "llm_stats.osworld-verified.score", + "metric_name": "OSWorld-Verified score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "osworld-verified", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OSWorld-Verified", + "raw_categories": "[\"multimodal\",\"general\",\"agents\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.625, + "details": { + "raw_score": "0.625", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "osworld-verified", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/osworld-verified\",\"https://api.llm-stats.com/leaderboard/benchmarks/osworld-verified\"]", + "raw_score_id": "osworld-verified::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "polymath::polymath-qwen3.6-plus", + "evaluation_name": "llm_stats.polymath", + "source_data": { + "dataset_name": "PolyMATH", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/polymath", + "https://api.llm-stats.com/leaderboard/benchmarks/polymath" + ], + "additional_details": { + "raw_benchmark_id": "polymath", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Polymath is a challenging multi-modal mathematical reasoning benchmark designed to evaluate the general cognitive reasoning abilities of Multi-modal Large Language Models (MLLMs). The benchmark comprises 5,000 manually collected high-quality images of cognitive textual and visual challenges across 10 distinct categories, including pattern recognition, spatial reasoning, and relative reasoning.", + "metric_id": "llm_stats.polymath.score", + "metric_name": "PolyMATH score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "polymath", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PolyMATH", + "raw_categories": "[\"math\",\"multimodal\",\"reasoning\",\"spatial_reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.774, + "details": { + "raw_score": "0.774", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "polymath", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/polymath\",\"https://api.llm-stats.com/leaderboard/benchmarks/polymath\"]", + "raw_score_id": "polymath::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "realworldqa::realworldqa-qwen3.6-plus", + "evaluation_name": "llm_stats.realworldqa", + "source_data": { + "dataset_name": "RealWorldQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/realworldqa", + "https://api.llm-stats.com/leaderboard/benchmarks/realworldqa" + ], + "additional_details": { + "raw_benchmark_id": "realworldqa", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "RealWorldQA is a benchmark designed to evaluate basic real-world spatial understanding capabilities of multimodal models. The initial release consists of over 700 anonymized images taken from vehicles and other real-world scenarios, each accompanied by a question and easily verifiable answer. Released by xAI as part of their Grok-1.5 Vision preview to test models' ability to understand natural scenes and spatial relationships in everyday visual contexts.", + "metric_id": "llm_stats.realworldqa.score", + "metric_name": "RealWorldQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "realworldqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "RealWorldQA", + "raw_categories": "[\"spatial_reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "25" + } + }, + "score_details": { + "score": 0.854, + "details": { + "raw_score": "0.854", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "realworldqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/realworldqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/realworldqa\"]", + "raw_score_id": "realworldqa::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "refcoco-avg::refcoco-avg-qwen3.6-plus", + "evaluation_name": "llm_stats.refcoco-avg", + "source_data": { + "dataset_name": "RefCOCO-avg", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/refcoco-avg", + "https://api.llm-stats.com/leaderboard/benchmarks/refcoco-avg" + ], + "additional_details": { + "raw_benchmark_id": "refcoco-avg", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "RefCOCO-avg measures object grounding accuracy averaged across RefCOCO, RefCOCO+, and RefCOCOg benchmarks.", + "metric_id": "llm_stats.refcoco-avg.score", + "metric_name": "RefCOCO-avg score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "refcoco-avg", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "RefCOCO-avg", + "raw_categories": "[\"spatial_reasoning\",\"grounding\",\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.935, + "details": { + "raw_score": "0.935", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "refcoco-avg", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/refcoco-avg\",\"https://api.llm-stats.com/leaderboard/benchmarks/refcoco-avg\"]", + "raw_score_id": "refcoco-avg::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "screenspot-pro::screenspot-pro-qwen3.6-plus", + "evaluation_name": "llm_stats.screenspot-pro", + "source_data": { + "dataset_name": "ScreenSpot Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/screenspot-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/screenspot-pro" + ], + "additional_details": { + "raw_benchmark_id": "screenspot-pro", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ScreenSpot-Pro is a novel GUI grounding benchmark designed to rigorously evaluate the grounding capabilities of multimodal large language models (MLLMs) in professional high-resolution computing environments. The benchmark comprises 1,581 instructions across 23 applications spanning 5 industries and 3 operating systems, featuring authentic high-resolution images from professional domains with expert annotations. Unlike previous benchmarks that focus on cropped screenshots in consumer applications, ScreenSpot-Pro addresses the complexity and diversity of real-world professional software scenarios, revealing significant performance gaps in current MLLM GUI perception capabilities.", + "metric_id": "llm_stats.screenspot-pro.score", + "metric_name": "ScreenSpot Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "screenspot-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ScreenSpot Pro", + "raw_categories": "[\"multimodal\",\"spatial_reasoning\",\"grounding\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.682, + "details": { + "raw_score": "0.682", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "screenspot-pro", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/screenspot-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/screenspot-pro\"]", + "raw_score_id": "screenspot-pro::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "simplevqa::simplevqa-qwen3.6-plus", + "evaluation_name": "llm_stats.simplevqa", + "source_data": { + "dataset_name": "SimpleVQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/simplevqa", + "https://api.llm-stats.com/leaderboard/benchmarks/simplevqa" + ], + "additional_details": { + "raw_benchmark_id": "simplevqa", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SimpleVQA is a visual question answering benchmark focused on simple queries.", + "metric_id": "llm_stats.simplevqa.score", + "metric_name": "SimpleVQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "simplevqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SimpleVQA", + "raw_categories": "[\"image_to_text\",\"multimodal\",\"general\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.673, + "details": { + "raw_score": "0.673", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "simplevqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/simplevqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/simplevqa\"]", + "raw_score_id": "simplevqa::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "skillsbench::skillsbench-qwen3.6-plus", + "evaluation_name": "llm_stats.skillsbench", + "source_data": { + "dataset_name": "SkillsBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/skillsbench", + "https://api.llm-stats.com/leaderboard/benchmarks/skillsbench" + ], + "additional_details": { + "raw_benchmark_id": "skillsbench", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SkillsBench evaluates coding agents on self-contained programming tasks, measuring practical engineering skills across diverse software development scenarios.", + "metric_id": "llm_stats.skillsbench.score", + "metric_name": "SkillsBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "skillsbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SkillsBench", + "raw_categories": "[\"agents\",\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.457, + "details": { + "raw_score": "0.457", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "skillsbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/skillsbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/skillsbench\"]", + "raw_score_id": "skillsbench::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "supergpqa::supergpqa-qwen3.6-plus", + "evaluation_name": "llm_stats.supergpqa", + "source_data": { + "dataset_name": "SuperGPQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/supergpqa", + "https://api.llm-stats.com/leaderboard/benchmarks/supergpqa" + ], + "additional_details": { + "raw_benchmark_id": "supergpqa", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SuperGPQA is a comprehensive benchmark that evaluates large language models across 285 graduate-level academic disciplines. The benchmark contains 25,957 questions covering 13 broad disciplinary areas including Engineering, Medicine, Science, and Law, with specialized fields in light industry, agriculture, and service-oriented domains. It employs a Human-LLM collaborative filtering mechanism with over 80 expert annotators to create challenging questions that assess graduate-level knowledge and reasoning capabilities.", + "metric_id": "llm_stats.supergpqa.score", + "metric_name": "SuperGPQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "supergpqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SuperGPQA", + "raw_categories": "[\"legal\",\"math\",\"physics\",\"reasoning\",\"finance\",\"general\",\"healthcare\",\"chemistry\",\"economics\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "34" + } + }, + "score_details": { + "score": 0.716, + "details": { + "raw_score": "0.716", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "supergpqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/supergpqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/supergpqa\"]", + "raw_score_id": "supergpqa::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-multilingual::swe-bench-multilingual-qwen3.6-plus", + "evaluation_name": "llm_stats.swe-bench-multilingual", + "source_data": { + "dataset_name": "SWE-bench Multilingual", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/swe-bench-multilingual", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-multilingual" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-multilingual", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multilingual benchmark for issue resolving in software engineering that covers Java, TypeScript, JavaScript, Go, Rust, C, and C++. Contains 1,632 high-quality instances carefully annotated from 2,456 candidates by 68 expert annotators, designed to evaluate Large Language Models across diverse software ecosystems beyond Python.", + "metric_id": "llm_stats.swe-bench-multilingual.score", + "metric_name": "SWE-bench Multilingual score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-multilingual", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-bench Multilingual", + "raw_categories": "[\"reasoning\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "33" + } + }, + "score_details": { + "score": 0.738, + "details": { + "raw_score": "0.738", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "swe-bench-multilingual", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/swe-bench-multilingual\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-multilingual\"]", + "raw_score_id": "swe-bench-multilingual::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-pro::swe-bench-pro-qwen3.6-plus", + "evaluation_name": "llm_stats.swe-bench-pro", + "source_data": { + "dataset_name": "SWE-Bench Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/swe-bench-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-pro" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-pro", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SWE-Bench Pro is an advanced version of SWE-Bench that evaluates language models on complex, real-world software engineering tasks requiring extended reasoning and multi-step problem solving.", + "metric_id": "llm_stats.swe-bench-pro.score", + "metric_name": "SWE-Bench Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Bench Pro", + "raw_categories": "[\"reasoning\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "35" + } + }, + "score_details": { + "score": 0.566, + "details": { + "raw_score": "0.566", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "swe-bench-pro", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/swe-bench-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-pro\"]", + "raw_score_id": "swe-bench-pro::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-verified::swe-bench-verified-qwen3.6-plus", + "evaluation_name": "llm_stats.swe-bench-verified", + "source_data": { + "dataset_name": "SWE-Bench Verified", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/swe-bench-verified", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-verified" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-verified", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A verified subset of 500 software engineering problems from real GitHub issues, validated by human annotators for evaluating language models' ability to resolve real-world coding issues by generating patches for Python codebases.", + "metric_id": "llm_stats.swe-bench-verified.score", + "metric_name": "SWE-Bench Verified score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-verified", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Bench Verified", + "raw_categories": "[\"reasoning\",\"frontend_development\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "103" + } + }, + "score_details": { + "score": 0.788, + "details": { + "raw_score": "0.788", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "swe-bench-verified", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/swe-bench-verified\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-verified\"]", + "raw_score_id": "swe-bench-verified::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "tau3-bench::tau3-bench-qwen3.6-plus", + "evaluation_name": "llm_stats.tau3-bench", + "source_data": { + "dataset_name": "TAU3-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/tau3-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/tau3-bench" + ], + "additional_details": { + "raw_benchmark_id": "tau3-bench", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "TAU3-Bench is a benchmark for evaluating general-purpose agent capabilities, testing models on multi-turn interactions with simulated user models, retrieval, and complex decision-making scenarios.", + "metric_id": "llm_stats.tau3-bench.score", + "metric_name": "TAU3-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "tau3-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TAU3-Bench", + "raw_categories": "[\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.707, + "details": { + "raw_score": "0.707", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "tau3-bench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/tau3-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/tau3-bench\"]", + "raw_score_id": "tau3-bench::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2::terminal-bench-2-qwen3.6-plus", + "evaluation_name": "llm_stats.terminal-bench-2", + "source_data": { + "dataset_name": "Terminal-Bench 2.0", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/terminal-bench-2", + "https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2" + ], + "additional_details": { + "raw_benchmark_id": "terminal-bench-2", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Terminal-Bench 2.0 is an updated benchmark for testing AI agents' tool use ability to operate a computer via terminal. It evaluates how well models can handle real-world, end-to-end tasks autonomously, including compiling code, training models, setting up servers, system administration, security tasks, data science workflows, and cybersecurity vulnerabilities.", + "metric_id": "llm_stats.terminal-bench-2.score", + "metric_name": "Terminal-Bench 2.0 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "terminal-bench-2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Terminal-Bench 2.0", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "49" + } + }, + "score_details": { + "score": 0.616, + "details": { + "raw_score": "0.616", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "terminal-bench-2", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/terminal-bench-2\",\"https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2\"]", + "raw_score_id": "terminal-bench-2::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "tir-bench::tir-bench-qwen3.6-plus", + "evaluation_name": "llm_stats.tir-bench", + "source_data": { + "dataset_name": "TIR-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/tir-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/tir-bench" + ], + "additional_details": { + "raw_benchmark_id": "tir-bench", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A tool-calling and multimodal interaction benchmark for testing visual instruction following and execution reliability.", + "metric_id": "llm_stats.tir-bench.score", + "metric_name": "TIR-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "tir-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TIR-Bench", + "raw_categories": "[\"multimodal\",\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.616, + "details": { + "raw_score": "0.616", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "tir-bench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/tir-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/tir-bench\"]", + "raw_score_id": "tir-bench::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "toolathlon::toolathlon-qwen3.6-plus", + "evaluation_name": "llm_stats.toolathlon", + "source_data": { + "dataset_name": "Toolathlon", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/toolathlon", + "https://api.llm-stats.com/leaderboard/benchmarks/toolathlon" + ], + "additional_details": { + "raw_benchmark_id": "toolathlon", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Tool Decathlon is a comprehensive benchmark for evaluating AI agents' ability to use multiple tools across diverse task categories. It measures proficiency in tool selection, sequencing, and execution across ten different tool-use scenarios.", + "metric_id": "llm_stats.toolathlon.score", + "metric_name": "Toolathlon score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "toolathlon", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Toolathlon", + "raw_categories": "[\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "24" + } + }, + "score_details": { + "score": 0.398, + "details": { + "raw_score": "0.398", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "toolathlon", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/toolathlon\",\"https://api.llm-stats.com/leaderboard/benchmarks/toolathlon\"]", + "raw_score_id": "toolathlon::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "v-star::v-star-qwen3.6-plus", + "evaluation_name": "llm_stats.v-star", + "source_data": { + "dataset_name": "V*", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/v-star", + "https://api.llm-stats.com/leaderboard/benchmarks/v-star" + ], + "additional_details": { + "raw_benchmark_id": "v-star", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A visual reasoning benchmark evaluating multimodal inference under challenging spatial and grounded tasks.", + "metric_id": "llm_stats.v-star.score", + "metric_name": "V* score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "v-star", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "V*", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.969, + "details": { + "raw_score": "0.969", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "v-star", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/v-star\",\"https://api.llm-stats.com/leaderboard/benchmarks/v-star\"]", + "raw_score_id": "v-star::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "video-mme::video-mme-qwen3.6-plus", + "evaluation_name": "llm_stats.video-mme", + "source_data": { + "dataset_name": "Video-MME", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/video-mme", + "https://api.llm-stats.com/leaderboard/benchmarks/video-mme" + ], + "additional_details": { + "raw_benchmark_id": "video-mme", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Video-MME is the first-ever comprehensive evaluation benchmark of Multi-modal Large Language Models (MLLMs) in video analysis. It features 900 videos totaling 254 hours with 2,700 human-annotated question-answer pairs across 6 primary visual domains (Knowledge, Film & Television, Sports Competition, Life Record, Multilingual, and others) and 30 subfields. The benchmark evaluates models across diverse temporal dimensions (11 seconds to 1 hour), integrates multi-modal inputs including video frames, subtitles, and audio, and uses rigorous manual labeling by expert annotators for precise assessment.", + "metric_id": "llm_stats.video-mme.score", + "metric_name": "Video-MME score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "video-mme", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Video-MME", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "17" + } + }, + "score_details": { + "score": 0.842, + "details": { + "raw_score": "0.842", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "video-mme", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/video-mme\",\"https://api.llm-stats.com/leaderboard/benchmarks/video-mme\"]", + "raw_score_id": "video-mme::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "videommmu::videommmu-qwen3.6-plus", + "evaluation_name": "llm_stats.videommmu", + "source_data": { + "dataset_name": "VideoMMMU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/videommmu", + "https://api.llm-stats.com/leaderboard/benchmarks/videommmu" + ], + "additional_details": { + "raw_benchmark_id": "videommmu", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Video-MMMU evaluates Large Multimodal Models' ability to acquire knowledge from expert-level professional videos across six disciplines through three cognitive stages: perception, comprehension, and adaptation. Contains 300 videos and 900 human-annotated questions spanning Art, Business, Science, Medicine, Humanities, and Engineering.", + "metric_id": "llm_stats.videommmu.score", + "metric_name": "VideoMMMU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "videommmu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VideoMMMU", + "raw_categories": "[\"multimodal\",\"reasoning\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "26" + } + }, + "score_details": { + "score": 0.84, + "details": { + "raw_score": "0.84", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "videommmu", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/videommmu\",\"https://api.llm-stats.com/leaderboard/benchmarks/videommmu\"]", + "raw_score_id": "videommmu::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "vita-bench::vita-bench-qwen3.6-plus", + "evaluation_name": "llm_stats.vita-bench", + "source_data": { + "dataset_name": "VITA-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/vita-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/vita-bench" + ], + "additional_details": { + "raw_benchmark_id": "vita-bench", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VITA-Bench evaluates AI agents on real-world virtual task automation, measuring their ability to complete complex multi-step tasks in simulated environments.", + "metric_id": "llm_stats.vita-bench.score", + "metric_name": "VITA-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "vita-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VITA-Bench", + "raw_categories": "[\"reasoning\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.443, + "details": { + "raw_score": "0.443", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "vita-bench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/vita-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/vita-bench\"]", + "raw_score_id": "vita-bench::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "we-math::we-math-qwen3.6-plus", + "evaluation_name": "llm_stats.we-math", + "source_data": { + "dataset_name": "We-Math", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/we-math", + "https://api.llm-stats.com/leaderboard/benchmarks/we-math" + ], + "additional_details": { + "raw_benchmark_id": "we-math", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "We-Math evaluates multimodal models on visual mathematical reasoning, requiring models to understand and solve math problems presented with visual elements such as diagrams, charts, and geometric figures.", + "metric_id": "llm_stats.we-math.score", + "metric_name": "We-Math score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "we-math", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "We-Math", + "raw_categories": "[\"math\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.89, + "details": { + "raw_score": "0.89", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "we-math", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/we-math\",\"https://api.llm-stats.com/leaderboard/benchmarks/we-math\"]", + "raw_score_id": "we-math::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "widesearch::widesearch-qwen3.6-plus", + "evaluation_name": "llm_stats.widesearch", + "source_data": { + "dataset_name": "WideSearch", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/widesearch", + "https://api.llm-stats.com/leaderboard/benchmarks/widesearch" + ], + "additional_details": { + "raw_benchmark_id": "widesearch", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WideSearch is an agentic search benchmark that evaluates models' ability to perform broad, parallel search operations across multiple sources. It tests wide-coverage information retrieval and synthesis capabilities.", + "metric_id": "llm_stats.widesearch.score", + "metric_name": "WideSearch score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "widesearch", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WideSearch", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.743, + "details": { + "raw_score": "0.743", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "widesearch", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/widesearch\",\"https://api.llm-stats.com/leaderboard/benchmarks/widesearch\"]", + "raw_score_id": "widesearch::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "wmt24::wmt24-qwen3.6-plus", + "evaluation_name": "llm_stats.wmt24", + "source_data": { + "dataset_name": "WMT24++", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.6-plus", + "https://llm-stats.com/benchmarks/wmt24", + "https://api.llm-stats.com/leaderboard/benchmarks/wmt24++" + ], + "additional_details": { + "raw_benchmark_id": "wmt24++", + "raw_model_id": "qwen3.6-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WMT24++ is a comprehensive multilingual machine translation benchmark that expands the WMT24 dataset to cover 55 languages and dialects. It includes human-written references and post-edits across four domains (literary, news, social, and speech) to evaluate machine translation systems and large language models across diverse linguistic contexts.", + "metric_id": "llm_stats.wmt24.score", + "metric_name": "WMT24++ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "wmt24++", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WMT24++", + "raw_categories": "[\"language\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.843, + "details": { + "raw_score": "0.843", + "raw_score_field": "score", + "raw_model_id": "qwen3.6-plus", + "raw_benchmark_id": "wmt24++", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.6-plus\",\"https://llm-stats.com/benchmarks/wmt24\",\"https://api.llm-stats.com/leaderboard/benchmarks/wmt24++\"]", + "raw_score_id": "wmt24++::qwen3.6-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/qwen/qwen3.7-plus/78b20080-2148-4734-bdd8-d14ece82db79.json b/data/llm_stats/qwen/qwen3.7-plus/78b20080-2148-4734-bdd8-d14ece82db79.json new file mode 100644 index 0000000000000000000000000000000000000000..fc2ac8df0ad3bde0f2243379cfc15c6039e6631f --- /dev/null +++ b/data/llm_stats/qwen/qwen3.7-plus/78b20080-2148-4734-bdd8-d14ece82db79.json @@ -0,0 +1,3480 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/qwen_qwen3.7-plus/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.7-plus", + "id": "qwen/qwen3.7-plus", + "developer": "qwen", + "additional_details": { + "raw_id": "qwen3.7-plus", + "raw_model_name": "qwen3.7-plus", + "raw_display_name": "Qwen3.7-Plus", + "raw_organization_id": "qwen", + "raw_organization_name": "Alibaba Cloud / Qwen Team", + "raw_context_length": "1000000", + "raw_input_modalities": "[\"image\",\"text\"]", + "raw_output_modalities": "[\"text\"]", + "raw_input_price": "0.32", + "raw_output_price": "1.28", + "raw_model_id": "qwen3.7-plus", + "raw_provider_slug": "qwen", + "raw_provider_name": "ZeroEval" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "androidworld::androidworld-qwen3.7-plus", + "evaluation_name": "llm_stats.androidworld", + "source_data": { + "dataset_name": "AndroidWorld", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/androidworld", + "https://api.llm-stats.com/leaderboard/benchmarks/androidworld" + ], + "additional_details": { + "raw_benchmark_id": "androidworld", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "AndroidWorld evaluates an agent's ability to operate in real Android GUI environments, completing multi-step tasks by perceiving screen content and executing touch/type actions.", + "metric_id": "llm_stats.androidworld.score", + "metric_name": "AndroidWorld score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "androidworld", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AndroidWorld", + "raw_categories": "[\"agents\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.81, + "details": { + "raw_score": "0.81", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "androidworld", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/androidworld\",\"https://api.llm-stats.com/leaderboard/benchmarks/androidworld\"]", + "raw_score_id": "androidworld::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "apex::apex-qwen3.7-plus", + "evaluation_name": "llm_stats.apex", + "source_data": { + "dataset_name": "Apex", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/apex", + "https://api.llm-stats.com/leaderboard/benchmarks/apex" + ], + "additional_details": { + "raw_benchmark_id": "apex", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Apex is a challenging frontier reasoning benchmark testing advanced multi-step problem solving across difficult STEM and logical tasks.", + "metric_id": "llm_stats.apex.score", + "metric_name": "Apex score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "apex", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Apex", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.227, + "details": { + "raw_score": "0.227", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "apex", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/apex\",\"https://api.llm-stats.com/leaderboard/benchmarks/apex\"]", + "raw_score_id": "apex::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "babyvision::babyvision-qwen3.7-plus", + "evaluation_name": "llm_stats.babyvision", + "source_data": { + "dataset_name": "BabyVision", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/babyvision", + "https://api.llm-stats.com/leaderboard/benchmarks/babyvision" + ], + "additional_details": { + "raw_benchmark_id": "babyvision", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A benchmark for early-stage visual reasoning and perception on child-like vision tasks.", + "metric_id": "llm_stats.babyvision.score", + "metric_name": "BabyVision score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "babyvision", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BabyVision", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.704, + "details": { + "raw_score": "0.704", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "babyvision", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/babyvision\",\"https://api.llm-stats.com/leaderboard/benchmarks/babyvision\"]", + "raw_score_id": "babyvision::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bc-vl::bc-vl-qwen3.7-plus", + "evaluation_name": "llm_stats.bc-vl", + "source_data": { + "dataset_name": "BC-VL", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/bc-vl", + "https://api.llm-stats.com/leaderboard/benchmarks/bc-vl" + ], + "additional_details": { + "raw_benchmark_id": "bc-vl", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BC-VL is a vision-language benchmark for knowledge-grounded multimodal question answering.", + "metric_id": "llm_stats.bc-vl.score", + "metric_name": "BC-VL score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bc-vl", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BC-VL", + "raw_categories": "[\"knowledge\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.511, + "details": { + "raw_score": "0.511", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "bc-vl", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/bc-vl\",\"https://api.llm-stats.com/leaderboard/benchmarks/bc-vl\"]", + "raw_score_id": "bc-vl::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bfcl-v4::bfcl-v4-qwen3.7-plus", + "evaluation_name": "llm_stats.bfcl-v4", + "source_data": { + "dataset_name": "BFCL-V4", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/bfcl-v4", + "https://api.llm-stats.com/leaderboard/benchmarks/bfcl-v4" + ], + "additional_details": { + "raw_benchmark_id": "bfcl-v4", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Berkeley Function Calling Leaderboard V4 (BFCL-V4) evaluates LLMs on their ability to accurately call functions and APIs, including simple, multiple, parallel, and nested function calls across diverse programming scenarios.", + "metric_id": "llm_stats.bfcl-v4.score", + "metric_name": "BFCL-V4 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bfcl-v4", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BFCL-V4", + "raw_categories": "[\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.729, + "details": { + "raw_score": "0.729", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "bfcl-v4", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/bfcl-v4\",\"https://api.llm-stats.com/leaderboard/benchmarks/bfcl-v4\"]", + "raw_score_id": "bfcl-v4::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "charxiv-r::charxiv-r-qwen3.7-plus", + "evaluation_name": "llm_stats.charxiv-r", + "source_data": { + "dataset_name": "CharXiv-R", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/charxiv-r", + "https://api.llm-stats.com/leaderboard/benchmarks/charxiv-r" + ], + "additional_details": { + "raw_benchmark_id": "charxiv-r", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CharXiv-R is the reasoning component of the CharXiv benchmark, focusing on complex reasoning questions that require synthesizing information across visual chart elements. It evaluates multimodal large language models on their ability to understand and reason about scientific charts from arXiv papers through various reasoning tasks.", + "metric_id": "llm_stats.charxiv-r.score", + "metric_name": "CharXiv-R score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "charxiv-r", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CharXiv-R", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "43" + } + }, + "score_details": { + "score": 0.859, + "details": { + "raw_score": "0.859", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "charxiv-r", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/charxiv-r\",\"https://api.llm-stats.com/leaderboard/benchmarks/charxiv-r\"]", + "raw_score_id": "charxiv-r::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "claw-eval-mm::claw-eval-mm-qwen3.7-plus", + "evaluation_name": "llm_stats.claw-eval-mm", + "source_data": { + "dataset_name": "ClawEval-MM", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/claw-eval-mm", + "https://api.llm-stats.com/leaderboard/benchmarks/claw-eval-mm" + ], + "additional_details": { + "raw_benchmark_id": "claw-eval-mm", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ClawEval-MM is the multimodal variant of ClawEval, evaluating agentic problem solving with visual inputs.", + "metric_id": "llm_stats.claw-eval-mm.score", + "metric_name": "ClawEval-MM score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "claw-eval-mm", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ClawEval-MM", + "raw_categories": "[\"multimodal\",\"agents\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.557, + "details": { + "raw_score": "0.557", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "claw-eval-mm", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/claw-eval-mm\",\"https://api.llm-stats.com/leaderboard/benchmarks/claw-eval-mm\"]", + "raw_score_id": "claw-eval-mm::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "claw-eval::claw-eval-qwen3.7-plus", + "evaluation_name": "llm_stats.claw-eval", + "source_data": { + "dataset_name": "Claw-Eval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/claw-eval", + "https://api.llm-stats.com/leaderboard/benchmarks/claw-eval" + ], + "additional_details": { + "raw_benchmark_id": "claw-eval", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Claw-Eval tests real-world agentic task completion across complex multi-step scenarios, evaluating a model's ability to use tools, navigate environments, and complete end-to-end tasks autonomously.", + "metric_id": "llm_stats.claw-eval.score", + "metric_name": "Claw-Eval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "claw-eval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Claw-Eval", + "raw_categories": "[\"agents\",\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.627, + "details": { + "raw_score": "0.627", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "claw-eval", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/claw-eval\",\"https://api.llm-stats.com/leaderboard/benchmarks/claw-eval\"]", + "raw_score_id": "claw-eval::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "countqa::countqa-qwen3.7-plus", + "evaluation_name": "llm_stats.countqa", + "source_data": { + "dataset_name": "CountQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/countqa", + "https://api.llm-stats.com/leaderboard/benchmarks/countqa" + ], + "additional_details": { + "raw_benchmark_id": "countqa", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CountQA is a benchmark for visual object counting and quantity reasoning over images.", + "metric_id": "llm_stats.countqa.score", + "metric_name": "CountQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "countqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CountQA", + "raw_categories": "[\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.77, + "details": { + "raw_score": "0.77", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "countqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/countqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/countqa\"]", + "raw_score_id": "countqa::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "coworkbench::coworkbench-qwen3.7-plus", + "evaluation_name": "llm_stats.coworkbench", + "source_data": { + "dataset_name": "CoWorkBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/coworkbench", + "https://api.llm-stats.com/leaderboard/benchmarks/coworkbench" + ], + "additional_details": { + "raw_benchmark_id": "coworkbench", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CoWorkBench is Qwen's internal cowork benchmark for evaluating long-horizon office and productivity agent tasks across domains such as computer science, finance, law, and medicine.", + "metric_id": "llm_stats.coworkbench.score", + "metric_name": "CoWorkBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "coworkbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CoWorkBench", + "raw_categories": "[\"productivity\",\"reasoning\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.651, + "details": { + "raw_score": "0.651", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "coworkbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/coworkbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/coworkbench\"]", + "raw_score_id": "coworkbench::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "critpt::critpt-qwen3.7-plus", + "evaluation_name": "llm_stats.critpt", + "source_data": { + "dataset_name": "CritPT", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/critpt", + "https://api.llm-stats.com/leaderboard/benchmarks/critpt" + ], + "additional_details": { + "raw_benchmark_id": "critpt", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CritPT is a challenging reasoning benchmark reported by Qwen for evaluating frontier mathematical and critical problem-solving capability.", + "metric_id": "llm_stats.critpt.score", + "metric_name": "CritPT score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "critpt", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CritPT", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.06, + "details": { + "raw_score": "0.06", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "critpt", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/critpt\",\"https://api.llm-stats.com/leaderboard/benchmarks/critpt\"]", + "raw_score_id": "critpt::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "deep-planning::deep-planning-qwen3.7-plus", + "evaluation_name": "llm_stats.deep-planning", + "source_data": { + "dataset_name": "DeepPlanning", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/deep-planning", + "https://api.llm-stats.com/leaderboard/benchmarks/deep-planning" + ], + "additional_details": { + "raw_benchmark_id": "deep-planning", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "DeepPlanning evaluates LLMs on complex multi-step planning tasks requiring long-horizon reasoning, goal decomposition, and strategic decision-making.", + "metric_id": "llm_stats.deep-planning.score", + "metric_name": "DeepPlanning score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "deep-planning", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DeepPlanning", + "raw_categories": "[\"reasoning\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.623, + "details": { + "raw_score": "0.623", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "deep-planning", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/deep-planning\",\"https://api.llm-stats.com/leaderboard/benchmarks/deep-planning\"]", + "raw_score_id": "deep-planning::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "erqa::erqa-qwen3.7-plus", + "evaluation_name": "llm_stats.erqa", + "source_data": { + "dataset_name": "ERQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/erqa", + "https://api.llm-stats.com/leaderboard/benchmarks/erqa" + ], + "additional_details": { + "raw_benchmark_id": "erqa", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Embodied Reasoning Question Answering benchmark consisting of 400 multiple-choice visual questions across spatial reasoning, trajectory reasoning, action reasoning, state estimation, and multi-view reasoning for evaluating AI capabilities in physical world interactions", + "metric_id": "llm_stats.erqa.score", + "metric_name": "ERQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "erqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ERQA", + "raw_categories": "[\"reasoning\",\"spatial_reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "22" + } + }, + "score_details": { + "score": 0.698, + "details": { + "raw_score": "0.698", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "erqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/erqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/erqa\"]", + "raw_score_id": "erqa::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "global-piqa::global-piqa-qwen3.7-plus", + "evaluation_name": "llm_stats.global-piqa", + "source_data": { + "dataset_name": "Global PIQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/global-piqa", + "https://api.llm-stats.com/leaderboard/benchmarks/global-piqa" + ], + "additional_details": { + "raw_benchmark_id": "global-piqa", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Global PIQA is a multilingual commonsense reasoning benchmark that evaluates physical interaction knowledge across 100 languages and cultures. It tests AI systems' understanding of physical world knowledge in diverse cultural contexts through multiple choice questions about everyday situations requiring physical commonsense.", + "metric_id": "llm_stats.global-piqa.score", + "metric_name": "Global PIQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "global-piqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Global PIQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.903, + "details": { + "raw_score": "0.903", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "global-piqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/global-piqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/global-piqa\"]", + "raw_score_id": "global-piqa::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gpqa::gpqa-qwen3.7-plus", + "evaluation_name": "llm_stats.gpqa", + "source_data": { + "dataset_name": "GPQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/gpqa", + "https://api.llm-stats.com/leaderboard/benchmarks/gpqa" + ], + "additional_details": { + "raw_benchmark_id": "gpqa", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A challenging dataset of 448 multiple-choice questions written by domain experts in biology, physics, and chemistry. Questions are Google-proof and extremely difficult, with PhD experts reaching 65% accuracy.", + "metric_id": "llm_stats.gpqa.score", + "metric_name": "GPQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gpqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GPQA", + "raw_categories": "[\"physics\",\"reasoning\",\"general\",\"biology\",\"chemistry\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "226" + } + }, + "score_details": { + "score": 0.903, + "details": { + "raw_score": "0.903", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "gpqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/gpqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/gpqa\"]", + "raw_score_id": "gpqa::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hipho::hipho-qwen3.7-plus", + "evaluation_name": "llm_stats.hipho", + "source_data": { + "dataset_name": "HiPhO", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/hipho", + "https://api.llm-stats.com/leaderboard/benchmarks/hipho" + ], + "additional_details": { + "raw_benchmark_id": "hipho", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "HiPhO is a high-school physics olympiad benchmark evaluating multimodal reasoning over physics problems that include diagrams and figures.", + "metric_id": "llm_stats.hipho.score", + "metric_name": "HiPhO score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hipho", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HiPhO", + "raw_categories": "[\"multimodal\",\"reasoning\",\"science\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.841, + "details": { + "raw_score": "0.841", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "hipho", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/hipho\",\"https://api.llm-stats.com/leaderboard/benchmarks/hipho\"]", + "raw_score_id": "hipho::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hmmt-feb-26::hmmt-feb-26-qwen3.7-plus", + "evaluation_name": "llm_stats.hmmt-feb-26", + "source_data": { + "dataset_name": "HMMT Feb 26", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/hmmt-feb-26", + "https://api.llm-stats.com/leaderboard/benchmarks/hmmt-feb-26" + ], + "additional_details": { + "raw_benchmark_id": "hmmt-feb-26", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "HMMT February 2026 is a math competition benchmark based on problems from the Harvard-MIT Mathematics Tournament, testing advanced mathematical problem-solving and reasoning.", + "metric_id": "llm_stats.hmmt-feb-26.score", + "metric_name": "HMMT Feb 26 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hmmt-feb-26", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HMMT Feb 26", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.929, + "details": { + "raw_score": "0.929", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "hmmt-feb-26", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/hmmt-feb-26\",\"https://api.llm-stats.com/leaderboard/benchmarks/hmmt-feb-26\"]", + "raw_score_id": "hmmt-feb-26::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ifbench::ifbench-qwen3.7-plus", + "evaluation_name": "llm_stats.ifbench", + "source_data": { + "dataset_name": "IFBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/ifbench", + "https://api.llm-stats.com/leaderboard/benchmarks/ifbench" + ], + "additional_details": { + "raw_benchmark_id": "ifbench", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Instruction Following Benchmark evaluating model's ability to follow complex instructions", + "metric_id": "llm_stats.ifbench.score", + "metric_name": "IFBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ifbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IFBench", + "raw_categories": "[\"instruction_following\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.791, + "details": { + "raw_score": "0.791", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "ifbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/ifbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/ifbench\"]", + "raw_score_id": "ifbench::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ifeval::ifeval-qwen3.7-plus", + "evaluation_name": "llm_stats.ifeval", + "source_data": { + "dataset_name": "IFEval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/ifeval", + "https://api.llm-stats.com/leaderboard/benchmarks/ifeval" + ], + "additional_details": { + "raw_benchmark_id": "ifeval", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Instruction-Following Evaluation (IFEval) benchmark for large language models, focusing on verifiable instructions with 25 types of instructions and around 500 prompts containing one or more verifiable constraints", + "metric_id": "llm_stats.ifeval.score", + "metric_name": "IFEval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ifeval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IFEval", + "raw_categories": "[\"instruction_following\",\"structured_output\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "65" + } + }, + "score_details": { + "score": 0.946, + "details": { + "raw_score": "0.946", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "ifeval", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/ifeval\",\"https://api.llm-stats.com/leaderboard/benchmarks/ifeval\"]", + "raw_score_id": "ifeval::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "imo-answerbench::imo-answerbench-qwen3.7-plus", + "evaluation_name": "llm_stats.imo-answerbench", + "source_data": { + "dataset_name": "IMO-AnswerBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/imo-answerbench", + "https://api.llm-stats.com/leaderboard/benchmarks/imo-answerbench" + ], + "additional_details": { + "raw_benchmark_id": "imo-answerbench", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "IMO-AnswerBench is a benchmark for evaluating mathematical reasoning capabilities on International Mathematical Olympiad (IMO) problems, focusing on answer generation and verification.", + "metric_id": "llm_stats.imo-answerbench.score", + "metric_name": "IMO-AnswerBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "imo-answerbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IMO-AnswerBench", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.86, + "details": { + "raw_score": "0.86", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "imo-answerbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/imo-answerbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/imo-answerbench\"]", + "raw_score_id": "imo-answerbench::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "include::include-qwen3.7-plus", + "evaluation_name": "llm_stats.include", + "source_data": { + "dataset_name": "Include", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/include", + "https://api.llm-stats.com/leaderboard/benchmarks/include" + ], + "additional_details": { + "raw_benchmark_id": "include", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Include benchmark - specific documentation not found in official sources", + "metric_id": "llm_stats.include.score", + "metric_name": "Include score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "include", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Include", + "raw_categories": "[\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "31" + } + }, + "score_details": { + "score": 0.83, + "details": { + "raw_score": "0.83", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "include", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/include\",\"https://api.llm-stats.com/leaderboard/benchmarks/include\"]", + "raw_score_id": "include::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "lingoqa::lingoqa-qwen3.7-plus", + "evaluation_name": "llm_stats.lingoqa", + "source_data": { + "dataset_name": "LingoQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/lingoqa", + "https://api.llm-stats.com/leaderboard/benchmarks/lingoqa" + ], + "additional_details": { + "raw_benchmark_id": "lingoqa", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A benchmark for multimodal spatial-language understanding and visual-linguistic question answering.", + "metric_id": "llm_stats.lingoqa.score", + "metric_name": "LingoQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "lingoqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LingoQA", + "raw_categories": "[\"language\",\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.834, + "details": { + "raw_score": "0.834", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "lingoqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/lingoqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/lingoqa\"]", + "raw_score_id": "lingoqa::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "livecodebench-v6::livecodebench-v6-qwen3.7-plus", + "evaluation_name": "llm_stats.livecodebench-v6", + "source_data": { + "dataset_name": "LiveCodeBench v6", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/livecodebench-v6", + "https://api.llm-stats.com/leaderboard/benchmarks/livecodebench-v6" + ], + "additional_details": { + "raw_benchmark_id": "livecodebench-v6", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LiveCodeBench is a holistic and contamination-free evaluation benchmark for large language models for code. It continuously collects new problems from programming contests (LeetCode, AtCoder, CodeForces) and evaluates four different scenarios: code generation, self-repair, code execution, and test output prediction. Problems are annotated with release dates to enable evaluation on unseen problems released after a model's training cutoff.", + "metric_id": "llm_stats.livecodebench-v6.score", + "metric_name": "LiveCodeBench v6 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "livecodebench-v6", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LiveCodeBench v6", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "53" + } + }, + "score_details": { + "score": 0.896, + "details": { + "raw_score": "0.896", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "livecodebench-v6", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/livecodebench-v6\",\"https://api.llm-stats.com/leaderboard/benchmarks/livecodebench-v6\"]", + "raw_score_id": "livecodebench-v6::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "lvbench::lvbench-qwen3.7-plus", + "evaluation_name": "llm_stats.lvbench", + "source_data": { + "dataset_name": "LVBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/lvbench", + "https://api.llm-stats.com/leaderboard/benchmarks/lvbench" + ], + "additional_details": { + "raw_benchmark_id": "lvbench", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "LVBench is an extreme long video understanding benchmark designed to evaluate multimodal models on videos up to two hours in duration. It contains 6 major categories and 21 subcategories, with videos averaging five times longer than existing datasets. The benchmark addresses applications requiring comprehension of extremely long videos.", + "metric_id": "llm_stats.lvbench.score", + "metric_name": "LVBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "lvbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "LVBench", + "raw_categories": "[\"long_context\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.762, + "details": { + "raw_score": "0.762", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "lvbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/lvbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/lvbench\"]", + "raw_score_id": "lvbench::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mathvision::mathvision-qwen3.7-plus", + "evaluation_name": "llm_stats.mathvision", + "source_data": { + "dataset_name": "MathVision", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/mathvision", + "https://api.llm-stats.com/leaderboard/benchmarks/mathvision" + ], + "additional_details": { + "raw_benchmark_id": "mathvision", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MATH-Vision is a dataset designed to measure multimodal mathematical reasoning capabilities. It focuses on evaluating how well models can solve mathematical problems that require both visual understanding and mathematical reasoning, bridging the gap between visual and mathematical domains.", + "metric_id": "llm_stats.mathvision.score", + "metric_name": "MathVision score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mathvision", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MathVision", + "raw_categories": "[\"math\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "31" + } + }, + "score_details": { + "score": 0.903, + "details": { + "raw_score": "0.903", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "mathvision", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/mathvision\",\"https://api.llm-stats.com/leaderboard/benchmarks/mathvision\"]", + "raw_score_id": "mathvision::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "maxife::maxife-qwen3.7-plus", + "evaluation_name": "llm_stats.maxife", + "source_data": { + "dataset_name": "MAXIFE", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/maxife", + "https://api.llm-stats.com/leaderboard/benchmarks/maxife" + ], + "additional_details": { + "raw_benchmark_id": "maxife", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MAXIFE is a multilingual benchmark evaluating LLMs on instruction following and execution across multiple languages and cultural contexts.", + "metric_id": "llm_stats.maxife.score", + "metric_name": "MAXIFE score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "maxife", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MAXIFE", + "raw_categories": "[\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.888, + "details": { + "raw_score": "0.888", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "maxife", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/maxife\",\"https://api.llm-stats.com/leaderboard/benchmarks/maxife\"]", + "raw_score_id": "maxife::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mcp-atlas::mcp-atlas-qwen3.7-plus", + "evaluation_name": "llm_stats.mcp-atlas", + "source_data": { + "dataset_name": "MCP Atlas", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/mcp-atlas", + "https://api.llm-stats.com/leaderboard/benchmarks/mcp-atlas" + ], + "additional_details": { + "raw_benchmark_id": "mcp-atlas", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MCP Atlas is a benchmark for evaluating AI models on scaled tool use capabilities, measuring how well models can coordinate and utilize multiple tools across complex multi-step tasks.", + "metric_id": "llm_stats.mcp-atlas.score", + "metric_name": "MCP Atlas score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mcp-atlas", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MCP Atlas", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.732, + "details": { + "raw_score": "0.732", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "mcp-atlas", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/mcp-atlas\",\"https://api.llm-stats.com/leaderboard/benchmarks/mcp-atlas\"]", + "raw_score_id": "mcp-atlas::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mcp-mark::mcp-mark-qwen3.7-plus", + "evaluation_name": "llm_stats.mcp-mark", + "source_data": { + "dataset_name": "MCP-Mark", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/mcp-mark", + "https://api.llm-stats.com/leaderboard/benchmarks/mcp-mark" + ], + "additional_details": { + "raw_benchmark_id": "mcp-mark", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MCP-Mark evaluates LLMs on their ability to use Model Context Protocol (MCP) tools effectively, testing tool discovery, selection, invocation, and result interpretation across diverse MCP server scenarios.", + "metric_id": "llm_stats.mcp-mark.score", + "metric_name": "MCP-Mark score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mcp-mark", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MCP-Mark", + "raw_categories": "[\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.587, + "details": { + "raw_score": "0.587", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "mcp-mark", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/mcp-mark\",\"https://api.llm-stats.com/leaderboard/benchmarks/mcp-mark\"]", + "raw_score_id": "mcp-mark::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "medxpertqa-mm::medxpertqa-mm-qwen3.7-plus", + "evaluation_name": "llm_stats.medxpertqa-mm", + "source_data": { + "dataset_name": "MedXpertQA-MM", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/medxpertqa-mm", + "https://api.llm-stats.com/leaderboard/benchmarks/medxpertqa-mm" + ], + "additional_details": { + "raw_benchmark_id": "medxpertqa-mm", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MedXpertQA-MM is the multimodal subset of MedXpertQA, evaluating expert-level medical question answering grounded in medical images.", + "metric_id": "llm_stats.medxpertqa-mm.score", + "metric_name": "MedXpertQA-MM score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "medxpertqa-mm", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MedXpertQA-MM", + "raw_categories": "[\"knowledge\",\"medical\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.71, + "details": { + "raw_score": "0.71", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "medxpertqa-mm", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/medxpertqa-mm\",\"https://api.llm-stats.com/leaderboard/benchmarks/medxpertqa-mm\"]", + "raw_score_id": "medxpertqa-mm::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mlvu::mlvu-qwen3.7-plus", + "evaluation_name": "llm_stats.mlvu", + "source_data": { + "dataset_name": "MLVU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/mlvu", + "https://api.llm-stats.com/leaderboard/benchmarks/mlvu" + ], + "additional_details": { + "raw_benchmark_id": "mlvu", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A comprehensive benchmark for multi-task long video understanding that evaluates multimodal large language models on videos ranging from 3 minutes to 2 hours across 9 distinct tasks including reasoning, captioning, recognition, and summarization.", + "metric_id": "llm_stats.mlvu.score", + "metric_name": "MLVU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mlvu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MLVU", + "raw_categories": "[\"long_context\",\"multimodal\",\"video\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.874, + "details": { + "raw_score": "0.874", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "mlvu", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/mlvu\",\"https://api.llm-stats.com/leaderboard/benchmarks/mlvu\"]", + "raw_score_id": "mlvu::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmbc::mmbc-qwen3.7-plus", + "evaluation_name": "llm_stats.mmbc", + "source_data": { + "dataset_name": "MMBC", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/mmbc", + "https://api.llm-stats.com/leaderboard/benchmarks/mmbc" + ], + "additional_details": { + "raw_benchmark_id": "mmbc", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MMBC is a multimodal benchmark for vision-language knowledge and reasoning.", + "metric_id": "llm_stats.mmbc.score", + "metric_name": "MMBC score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmbc", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMBC", + "raw_categories": "[\"knowledge\",\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.463, + "details": { + "raw_score": "0.463", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "mmbc", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/mmbc\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmbc\"]", + "raw_score_id": "mmbc::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu-pro::mmlu-pro-qwen3.7-plus", + "evaluation_name": "llm_stats.mmlu-pro", + "source_data": { + "dataset_name": "MMLU-Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/mmlu-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu-pro" + ], + "additional_details": { + "raw_benchmark_id": "mmlu-pro", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A more robust and challenging multi-task language understanding benchmark that extends MMLU by expanding multiple-choice options from 4 to 10, eliminating trivial questions, and focusing on reasoning-intensive tasks. Features over 12,000 curated questions across 14 domains and causes a 16-33% accuracy drop compared to original MMLU.", + "metric_id": "llm_stats.mmlu-pro.score", + "metric_name": "MMLU-Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU-Pro", + "raw_categories": "[\"language\",\"legal\",\"math\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "129" + } + }, + "score_details": { + "score": 0.885, + "details": { + "raw_score": "0.885", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "mmlu-pro", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/mmlu-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu-pro\"]", + "raw_score_id": "mmlu-pro::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu-prox::mmlu-prox-qwen3.7-plus", + "evaluation_name": "llm_stats.mmlu-prox", + "source_data": { + "dataset_name": "MMLU-ProX", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/mmlu-prox", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu-prox" + ], + "additional_details": { + "raw_benchmark_id": "mmlu-prox", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Extended version of MMLU-Pro providing additional challenging multiple-choice questions for evaluating language models across diverse academic and professional domains. Built on the foundation of the Massive Multitask Language Understanding benchmark framework.", + "metric_id": "llm_stats.mmlu-prox.score", + "metric_name": "MMLU-ProX score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu-prox", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU-ProX", + "raw_categories": "[\"language\",\"legal\",\"math\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "32" + } + }, + "score_details": { + "score": 0.854, + "details": { + "raw_score": "0.854", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "mmlu-prox", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/mmlu-prox\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu-prox\"]", + "raw_score_id": "mmlu-prox::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu-redux::mmlu-redux-qwen3.7-plus", + "evaluation_name": "llm_stats.mmlu-redux", + "source_data": { + "dataset_name": "MMLU-Redux", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/mmlu-redux", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu-redux" + ], + "additional_details": { + "raw_benchmark_id": "mmlu-redux", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "An improved version of the MMLU benchmark featuring manually re-annotated questions to identify and correct errors in the original dataset. Provides more reliable evaluation metrics for language models by addressing dataset quality issues found in the original MMLU.", + "metric_id": "llm_stats.mmlu-redux.score", + "metric_name": "MMLU-Redux score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu-redux", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU-Redux", + "raw_categories": "[\"language\",\"math\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "48" + } + }, + "score_details": { + "score": 0.945, + "details": { + "raw_score": "0.945", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "mmlu-redux", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/mmlu-redux\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu-redux\"]", + "raw_score_id": "mmlu-redux::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmmlu::mmmlu-qwen3.7-plus", + "evaluation_name": "llm_stats.mmmlu", + "source_data": { + "dataset_name": "MMMLU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/mmmlu", + "https://api.llm-stats.com/leaderboard/benchmarks/mmmlu" + ], + "additional_details": { + "raw_benchmark_id": "mmmlu", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Multilingual Massive Multitask Language Understanding dataset released by OpenAI, featuring professionally translated MMLU test questions across 14 languages including Arabic, Bengali, German, Spanish, French, Hindi, Indonesian, Italian, Japanese, Korean, Portuguese, Swahili, Yoruba, and Chinese. Contains approximately 15,908 multiple-choice questions per language covering 57 subjects.", + "metric_id": "llm_stats.mmmlu.score", + "metric_name": "MMMLU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmmlu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMMLU", + "raw_categories": "[\"language\",\"math\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "49" + } + }, + "score_details": { + "score": 0.89, + "details": { + "raw_score": "0.89", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "mmmlu", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/mmmlu\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmmlu\"]", + "raw_score_id": "mmmlu::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmmu-pro::mmmu-pro-qwen3.7-plus", + "evaluation_name": "llm_stats.mmmu-pro", + "source_data": { + "dataset_name": "MMMU-Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/mmmu-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/mmmu-pro" + ], + "additional_details": { + "raw_benchmark_id": "mmmu-pro", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A more robust multi-discipline multimodal understanding benchmark that enhances MMMU through a three-step process: filtering text-only answerable questions, augmenting candidate options, and introducing vision-only input settings. Achieves significantly lower model performance (16.8-26.9%) compared to original MMMU, providing more rigorous evaluation that closely mimics real-world scenarios.", + "metric_id": "llm_stats.mmmu-pro.score", + "metric_name": "MMMU-Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmmu-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMMU-Pro", + "raw_categories": "[\"multimodal\",\"reasoning\",\"general\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "60" + } + }, + "score_details": { + "score": 0.79, + "details": { + "raw_score": "0.79", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "mmmu-pro", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/mmmu-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmmu-pro\"]", + "raw_score_id": "mmmu-pro::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmsearch-plus::mmsearch-plus-qwen3.7-plus", + "evaluation_name": "llm_stats.mmsearch-plus", + "source_data": { + "dataset_name": "MMSearch-Plus", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/mmsearch-plus", + "https://api.llm-stats.com/leaderboard/benchmarks/mmsearch-plus" + ], + "additional_details": { + "raw_benchmark_id": "mmsearch-plus", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MMSearch-Plus is an extended variant of MMSearch with harder multimodal search and retrieval tasks requiring deeper reasoning over visual and textual search results.", + "metric_id": "llm_stats.mmsearch-plus.score", + "metric_name": "MMSearch-Plus score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmsearch-plus", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMSearch-Plus", + "raw_categories": "[\"multimodal\",\"search\",\"agents\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.414, + "details": { + "raw_score": "0.414", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "mmsearch-plus", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/mmsearch-plus\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmsearch-plus\"]", + "raw_score_id": "mmsearch-plus::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mrcr-v2::mrcr-v2-qwen3.7-plus", + "evaluation_name": "llm_stats.mrcr-v2", + "source_data": { + "dataset_name": "MRCR v2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/mrcr-v2", + "https://api.llm-stats.com/leaderboard/benchmarks/mrcr-v2" + ], + "additional_details": { + "raw_benchmark_id": "mrcr-v2", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MRCR v2 (Multi-Round Coreference Resolution version 2) is an enhanced version of the synthetic long-context reasoning task. It extends the original MRCR framework with improved evaluation criteria and additional complexity for testing models' ability to maintain attention and reasoning across extended contexts.", + "metric_id": "llm_stats.mrcr-v2.score", + "metric_name": "MRCR v2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mrcr-v2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MRCR v2", + "raw_categories": "[\"long_context\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "8" + } + }, + "score_details": { + "score": 0.917, + "details": { + "raw_score": "0.917", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "mrcr-v2", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/mrcr-v2\",\"https://api.llm-stats.com/leaderboard/benchmarks/mrcr-v2\"]", + "raw_score_id": "mrcr-v2::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "nl2repo::nl2repo-qwen3.7-plus", + "evaluation_name": "llm_stats.nl2repo", + "source_data": { + "dataset_name": "NL2Repo", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/nl2repo", + "https://api.llm-stats.com/leaderboard/benchmarks/nl2repo" + ], + "additional_details": { + "raw_benchmark_id": "nl2repo", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "NL2Repo evaluates long-horizon coding capabilities including repository-level understanding, where models must generate or modify code across entire repositories from natural language specifications.", + "metric_id": "llm_stats.nl2repo.score", + "metric_name": "NL2Repo score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "nl2repo", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "NL2Repo", + "raw_categories": "[\"agents\",\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.411, + "details": { + "raw_score": "0.411", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "nl2repo", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/nl2repo\",\"https://api.llm-stats.com/leaderboard/benchmarks/nl2repo\"]", + "raw_score_id": "nl2repo::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "nova-63::nova-63-qwen3.7-plus", + "evaluation_name": "llm_stats.nova-63", + "source_data": { + "dataset_name": "NOVA-63", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/nova-63", + "https://api.llm-stats.com/leaderboard/benchmarks/nova-63" + ], + "additional_details": { + "raw_benchmark_id": "nova-63", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "NOVA-63 is a multilingual evaluation benchmark covering 63 languages, designed to assess LLM performance across diverse linguistic contexts and tasks.", + "metric_id": "llm_stats.nova-63.score", + "metric_name": "NOVA-63 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "nova-63", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "NOVA-63", + "raw_categories": "[\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.588, + "details": { + "raw_score": "0.588", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "nova-63", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/nova-63\",\"https://api.llm-stats.com/leaderboard/benchmarks/nova-63\"]", + "raw_score_id": "nova-63::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "ocrbench-v2::ocrbench-v2-qwen3.7-plus", + "evaluation_name": "llm_stats.ocrbench-v2", + "source_data": { + "dataset_name": "OCRBench_V2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/ocrbench-v2", + "https://api.llm-stats.com/leaderboard/benchmarks/ocrbench-v2" + ], + "additional_details": { + "raw_benchmark_id": "ocrbench-v2", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OCRBench v2: Enhanced large-scale bilingual benchmark for evaluating Large Multimodal Models on visual text localization and reasoning with 10,000 human-verified question-answering pairs across 8 core OCR capabilities", + "metric_id": "llm_stats.ocrbench-v2.score", + "metric_name": "OCRBench_V2 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "ocrbench-v2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OCRBench_V2", + "raw_categories": "[\"image_to_text\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "7" + } + }, + "score_details": { + "score": 0.671, + "details": { + "raw_score": "0.671", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "ocrbench-v2", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/ocrbench-v2\",\"https://api.llm-stats.com/leaderboard/benchmarks/ocrbench-v2\"]", + "raw_score_id": "ocrbench-v2::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "odinw::odinw-qwen3.7-plus", + "evaluation_name": "llm_stats.odinw", + "source_data": { + "dataset_name": "ODinW", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/odinw", + "https://api.llm-stats.com/leaderboard/benchmarks/odinw" + ], + "additional_details": { + "raw_benchmark_id": "odinw", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Object Detection in the Wild (ODinW) benchmark for evaluating object detection models' task-level transfer ability across diverse real-world datasets in terms of prediction accuracy and adaptation efficiency", + "metric_id": "llm_stats.odinw.score", + "metric_name": "ODinW score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "odinw", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ODinW", + "raw_categories": "[\"vision\"]", + "raw_modality": "image", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.511, + "details": { + "raw_score": "0.511", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "odinw", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/odinw\",\"https://api.llm-stats.com/leaderboard/benchmarks/odinw\"]", + "raw_score_id": "odinw::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "omnidocbench-1.5::omnidocbench-1.5-qwen3.7-plus", + "evaluation_name": "llm_stats.omnidocbench-1.5", + "source_data": { + "dataset_name": "OmniDocBench 1.5", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/omnidocbench-1.5", + "https://api.llm-stats.com/leaderboard/benchmarks/omnidocbench-1.5" + ], + "additional_details": { + "raw_benchmark_id": "omnidocbench-1.5", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OmniDocBench 1.5 is a comprehensive benchmark for evaluating multimodal large language models on document understanding tasks, including OCR, document parsing, information extraction, and visual question answering across diverse document types. Lower Overall Edit Distance scores are better.", + "metric_id": "llm_stats.omnidocbench-1.5.score", + "metric_name": "OmniDocBench 1.5 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "omnidocbench-1.5", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OmniDocBench 1.5", + "raw_categories": "[\"multimodal\",\"reasoning\",\"structured_output\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "16" + } + }, + "score_details": { + "score": 0.914, + "details": { + "raw_score": "0.914", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "omnidocbench-1.5", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/omnidocbench-1.5\",\"https://api.llm-stats.com/leaderboard/benchmarks/omnidocbench-1.5\"]", + "raw_score_id": "omnidocbench-1.5::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "osworld-verified::osworld-verified-qwen3.7-plus", + "evaluation_name": "llm_stats.osworld-verified", + "source_data": { + "dataset_name": "OSWorld-Verified", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/osworld-verified", + "https://api.llm-stats.com/leaderboard/benchmarks/osworld-verified" + ], + "additional_details": { + "raw_benchmark_id": "osworld-verified", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "OSWorld-Verified is a verified subset of OSWorld, a scalable real computer environment for multimodal agents supporting task setup, execution-based evaluation, and interactive learning across Ubuntu, Windows, and macOS.", + "metric_id": "llm_stats.osworld-verified.score", + "metric_name": "OSWorld-Verified score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "osworld-verified", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "OSWorld-Verified", + "raw_categories": "[\"multimodal\",\"general\",\"agents\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.733, + "details": { + "raw_score": "0.733", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "osworld-verified", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/osworld-verified\",\"https://api.llm-stats.com/leaderboard/benchmarks/osworld-verified\"]", + "raw_score_id": "osworld-verified::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "polymath::polymath-qwen3.7-plus", + "evaluation_name": "llm_stats.polymath", + "source_data": { + "dataset_name": "PolyMATH", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/polymath", + "https://api.llm-stats.com/leaderboard/benchmarks/polymath" + ], + "additional_details": { + "raw_benchmark_id": "polymath", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Polymath is a challenging multi-modal mathematical reasoning benchmark designed to evaluate the general cognitive reasoning abilities of Multi-modal Large Language Models (MLLMs). The benchmark comprises 5,000 manually collected high-quality images of cognitive textual and visual challenges across 10 distinct categories, including pattern recognition, spatial reasoning, and relative reasoning.", + "metric_id": "llm_stats.polymath.score", + "metric_name": "PolyMATH score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "polymath", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "PolyMATH", + "raw_categories": "[\"math\",\"multimodal\",\"reasoning\",\"spatial_reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.84, + "details": { + "raw_score": "0.84", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "polymath", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/polymath\",\"https://api.llm-stats.com/leaderboard/benchmarks/polymath\"]", + "raw_score_id": "polymath::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "qwenclawbench::qwenclawbench-qwen3.7-plus", + "evaluation_name": "llm_stats.qwenclawbench", + "source_data": { + "dataset_name": "QwenClawBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/qwenclawbench", + "https://api.llm-stats.com/leaderboard/benchmarks/qwenclawbench" + ], + "additional_details": { + "raw_benchmark_id": "qwenclawbench", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "QwenClawBench is a real-user-distribution Claw agent benchmark for evaluating coding agents on realistic developer tasks.", + "metric_id": "llm_stats.qwenclawbench.score", + "metric_name": "QwenClawBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "qwenclawbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "QwenClawBench", + "raw_categories": "[\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.618, + "details": { + "raw_score": "0.618", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "qwenclawbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/qwenclawbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/qwenclawbench\"]", + "raw_score_id": "qwenclawbench::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "qwenworldbench::qwenworldbench-qwen3.7-plus", + "evaluation_name": "llm_stats.qwenworldbench", + "source_data": { + "dataset_name": "QwenWorldBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/qwenworldbench", + "https://api.llm-stats.com/leaderboard/benchmarks/qwenworldbench" + ], + "additional_details": { + "raw_benchmark_id": "qwenworldbench", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "QwenWorldBench is Qwen's internal benchmark for evaluating LLMs as world models that simulate agentic environments across Terminal, SWE, MCP, Search, OS, Android, and Web domains.", + "metric_id": "llm_stats.qwenworldbench.score", + "metric_name": "QwenWorldBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "qwenworldbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "QwenWorldBench", + "raw_categories": "[\"reasoning\",\"simulation\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.621, + "details": { + "raw_score": "0.621", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "qwenworldbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/qwenworldbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/qwenworldbench\"]", + "raw_score_id": "qwenworldbench::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "realworldqa::realworldqa-qwen3.7-plus", + "evaluation_name": "llm_stats.realworldqa", + "source_data": { + "dataset_name": "RealWorldQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/realworldqa", + "https://api.llm-stats.com/leaderboard/benchmarks/realworldqa" + ], + "additional_details": { + "raw_benchmark_id": "realworldqa", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "RealWorldQA is a benchmark designed to evaluate basic real-world spatial understanding capabilities of multimodal models. The initial release consists of over 700 anonymized images taken from vehicles and other real-world scenarios, each accompanied by a question and easily verifiable answer. Released by xAI as part of their Grok-1.5 Vision preview to test models' ability to understand natural scenes and spatial relationships in everyday visual contexts.", + "metric_id": "llm_stats.realworldqa.score", + "metric_name": "RealWorldQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "realworldqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "RealWorldQA", + "raw_categories": "[\"spatial_reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "25" + } + }, + "score_details": { + "score": 0.869, + "details": { + "raw_score": "0.869", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "realworldqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/realworldqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/realworldqa\"]", + "raw_score_id": "realworldqa::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "scicode::scicode-qwen3.7-plus", + "evaluation_name": "llm_stats.scicode", + "source_data": { + "dataset_name": "SciCode", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/scicode", + "https://api.llm-stats.com/leaderboard/benchmarks/scicode" + ], + "additional_details": { + "raw_benchmark_id": "scicode", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SciCode is a research coding benchmark curated by scientists that challenges language models to code solutions for scientific problems. It contains 338 subproblems decomposed from 80 challenging main problems across 16 natural science sub-fields including mathematics, physics, chemistry, biology, and materials science. Problems require knowledge recall, reasoning, and code synthesis skills.", + "metric_id": "llm_stats.scicode.score", + "metric_name": "SciCode score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "scicode", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SciCode", + "raw_categories": "[\"math\",\"physics\",\"reasoning\",\"biology\",\"chemistry\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.513, + "details": { + "raw_score": "0.513", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "scicode", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/scicode\",\"https://api.llm-stats.com/leaderboard/benchmarks/scicode\"]", + "raw_score_id": "scicode::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "screenspot-pro::screenspot-pro-qwen3.7-plus", + "evaluation_name": "llm_stats.screenspot-pro", + "source_data": { + "dataset_name": "ScreenSpot Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/screenspot-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/screenspot-pro" + ], + "additional_details": { + "raw_benchmark_id": "screenspot-pro", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "ScreenSpot-Pro is a novel GUI grounding benchmark designed to rigorously evaluate the grounding capabilities of multimodal large language models (MLLMs) in professional high-resolution computing environments. The benchmark comprises 1,581 instructions across 23 applications spanning 5 industries and 3 operating systems, featuring authentic high-resolution images from professional domains with expert annotations. Unlike previous benchmarks that focus on cropped screenshots in consumer applications, ScreenSpot-Pro addresses the complexity and diversity of real-world professional software scenarios, revealing significant performance gaps in current MLLM GUI perception capabilities.", + "metric_id": "llm_stats.screenspot-pro.score", + "metric_name": "ScreenSpot Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "screenspot-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ScreenSpot Pro", + "raw_categories": "[\"multimodal\",\"spatial_reasoning\",\"grounding\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.79, + "details": { + "raw_score": "0.79", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "screenspot-pro", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/screenspot-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/screenspot-pro\"]", + "raw_score_id": "screenspot-pro::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "simplevqa::simplevqa-qwen3.7-plus", + "evaluation_name": "llm_stats.simplevqa", + "source_data": { + "dataset_name": "SimpleVQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/simplevqa", + "https://api.llm-stats.com/leaderboard/benchmarks/simplevqa" + ], + "additional_details": { + "raw_benchmark_id": "simplevqa", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SimpleVQA is a visual question answering benchmark focused on simple queries.", + "metric_id": "llm_stats.simplevqa.score", + "metric_name": "SimpleVQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "simplevqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SimpleVQA", + "raw_categories": "[\"image_to_text\",\"multimodal\",\"general\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "13" + } + }, + "score_details": { + "score": 0.817, + "details": { + "raw_score": "0.817", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "simplevqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/simplevqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/simplevqa\"]", + "raw_score_id": "simplevqa::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "skillsbench::skillsbench-qwen3.7-plus", + "evaluation_name": "llm_stats.skillsbench", + "source_data": { + "dataset_name": "SkillsBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/skillsbench", + "https://api.llm-stats.com/leaderboard/benchmarks/skillsbench" + ], + "additional_details": { + "raw_benchmark_id": "skillsbench", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SkillsBench evaluates coding agents on self-contained programming tasks, measuring practical engineering skills across diverse software development scenarios.", + "metric_id": "llm_stats.skillsbench.score", + "metric_name": "SkillsBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "skillsbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SkillsBench", + "raw_categories": "[\"agents\",\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.549, + "details": { + "raw_score": "0.549", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "skillsbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/skillsbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/skillsbench\"]", + "raw_score_id": "skillsbench::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "spreadsheetbench-v1::spreadsheetbench-v1-qwen3.7-plus", + "evaluation_name": "llm_stats.spreadsheetbench-v1", + "source_data": { + "dataset_name": "SpreadSheetBench-v1", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/spreadsheetbench-v1", + "https://api.llm-stats.com/leaderboard/benchmarks/spreadsheetbench-v1" + ], + "additional_details": { + "raw_benchmark_id": "spreadsheetbench-v1", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SpreadSheetBench-v1 evaluates office automation agents on spreadsheet reasoning and manipulation tasks, measuring the ability to analyze, transform, and operate on spreadsheet data through tools.", + "metric_id": "llm_stats.spreadsheetbench-v1.score", + "metric_name": "SpreadSheetBench-v1 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "spreadsheetbench-v1", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SpreadSheetBench-v1", + "raw_categories": "[\"productivity\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.863, + "details": { + "raw_score": "0.863", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "spreadsheetbench-v1", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/spreadsheetbench-v1\",\"https://api.llm-stats.com/leaderboard/benchmarks/spreadsheetbench-v1\"]", + "raw_score_id": "spreadsheetbench-v1::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "supergpqa::supergpqa-qwen3.7-plus", + "evaluation_name": "llm_stats.supergpqa", + "source_data": { + "dataset_name": "SuperGPQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/supergpqa", + "https://api.llm-stats.com/leaderboard/benchmarks/supergpqa" + ], + "additional_details": { + "raw_benchmark_id": "supergpqa", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SuperGPQA is a comprehensive benchmark that evaluates large language models across 285 graduate-level academic disciplines. The benchmark contains 25,957 questions covering 13 broad disciplinary areas including Engineering, Medicine, Science, and Law, with specialized fields in light industry, agriculture, and service-oriented domains. It employs a Human-LLM collaborative filtering mechanism with over 80 expert annotators to create challenging questions that assess graduate-level knowledge and reasoning capabilities.", + "metric_id": "llm_stats.supergpqa.score", + "metric_name": "SuperGPQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "supergpqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SuperGPQA", + "raw_categories": "[\"legal\",\"math\",\"physics\",\"reasoning\",\"finance\",\"general\",\"healthcare\",\"chemistry\",\"economics\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "34" + } + }, + "score_details": { + "score": 0.714, + "details": { + "raw_score": "0.714", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "supergpqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/supergpqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/supergpqa\"]", + "raw_score_id": "supergpqa::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "surds::surds-qwen3.7-plus", + "evaluation_name": "llm_stats.surds", + "source_data": { + "dataset_name": "SURDS", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/surds", + "https://api.llm-stats.com/leaderboard/benchmarks/surds" + ], + "additional_details": { + "raw_benchmark_id": "surds", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SURDS is a benchmark for spatial understanding and reasoning in autonomous-driving scenes.", + "metric_id": "llm_stats.surds.score", + "metric_name": "SURDS score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "surds", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SURDS", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.772, + "details": { + "raw_score": "0.772", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "surds", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/surds\",\"https://api.llm-stats.com/leaderboard/benchmarks/surds\"]", + "raw_score_id": "surds::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-multilingual::swe-bench-multilingual-qwen3.7-plus", + "evaluation_name": "llm_stats.swe-bench-multilingual", + "source_data": { + "dataset_name": "SWE-bench Multilingual", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/swe-bench-multilingual", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-multilingual" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-multilingual", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A multilingual benchmark for issue resolving in software engineering that covers Java, TypeScript, JavaScript, Go, Rust, C, and C++. Contains 1,632 high-quality instances carefully annotated from 2,456 candidates by 68 expert annotators, designed to evaluate Large Language Models across diverse software ecosystems beyond Python.", + "metric_id": "llm_stats.swe-bench-multilingual.score", + "metric_name": "SWE-bench Multilingual score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-multilingual", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-bench Multilingual", + "raw_categories": "[\"reasoning\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "33" + } + }, + "score_details": { + "score": 0.758, + "details": { + "raw_score": "0.758", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "swe-bench-multilingual", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/swe-bench-multilingual\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-multilingual\"]", + "raw_score_id": "swe-bench-multilingual::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-pro::swe-bench-pro-qwen3.7-plus", + "evaluation_name": "llm_stats.swe-bench-pro", + "source_data": { + "dataset_name": "SWE-Bench Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/swe-bench-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-pro" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-pro", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SWE-Bench Pro is an advanced version of SWE-Bench that evaluates language models on complex, real-world software engineering tasks requiring extended reasoning and multi-step problem solving.", + "metric_id": "llm_stats.swe-bench-pro.score", + "metric_name": "SWE-Bench Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Bench Pro", + "raw_categories": "[\"reasoning\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "35" + } + }, + "score_details": { + "score": 0.576, + "details": { + "raw_score": "0.576", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "swe-bench-pro", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/swe-bench-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-pro\"]", + "raw_score_id": "swe-bench-pro::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2::terminal-bench-2-qwen3.7-plus", + "evaluation_name": "llm_stats.terminal-bench-2", + "source_data": { + "dataset_name": "Terminal-Bench 2.0", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/terminal-bench-2", + "https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2" + ], + "additional_details": { + "raw_benchmark_id": "terminal-bench-2", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Terminal-Bench 2.0 is an updated benchmark for testing AI agents' tool use ability to operate a computer via terminal. It evaluates how well models can handle real-world, end-to-end tasks autonomously, including compiling code, training models, setting up servers, system administration, security tasks, data science workflows, and cybersecurity vulnerabilities.", + "metric_id": "llm_stats.terminal-bench-2.score", + "metric_name": "Terminal-Bench 2.0 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "terminal-bench-2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Terminal-Bench 2.0", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "49" + } + }, + "score_details": { + "score": 0.703, + "details": { + "raw_score": "0.703", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "terminal-bench-2", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/terminal-bench-2\",\"https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2\"]", + "raw_score_id": "terminal-bench-2::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "tvbench::tvbench-qwen3.7-plus", + "evaluation_name": "llm_stats.tvbench", + "source_data": { + "dataset_name": "TVBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/tvbench", + "https://api.llm-stats.com/leaderboard/benchmarks/tvbench" + ], + "additional_details": { + "raw_benchmark_id": "tvbench", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "TVBench is a temporal video understanding benchmark evaluating reasoning over actions, events, and temporal dynamics in videos.", + "metric_id": "llm_stats.tvbench.score", + "metric_name": "TVBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "tvbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TVBench", + "raw_categories": "[\"multimodal\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.782, + "details": { + "raw_score": "0.782", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "tvbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/tvbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/tvbench\"]", + "raw_score_id": "tvbench::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "video-mme::video-mme-qwen3.7-plus", + "evaluation_name": "llm_stats.video-mme", + "source_data": { + "dataset_name": "Video-MME", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/video-mme", + "https://api.llm-stats.com/leaderboard/benchmarks/video-mme" + ], + "additional_details": { + "raw_benchmark_id": "video-mme", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Video-MME is the first-ever comprehensive evaluation benchmark of Multi-modal Large Language Models (MLLMs) in video analysis. It features 900 videos totaling 254 hours with 2,700 human-annotated question-answer pairs across 6 primary visual domains (Knowledge, Film & Television, Sports Competition, Life Record, Multilingual, and others) and 30 subfields. The benchmark evaluates models across diverse temporal dimensions (11 seconds to 1 hour), integrates multi-modal inputs including video frames, subtitles, and audio, and uses rigorous manual labeling by expert annotators for precise assessment.", + "metric_id": "llm_stats.video-mme.score", + "metric_name": "Video-MME score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "video-mme", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Video-MME", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "17" + } + }, + "score_details": { + "score": 0.88, + "details": { + "raw_score": "0.88", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "video-mme", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/video-mme\",\"https://api.llm-stats.com/leaderboard/benchmarks/video-mme\"]", + "raw_score_id": "video-mme::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "videommmu::videommmu-qwen3.7-plus", + "evaluation_name": "llm_stats.videommmu", + "source_data": { + "dataset_name": "VideoMMMU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/videommmu", + "https://api.llm-stats.com/leaderboard/benchmarks/videommmu" + ], + "additional_details": { + "raw_benchmark_id": "videommmu", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Video-MMMU evaluates Large Multimodal Models' ability to acquire knowledge from expert-level professional videos across six disciplines through three cognitive stages: perception, comprehension, and adaptation. Contains 300 videos and 900 human-annotated questions spanning Art, Business, Science, Medicine, Humanities, and Engineering.", + "metric_id": "llm_stats.videommmu.score", + "metric_name": "VideoMMMU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "videommmu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VideoMMMU", + "raw_categories": "[\"multimodal\",\"reasoning\",\"healthcare\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "26" + } + }, + "score_details": { + "score": 0.854, + "details": { + "raw_score": "0.854", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "videommmu", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/videommmu\",\"https://api.llm-stats.com/leaderboard/benchmarks/videommmu\"]", + "raw_score_id": "videommmu::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "visfactor::visfactor-qwen3.7-plus", + "evaluation_name": "llm_stats.visfactor", + "source_data": { + "dataset_name": "VisFactor", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/visfactor", + "https://api.llm-stats.com/leaderboard/benchmarks/visfactor" + ], + "additional_details": { + "raw_benchmark_id": "visfactor", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VisFactor is a benchmark evaluating fine-grained visual factor perception and reasoning over images.", + "metric_id": "llm_stats.visfactor.score", + "metric_name": "VisFactor score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "visfactor", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VisFactor", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.428, + "details": { + "raw_score": "0.428", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "visfactor", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/visfactor\",\"https://api.llm-stats.com/leaderboard/benchmarks/visfactor\"]", + "raw_score_id": "visfactor::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "vita-bench::vita-bench-qwen3.7-plus", + "evaluation_name": "llm_stats.vita-bench", + "source_data": { + "dataset_name": "VITA-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/vita-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/vita-bench" + ], + "additional_details": { + "raw_benchmark_id": "vita-bench", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VITA-Bench evaluates AI agents on real-world virtual task automation, measuring their ability to complete complex multi-step tasks in simulated environments.", + "metric_id": "llm_stats.vita-bench.score", + "metric_name": "VITA-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "vita-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VITA-Bench", + "raw_categories": "[\"reasoning\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.456, + "details": { + "raw_score": "0.456", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "vita-bench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/vita-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/vita-bench\"]", + "raw_score_id": "vita-bench::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "vladbench::vladbench-qwen3.7-plus", + "evaluation_name": "llm_stats.vladbench", + "source_data": { + "dataset_name": "VLADBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/vladbench", + "https://api.llm-stats.com/leaderboard/benchmarks/vladbench" + ], + "additional_details": { + "raw_benchmark_id": "vladbench", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "VLADBench is a vision-language autonomous-driving benchmark evaluating understanding of dynamic traffic scenes and participants.", + "metric_id": "llm_stats.vladbench.score", + "metric_name": "VLADBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "vladbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "VLADBench", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 0.772, + "details": { + "raw_score": "0.772", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "vladbench", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/vladbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/vladbench\"]", + "raw_score_id": "vladbench::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "wmt24::wmt24-qwen3.7-plus", + "evaluation_name": "llm_stats.wmt24", + "source_data": { + "dataset_name": "WMT24++", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/wmt24", + "https://api.llm-stats.com/leaderboard/benchmarks/wmt24++" + ], + "additional_details": { + "raw_benchmark_id": "wmt24++", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WMT24++ is a comprehensive multilingual machine translation benchmark that expands the WMT24 dataset to cover 55 languages and dialects. It includes human-written references and post-edits across four domains (literary, news, social, and speech) to evaluate machine translation systems and large language models across diverse linguistic contexts.", + "metric_id": "llm_stats.wmt24.score", + "metric_name": "WMT24++ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "wmt24++", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WMT24++", + "raw_categories": "[\"language\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.846, + "details": { + "raw_score": "0.846", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "wmt24++", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/wmt24\",\"https://api.llm-stats.com/leaderboard/benchmarks/wmt24++\"]", + "raw_score_id": "wmt24++::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "worldvqa::worldvqa-qwen3.7-plus", + "evaluation_name": "llm_stats.worldvqa", + "source_data": { + "dataset_name": "WorldVQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/qwen3.7-plus", + "https://llm-stats.com/benchmarks/worldvqa", + "https://api.llm-stats.com/leaderboard/benchmarks/worldvqa" + ], + "additional_details": { + "raw_benchmark_id": "worldvqa", + "raw_model_id": "qwen3.7-plus", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WorldVQA is a benchmark designed to evaluate atomic vision-centric world knowledge. It assesses models' ability to understand and reason about visual elements representing real-world knowledge.", + "metric_id": "llm_stats.worldvqa.score", + "metric_name": "WorldVQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "worldvqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WorldVQA", + "raw_categories": "[\"multimodal\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.611, + "details": { + "raw_score": "0.611", + "raw_score_field": "score", + "raw_model_id": "qwen3.7-plus", + "raw_benchmark_id": "worldvqa", + "source_urls_json": "[\"https://llm-stats.com/models/qwen3.7-plus\",\"https://llm-stats.com/benchmarks/worldvqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/worldvqa\"]", + "raw_score_id": "worldvqa::qwen3.7-plus", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/xiaomi/mimo-v2.5-pro/3b623760-e06b-4a40-929a-0ac211dc23ab.json b/data/llm_stats/xiaomi/mimo-v2.5-pro/3b623760-e06b-4a40-929a-0ac211dc23ab.json new file mode 100644 index 0000000000000000000000000000000000000000..079d58fb84dd5b4e1afdfc979593fb54e30a5a04 --- /dev/null +++ b/data/llm_stats/xiaomi/mimo-v2.5-pro/3b623760-e06b-4a40-929a-0ac211dc23ab.json @@ -0,0 +1,1400 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/xiaomi_mimo-v2.5-pro/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5-pro", + "id": "xiaomi/mimo-v2.5-pro", + "developer": "xiaomi", + "additional_details": { + "raw_id": "mimo-v2.5-pro", + "raw_model_name": "mimo-v2.5-pro", + "raw_display_name": "MiMo-V2.5-Pro", + "raw_organization_id": "xiaomi", + "raw_organization_name": "Xiaomi", + "raw_context_length": "1048576", + "raw_input_modalities": "[\"text\"]", + "raw_output_modalities": "[\"text\"]", + "raw_input_price": "0.435", + "raw_output_price": "0.87", + "raw_model_id": "mimo-v2.5-pro", + "raw_provider_slug": "xiaomi", + "raw_provider_name": "ZeroEval" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aime::aime-mimo-v2.5-pro", + "evaluation_name": "llm_stats.aime", + "source_data": { + "dataset_name": "AIME", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mimo-v2.5-pro", + "https://llm-stats.com/benchmarks/aime", + "https://api.llm-stats.com/leaderboard/benchmarks/aime" + ], + "additional_details": { + "raw_benchmark_id": "aime", + "raw_model_id": "mimo-v2.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "American Invitational Mathematics Examination (AIME) benchmark for evaluating mathematical reasoning capabilities of large language models. Contains 30 challenging mathematical problems from AIME 2024 competition that require multi-step reasoning and advanced mathematical insight. Each problem has an integer answer between 000-999.", + "metric_id": "llm_stats.aime.score", + "metric_name": "AIME score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aime", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AIME", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.373, + "details": { + "raw_score": "0.373", + "raw_score_field": "score", + "raw_model_id": "mimo-v2.5-pro", + "raw_benchmark_id": "aime", + "source_urls_json": "[\"https://llm-stats.com/models/mimo-v2.5-pro\",\"https://llm-stats.com/benchmarks/aime\",\"https://api.llm-stats.com/leaderboard/benchmarks/aime\"]", + "raw_score_id": "aime::mimo-v2.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "arc-c::arc-c-mimo-v2.5-pro", + "evaluation_name": "llm_stats.arc-c", + "source_data": { + "dataset_name": "ARC-C", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mimo-v2.5-pro", + "https://llm-stats.com/benchmarks/arc-c", + "https://api.llm-stats.com/leaderboard/benchmarks/arc-c" + ], + "additional_details": { + "raw_benchmark_id": "arc-c", + "raw_model_id": "mimo-v2.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "The AI2 Reasoning Challenge (ARC) Challenge Set is a multiple-choice question-answering benchmark containing grade-school level science questions that require advanced reasoning capabilities. ARC-C specifically contains questions that were answered incorrectly by both retrieval-based and word co-occurrence algorithms, making it a particularly challenging subset designed to test commonsense reasoning abilities in AI systems.", + "metric_id": "llm_stats.arc-c.score", + "metric_name": "ARC-C score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "arc-c", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "ARC-C", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "34" + } + }, + "score_details": { + "score": 0.972, + "details": { + "raw_score": "0.972", + "raw_score_field": "score", + "raw_model_id": "mimo-v2.5-pro", + "raw_benchmark_id": "arc-c", + "source_urls_json": "[\"https://llm-stats.com/models/mimo-v2.5-pro\",\"https://llm-stats.com/benchmarks/arc-c\",\"https://api.llm-stats.com/leaderboard/benchmarks/arc-c\"]", + "raw_score_id": "arc-c::mimo-v2.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "bbh::bbh-mimo-v2.5-pro", + "evaluation_name": "llm_stats.bbh", + "source_data": { + "dataset_name": "BBH", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mimo-v2.5-pro", + "https://llm-stats.com/benchmarks/bbh", + "https://api.llm-stats.com/leaderboard/benchmarks/bbh" + ], + "additional_details": { + "raw_benchmark_id": "bbh", + "raw_model_id": "mimo-v2.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Big-Bench Hard (BBH) is a suite of 23 challenging tasks selected from BIG-Bench for which prior language model evaluations did not outperform the average human-rater. These tasks require multi-step reasoning across diverse domains including arithmetic, logical reasoning, reading comprehension, and commonsense reasoning. The benchmark was designed to test capabilities believed to be beyond current language models and focuses on evaluating complex reasoning skills including temporal understanding, spatial reasoning, causal understanding, and deductive logical reasoning.", + "metric_id": "llm_stats.bbh.score", + "metric_name": "BBH score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "bbh", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BBH", + "raw_categories": "[\"language\",\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.884, + "details": { + "raw_score": "0.884", + "raw_score_field": "score", + "raw_model_id": "mimo-v2.5-pro", + "raw_benchmark_id": "bbh", + "source_urls_json": "[\"https://llm-stats.com/models/mimo-v2.5-pro\",\"https://llm-stats.com/benchmarks/bbh\",\"https://api.llm-stats.com/leaderboard/benchmarks/bbh\"]", + "raw_score_id": "bbh::mimo-v2.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "c-eval::c-eval-mimo-v2.5-pro", + "evaluation_name": "llm_stats.c-eval", + "source_data": { + "dataset_name": "C-Eval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mimo-v2.5-pro", + "https://llm-stats.com/benchmarks/c-eval", + "https://api.llm-stats.com/leaderboard/benchmarks/c-eval" + ], + "additional_details": { + "raw_benchmark_id": "c-eval", + "raw_model_id": "mimo-v2.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "C-Eval is a comprehensive Chinese evaluation suite designed to assess advanced knowledge and reasoning abilities of foundation models in a Chinese context. It comprises 13,948 multiple-choice questions across 52 diverse disciplines spanning humanities, science, and engineering, with four difficulty levels: middle school, high school, college, and professional. The benchmark includes C-Eval Hard, a subset of very challenging subjects requiring advanced reasoning abilities.", + "metric_id": "llm_stats.c-eval.score", + "metric_name": "C-Eval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "c-eval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "C-Eval", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.915, + "details": { + "raw_score": "0.915", + "raw_score_field": "score", + "raw_model_id": "mimo-v2.5-pro", + "raw_benchmark_id": "c-eval", + "source_urls_json": "[\"https://llm-stats.com/models/mimo-v2.5-pro\",\"https://llm-stats.com/benchmarks/c-eval\",\"https://api.llm-stats.com/leaderboard/benchmarks/c-eval\"]", + "raw_score_id": "c-eval::mimo-v2.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "claw-eval::claw-eval-mimo-v2.5-pro", + "evaluation_name": "llm_stats.claw-eval", + "source_data": { + "dataset_name": "Claw-Eval", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mimo-v2.5-pro", + "https://llm-stats.com/benchmarks/claw-eval", + "https://api.llm-stats.com/leaderboard/benchmarks/claw-eval" + ], + "additional_details": { + "raw_benchmark_id": "claw-eval", + "raw_model_id": "mimo-v2.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Claw-Eval tests real-world agentic task completion across complex multi-step scenarios, evaluating a model's ability to use tools, navigate environments, and complete end-to-end tasks autonomously.", + "metric_id": "llm_stats.claw-eval.score", + "metric_name": "Claw-Eval score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "claw-eval", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Claw-Eval", + "raw_categories": "[\"agents\",\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "12" + } + }, + "score_details": { + "score": 0.64, + "details": { + "raw_score": "0.64", + "raw_score_field": "score", + "raw_model_id": "mimo-v2.5-pro", + "raw_benchmark_id": "claw-eval", + "source_urls_json": "[\"https://llm-stats.com/models/mimo-v2.5-pro\",\"https://llm-stats.com/benchmarks/claw-eval\",\"https://api.llm-stats.com/leaderboard/benchmarks/claw-eval\"]", + "raw_score_id": "claw-eval::mimo-v2.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "cmmlu::cmmlu-mimo-v2.5-pro", + "evaluation_name": "llm_stats.cmmlu", + "source_data": { + "dataset_name": "CMMLU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mimo-v2.5-pro", + "https://llm-stats.com/benchmarks/cmmlu", + "https://api.llm-stats.com/leaderboard/benchmarks/cmmlu" + ], + "additional_details": { + "raw_benchmark_id": "cmmlu", + "raw_model_id": "mimo-v2.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CMMLU (Chinese Massive Multitask Language Understanding) is a comprehensive Chinese benchmark that evaluates the knowledge and reasoning capabilities of large language models across 67 different subject topics. The benchmark covers natural sciences, social sciences, engineering, and humanities with multiple-choice questions ranging from basic to advanced professional levels.", + "metric_id": "llm_stats.cmmlu.score", + "metric_name": "CMMLU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "cmmlu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CMMLU", + "raw_categories": "[\"language\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "6" + } + }, + "score_details": { + "score": 0.902, + "details": { + "raw_score": "0.902", + "raw_score_field": "score", + "raw_model_id": "mimo-v2.5-pro", + "raw_benchmark_id": "cmmlu", + "source_urls_json": "[\"https://llm-stats.com/models/mimo-v2.5-pro\",\"https://llm-stats.com/benchmarks/cmmlu\",\"https://api.llm-stats.com/leaderboard/benchmarks/cmmlu\"]", + "raw_score_id": "cmmlu::mimo-v2.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "drop::drop-mimo-v2.5-pro", + "evaluation_name": "llm_stats.drop", + "source_data": { + "dataset_name": "DROP", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mimo-v2.5-pro", + "https://llm-stats.com/benchmarks/drop", + "https://api.llm-stats.com/leaderboard/benchmarks/drop" + ], + "additional_details": { + "raw_benchmark_id": "drop", + "raw_model_id": "mimo-v2.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "DROP (Discrete Reasoning Over Paragraphs) is a reading comprehension benchmark requiring discrete reasoning over paragraph content. It contains crowdsourced, adversarially-created questions that require resolving references and performing discrete operations like addition, counting, or sorting, demanding comprehensive paragraph understanding beyond paraphrase-and-entity-typing shortcuts.", + "metric_id": "llm_stats.drop.score", + "metric_name": "DROP score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "drop", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "DROP", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "30" + } + }, + "score_details": { + "score": 0.863, + "details": { + "raw_score": "0.863", + "raw_score_field": "score", + "raw_model_id": "mimo-v2.5-pro", + "raw_benchmark_id": "drop", + "source_urls_json": "[\"https://llm-stats.com/models/mimo-v2.5-pro\",\"https://llm-stats.com/benchmarks/drop\",\"https://api.llm-stats.com/leaderboard/benchmarks/drop\"]", + "raw_score_id": "drop::mimo-v2.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "frontier-swe-impl::frontier-swe-impl-mimo-v2.5-pro", + "evaluation_name": "llm_stats.frontier-swe-impl", + "source_data": { + "dataset_name": "FrontierSWE (Impl.)", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mimo-v2.5-pro", + "https://llm-stats.com/benchmarks/frontier-swe-impl", + "https://api.llm-stats.com/leaderboard/benchmarks/frontier-swe-impl" + ], + "additional_details": { + "raw_benchmark_id": "frontier-swe-impl", + "raw_model_id": "mimo-v2.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "FrontierSWE (Impl.) evaluates software engineering implementation ability and reports model ranking on implementation tasks. Lower rank is better.", + "metric_id": "llm_stats.frontier-swe-impl.score", + "metric_name": "FrontierSWE (Impl.) score", + "metric_kind": "benchmark_score", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "raw_benchmark_id": "frontier-swe-impl", + "raw_score_field": "score", + "bound_strategy": "inferred_percent", + "raw_name": "FrontierSWE (Impl.)", + "raw_categories": "[\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "1" + } + }, + "score_details": { + "score": 3.4, + "details": { + "raw_score": "3.4", + "raw_score_field": "score", + "raw_model_id": "mimo-v2.5-pro", + "raw_benchmark_id": "frontier-swe-impl", + "source_urls_json": "[\"https://llm-stats.com/models/mimo-v2.5-pro\",\"https://llm-stats.com/benchmarks/frontier-swe-impl\",\"https://api.llm-stats.com/leaderboard/benchmarks/frontier-swe-impl\"]", + "raw_score_id": "frontier-swe-impl::mimo-v2.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "global-mmlu::global-mmlu-mimo-v2.5-pro", + "evaluation_name": "llm_stats.global-mmlu", + "source_data": { + "dataset_name": "Global-MMLU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mimo-v2.5-pro", + "https://llm-stats.com/benchmarks/global-mmlu", + "https://api.llm-stats.com/leaderboard/benchmarks/global-mmlu" + ], + "additional_details": { + "raw_benchmark_id": "global-mmlu", + "raw_model_id": "mimo-v2.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A comprehensive multilingual benchmark covering 42 languages that addresses cultural and linguistic biases in evaluation, with improved translation quality and culturally sensitive question subsets.", + "metric_id": "llm_stats.global-mmlu.score", + "metric_name": "Global-MMLU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "global-mmlu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Global-MMLU", + "raw_categories": "[\"language\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.836, + "details": { + "raw_score": "0.836", + "raw_score_field": "score", + "raw_model_id": "mimo-v2.5-pro", + "raw_benchmark_id": "global-mmlu", + "source_urls_json": "[\"https://llm-stats.com/models/mimo-v2.5-pro\",\"https://llm-stats.com/benchmarks/global-mmlu\",\"https://api.llm-stats.com/leaderboard/benchmarks/global-mmlu\"]", + "raw_score_id": "global-mmlu::mimo-v2.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "graphwalks::graphwalks-mimo-v2.5-pro", + "evaluation_name": "llm_stats.graphwalks", + "source_data": { + "dataset_name": "GraphWalks", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mimo-v2.5-pro", + "https://llm-stats.com/benchmarks/graphwalks", + "https://api.llm-stats.com/leaderboard/benchmarks/graphwalks" + ], + "additional_details": { + "raw_benchmark_id": "graphwalks", + "raw_model_id": "mimo-v2.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "GraphWalks is a synthetic multi-hop long-context reasoning benchmark in which a model is given an edge-list representation of a graph and must traverse it to find neighboring nodes (via breadth-first search) or parent nodes for a given start node. Performance is reported as F1 of the model-predicted answer set versus the ground truth.", + "metric_id": "llm_stats.graphwalks.score", + "metric_name": "GraphWalks score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "graphwalks", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GraphWalks", + "raw_categories": "[\"long_context\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.62, + "details": { + "raw_score": "0.62", + "raw_score_field": "score", + "raw_model_id": "mimo-v2.5-pro", + "raw_benchmark_id": "graphwalks", + "source_urls_json": "[\"https://llm-stats.com/models/mimo-v2.5-pro\",\"https://llm-stats.com/benchmarks/graphwalks\",\"https://api.llm-stats.com/leaderboard/benchmarks/graphwalks\"]", + "raw_score_id": "graphwalks::mimo-v2.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "gsm8k::gsm8k-mimo-v2.5-pro", + "evaluation_name": "llm_stats.gsm8k", + "source_data": { + "dataset_name": "GSM8k", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mimo-v2.5-pro", + "https://llm-stats.com/benchmarks/gsm8k", + "https://api.llm-stats.com/leaderboard/benchmarks/gsm8k" + ], + "additional_details": { + "raw_benchmark_id": "gsm8k", + "raw_model_id": "mimo-v2.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Grade School Math 8K, a dataset of 8.5K high-quality linguistically diverse grade school math word problems requiring multi-step reasoning and elementary arithmetic operations.", + "metric_id": "llm_stats.gsm8k.score", + "metric_name": "GSM8k score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "gsm8k", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "GSM8k", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "48" + } + }, + "score_details": { + "score": 0.996, + "details": { + "raw_score": "0.996", + "raw_score_field": "score", + "raw_model_id": "mimo-v2.5-pro", + "raw_benchmark_id": "gsm8k", + "source_urls_json": "[\"https://llm-stats.com/models/mimo-v2.5-pro\",\"https://llm-stats.com/benchmarks/gsm8k\",\"https://api.llm-stats.com/leaderboard/benchmarks/gsm8k\"]", + "raw_score_id": "gsm8k::mimo-v2.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hellaswag::hellaswag-mimo-v2.5-pro", + "evaluation_name": "llm_stats.hellaswag", + "source_data": { + "dataset_name": "HellaSwag", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mimo-v2.5-pro", + "https://llm-stats.com/benchmarks/hellaswag", + "https://api.llm-stats.com/leaderboard/benchmarks/hellaswag" + ], + "additional_details": { + "raw_benchmark_id": "hellaswag", + "raw_model_id": "mimo-v2.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A challenging commonsense natural language inference dataset that uses Adversarial Filtering to create questions trivial for humans (>95% accuracy) but difficult for state-of-the-art models, requiring completion of sentence endings based on physical situations and everyday activities", + "metric_id": "llm_stats.hellaswag.score", + "metric_name": "HellaSwag score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hellaswag", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HellaSwag", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.898, + "details": { + "raw_score": "0.898", + "raw_score_field": "score", + "raw_model_id": "mimo-v2.5-pro", + "raw_benchmark_id": "hellaswag", + "source_urls_json": "[\"https://llm-stats.com/models/mimo-v2.5-pro\",\"https://llm-stats.com/benchmarks/hellaswag\",\"https://api.llm-stats.com/leaderboard/benchmarks/hellaswag\"]", + "raw_score_id": "hellaswag::mimo-v2.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "humaneval::humaneval-mimo-v2.5-pro", + "evaluation_name": "llm_stats.humaneval", + "source_data": { + "dataset_name": "HumanEval+", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mimo-v2.5-pro", + "https://llm-stats.com/benchmarks/humaneval", + "https://api.llm-stats.com/leaderboard/benchmarks/humaneval+" + ], + "additional_details": { + "raw_benchmark_id": "humaneval+", + "raw_model_id": "mimo-v2.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Enhanced version of HumanEval that extends the original test cases by 80x using EvalPlus framework for rigorous evaluation of LLM-synthesized code functional correctness, detecting previously undetected wrong code", + "metric_id": "llm_stats.humaneval.score", + "metric_name": "HumanEval+ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "humaneval+", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HumanEval+", + "raw_categories": "[\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "10" + } + }, + "score_details": { + "score": 0.756, + "details": { + "raw_score": "0.756", + "raw_score_field": "score", + "raw_model_id": "mimo-v2.5-pro", + "raw_benchmark_id": "humaneval+", + "source_urls_json": "[\"https://llm-stats.com/models/mimo-v2.5-pro\",\"https://llm-stats.com/benchmarks/humaneval\",\"https://api.llm-stats.com/leaderboard/benchmarks/humaneval+\"]", + "raw_score_id": "humaneval+::mimo-v2.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "math::math-mimo-v2.5-pro", + "evaluation_name": "llm_stats.math", + "source_data": { + "dataset_name": "MATH", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mimo-v2.5-pro", + "https://llm-stats.com/benchmarks/math", + "https://api.llm-stats.com/leaderboard/benchmarks/math" + ], + "additional_details": { + "raw_benchmark_id": "math", + "raw_model_id": "mimo-v2.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MATH dataset contains 12,500 challenging competition mathematics problems from AMC 10, AMC 12, AIME, and other mathematics competitions. Each problem includes full step-by-step solutions and spans multiple difficulty levels (1-5) across seven mathematical subjects including Prealgebra, Algebra, Number Theory, Counting and Probability, Geometry, Intermediate Algebra, and Precalculus.", + "metric_id": "llm_stats.math.score", + "metric_name": "MATH score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "math", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MATH", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "71" + } + }, + "score_details": { + "score": 0.862, + "details": { + "raw_score": "0.862", + "raw_score_field": "score", + "raw_model_id": "mimo-v2.5-pro", + "raw_benchmark_id": "math", + "source_urls_json": "[\"https://llm-stats.com/models/mimo-v2.5-pro\",\"https://llm-stats.com/benchmarks/math\",\"https://api.llm-stats.com/leaderboard/benchmarks/math\"]", + "raw_score_id": "math::mimo-v2.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mbpp::mbpp-mimo-v2.5-pro", + "evaluation_name": "llm_stats.mbpp", + "source_data": { + "dataset_name": "MBPP+", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mimo-v2.5-pro", + "https://llm-stats.com/benchmarks/mbpp", + "https://api.llm-stats.com/leaderboard/benchmarks/mbpp+" + ], + "additional_details": { + "raw_benchmark_id": "mbpp+", + "raw_model_id": "mimo-v2.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MBPP+ is an enhanced version of MBPP (Mostly Basic Python Problems) with significantly more test cases (35x) for more rigorous evaluation. MBPP is a benchmark of 974 crowd-sourced Python programming problems designed to be solvable by entry-level programmers, covering programming fundamentals and standard library functionality.", + "metric_id": "llm_stats.mbpp.score", + "metric_name": "MBPP+ score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mbpp+", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MBPP+", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 0.741, + "details": { + "raw_score": "0.741", + "raw_score_field": "score", + "raw_model_id": "mimo-v2.5-pro", + "raw_benchmark_id": "mbpp+", + "source_urls_json": "[\"https://llm-stats.com/models/mimo-v2.5-pro\",\"https://llm-stats.com/benchmarks/mbpp\",\"https://api.llm-stats.com/leaderboard/benchmarks/mbpp+\"]", + "raw_score_id": "mbpp+::mimo-v2.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mimo-coding-bench::mimo-coding-bench-mimo-v2.5-pro", + "evaluation_name": "llm_stats.mimo-coding-bench", + "source_data": { + "dataset_name": "MiMo Coding Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mimo-v2.5-pro", + "https://llm-stats.com/benchmarks/mimo-coding-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/mimo-coding-bench" + ], + "additional_details": { + "raw_benchmark_id": "mimo-coding-bench", + "raw_model_id": "mimo-v2.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MiMo Coding Bench evaluates coding-agent capabilities on software engineering tasks reported with the MiMo model family.", + "metric_id": "llm_stats.mimo-coding-bench.score", + "metric_name": "MiMo Coding Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mimo-coding-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MiMo Coding Bench", + "raw_categories": "[\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.737, + "details": { + "raw_score": "0.737", + "raw_score_field": "score", + "raw_model_id": "mimo-v2.5-pro", + "raw_benchmark_id": "mimo-coding-bench", + "source_urls_json": "[\"https://llm-stats.com/models/mimo-v2.5-pro\",\"https://llm-stats.com/benchmarks/mimo-coding-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/mimo-coding-bench\"]", + "raw_score_id": "mimo-coding-bench::mimo-v2.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu-redux::mmlu-redux-mimo-v2.5-pro", + "evaluation_name": "llm_stats.mmlu-redux", + "source_data": { + "dataset_name": "MMLU-Redux", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mimo-v2.5-pro", + "https://llm-stats.com/benchmarks/mmlu-redux", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu-redux" + ], + "additional_details": { + "raw_benchmark_id": "mmlu-redux", + "raw_model_id": "mimo-v2.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "An improved version of the MMLU benchmark featuring manually re-annotated questions to identify and correct errors in the original dataset. Provides more reliable evaluation metrics for language models by addressing dataset quality issues found in the original MMLU.", + "metric_id": "llm_stats.mmlu-redux.score", + "metric_name": "MMLU-Redux score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu-redux", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU-Redux", + "raw_categories": "[\"language\",\"math\",\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "48" + } + }, + "score_details": { + "score": 0.928, + "details": { + "raw_score": "0.928", + "raw_score_field": "score", + "raw_model_id": "mimo-v2.5-pro", + "raw_benchmark_id": "mmlu-redux", + "source_urls_json": "[\"https://llm-stats.com/models/mimo-v2.5-pro\",\"https://llm-stats.com/benchmarks/mmlu-redux\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu-redux\"]", + "raw_score_id": "mmlu-redux::mimo-v2.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mmlu::mmlu-mimo-v2.5-pro", + "evaluation_name": "llm_stats.mmlu", + "source_data": { + "dataset_name": "MMLU", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mimo-v2.5-pro", + "https://llm-stats.com/benchmarks/mmlu", + "https://api.llm-stats.com/leaderboard/benchmarks/mmlu" + ], + "additional_details": { + "raw_benchmark_id": "mmlu", + "raw_model_id": "mimo-v2.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Massive Multitask Language Understanding benchmark testing knowledge across 57 diverse subjects including STEM, humanities, social sciences, and professional domains", + "metric_id": "llm_stats.mmlu.score", + "metric_name": "MMLU score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mmlu", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MMLU", + "raw_categories": "[\"language\",\"legal\",\"math\",\"reasoning\",\"finance\",\"general\",\"healthcare\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "100" + } + }, + "score_details": { + "score": 0.894, + "details": { + "raw_score": "0.894", + "raw_score_field": "score", + "raw_model_id": "mimo-v2.5-pro", + "raw_benchmark_id": "mmlu", + "source_urls_json": "[\"https://llm-stats.com/models/mimo-v2.5-pro\",\"https://llm-stats.com/benchmarks/mmlu\",\"https://api.llm-stats.com/leaderboard/benchmarks/mmlu\"]", + "raw_score_id": "mmlu::mimo-v2.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-pro::swe-bench-pro-mimo-v2.5-pro", + "evaluation_name": "llm_stats.swe-bench-pro", + "source_data": { + "dataset_name": "SWE-Bench Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mimo-v2.5-pro", + "https://llm-stats.com/benchmarks/swe-bench-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-pro" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-pro", + "raw_model_id": "mimo-v2.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SWE-Bench Pro is an advanced version of SWE-Bench that evaluates language models on complex, real-world software engineering tasks requiring extended reasoning and multi-step problem solving.", + "metric_id": "llm_stats.swe-bench-pro.score", + "metric_name": "SWE-Bench Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Bench Pro", + "raw_categories": "[\"reasoning\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "35" + } + }, + "score_details": { + "score": 0.572, + "details": { + "raw_score": "0.572", + "raw_score_field": "score", + "raw_model_id": "mimo-v2.5-pro", + "raw_benchmark_id": "swe-bench-pro", + "source_urls_json": "[\"https://llm-stats.com/models/mimo-v2.5-pro\",\"https://llm-stats.com/benchmarks/swe-bench-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-pro\"]", + "raw_score_id": "swe-bench-pro::mimo-v2.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-verified-agentless::swe-bench-verified-agentless-mimo-v2.5-pro", + "evaluation_name": "llm_stats.swe-bench-verified-agentless", + "source_data": { + "dataset_name": "SWE-bench Verified (Agentless)", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mimo-v2.5-pro", + "https://llm-stats.com/benchmarks/swe-bench-verified-agentless", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-verified-(agentless)" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-verified-(agentless)", + "raw_model_id": "mimo-v2.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A human-validated subset of SWE-bench that evaluates language models' ability to resolve real-world GitHub issues using an agentless approach. The benchmark tests models on software engineering problems requiring understanding and coordinating changes across multiple functions, classes, and files simultaneously.", + "metric_id": "llm_stats.swe-bench-verified-agentless.score", + "metric_name": "SWE-bench Verified (Agentless) score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-verified-(agentless)", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-bench Verified (Agentless)", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "2" + } + }, + "score_details": { + "score": 0.357, + "details": { + "raw_score": "0.357", + "raw_score_field": "score", + "raw_model_id": "mimo-v2.5-pro", + "raw_benchmark_id": "swe-bench-verified-(agentless)", + "source_urls_json": "[\"https://llm-stats.com/models/mimo-v2.5-pro\",\"https://llm-stats.com/benchmarks/swe-bench-verified-agentless\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-verified-(agentless)\"]", + "raw_score_id": "swe-bench-verified-(agentless)::mimo-v2.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-verified::swe-bench-verified-mimo-v2.5-pro", + "evaluation_name": "llm_stats.swe-bench-verified", + "source_data": { + "dataset_name": "SWE-Bench Verified", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mimo-v2.5-pro", + "https://llm-stats.com/benchmarks/swe-bench-verified", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-verified" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-verified", + "raw_model_id": "mimo-v2.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A verified subset of 500 software engineering problems from real GitHub issues, validated by human annotators for evaluating language models' ability to resolve real-world coding issues by generating patches for Python codebases.", + "metric_id": "llm_stats.swe-bench-verified.score", + "metric_name": "SWE-Bench Verified score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-verified", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Bench Verified", + "raw_categories": "[\"reasoning\",\"frontend_development\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "103" + } + }, + "score_details": { + "score": 0.789, + "details": { + "raw_score": "0.789", + "raw_score_field": "score", + "raw_model_id": "mimo-v2.5-pro", + "raw_benchmark_id": "swe-bench-verified", + "source_urls_json": "[\"https://llm-stats.com/models/mimo-v2.5-pro\",\"https://llm-stats.com/benchmarks/swe-bench-verified\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-verified\"]", + "raw_score_id": "swe-bench-verified::mimo-v2.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "tau3-bench::tau3-bench-mimo-v2.5-pro", + "evaluation_name": "llm_stats.tau3-bench", + "source_data": { + "dataset_name": "TAU3-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mimo-v2.5-pro", + "https://llm-stats.com/benchmarks/tau3-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/tau3-bench" + ], + "additional_details": { + "raw_benchmark_id": "tau3-bench", + "raw_model_id": "mimo-v2.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "TAU3-Bench is a benchmark for evaluating general-purpose agent capabilities, testing models on multi-turn interactions with simulated user models, retrieval, and complex decision-making scenarios.", + "metric_id": "llm_stats.tau3-bench.score", + "metric_name": "TAU3-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "tau3-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TAU3-Bench", + "raw_categories": "[\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.729, + "details": { + "raw_score": "0.729", + "raw_score_field": "score", + "raw_model_id": "mimo-v2.5-pro", + "raw_benchmark_id": "tau3-bench", + "source_urls_json": "[\"https://llm-stats.com/models/mimo-v2.5-pro\",\"https://llm-stats.com/benchmarks/tau3-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/tau3-bench\"]", + "raw_score_id": "tau3-bench::mimo-v2.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2::terminal-bench-2-mimo-v2.5-pro", + "evaluation_name": "llm_stats.terminal-bench-2", + "source_data": { + "dataset_name": "Terminal-Bench 2.0", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mimo-v2.5-pro", + "https://llm-stats.com/benchmarks/terminal-bench-2", + "https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2" + ], + "additional_details": { + "raw_benchmark_id": "terminal-bench-2", + "raw_model_id": "mimo-v2.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Terminal-Bench 2.0 is an updated benchmark for testing AI agents' tool use ability to operate a computer via terminal. It evaluates how well models can handle real-world, end-to-end tasks autonomously, including compiling code, training models, setting up servers, system administration, security tasks, data science workflows, and cybersecurity vulnerabilities.", + "metric_id": "llm_stats.terminal-bench-2.score", + "metric_name": "Terminal-Bench 2.0 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "terminal-bench-2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Terminal-Bench 2.0", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "49" + } + }, + "score_details": { + "score": 0.684, + "details": { + "raw_score": "0.684", + "raw_score_field": "score", + "raw_model_id": "mimo-v2.5-pro", + "raw_benchmark_id": "terminal-bench-2", + "source_urls_json": "[\"https://llm-stats.com/models/mimo-v2.5-pro\",\"https://llm-stats.com/benchmarks/terminal-bench-2\",\"https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2\"]", + "raw_score_id": "terminal-bench-2::mimo-v2.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "triviaqa::triviaqa-mimo-v2.5-pro", + "evaluation_name": "llm_stats.triviaqa", + "source_data": { + "dataset_name": "TriviaQA", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mimo-v2.5-pro", + "https://llm-stats.com/benchmarks/triviaqa", + "https://api.llm-stats.com/leaderboard/benchmarks/triviaqa" + ], + "additional_details": { + "raw_benchmark_id": "triviaqa", + "raw_model_id": "mimo-v2.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "A large-scale reading comprehension dataset containing over 650K question-answer-evidence triples. TriviaQA includes 95K question-answer pairs authored by trivia enthusiasts and independently gathered evidence documents (six per question on average) that provide high quality distant supervision for answering the questions. The dataset features relatively complex, compositional questions with considerable syntactic and lexical variability, requiring cross-sentence reasoning to find answers.", + "metric_id": "llm_stats.triviaqa.score", + "metric_name": "TriviaQA score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "triviaqa", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TriviaQA", + "raw_categories": "[\"reasoning\",\"general\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.813, + "details": { + "raw_score": "0.813", + "raw_score_field": "score", + "raw_model_id": "mimo-v2.5-pro", + "raw_benchmark_id": "triviaqa", + "source_urls_json": "[\"https://llm-stats.com/models/mimo-v2.5-pro\",\"https://llm-stats.com/benchmarks/triviaqa\",\"https://api.llm-stats.com/leaderboard/benchmarks/triviaqa\"]", + "raw_score_id": "triviaqa::mimo-v2.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "wildclawbench::wildclawbench-mimo-v2.5-pro", + "evaluation_name": "llm_stats.wildclawbench", + "source_data": { + "dataset_name": "WildClawBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mimo-v2.5-pro", + "https://llm-stats.com/benchmarks/wildclawbench", + "https://api.llm-stats.com/leaderboard/benchmarks/wildclawbench" + ], + "additional_details": { + "raw_benchmark_id": "wildclawbench", + "raw_model_id": "mimo-v2.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WildClawBench is an agentic coding benchmark from InternLM/Claw-Eval that reports overall model performance on real-world tool-using development tasks.", + "metric_id": "llm_stats.wildclawbench.score", + "metric_name": "WildClawBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "wildclawbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "WildClawBench", + "raw_categories": "[\"agents\",\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "3" + } + }, + "score_details": { + "score": 0.43, + "details": { + "raw_score": "0.43", + "raw_score_field": "score", + "raw_model_id": "mimo-v2.5-pro", + "raw_benchmark_id": "wildclawbench", + "source_urls_json": "[\"https://llm-stats.com/models/mimo-v2.5-pro\",\"https://llm-stats.com/benchmarks/wildclawbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/wildclawbench\"]", + "raw_score_id": "wildclawbench::mimo-v2.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "winogrande::winogrande-mimo-v2.5-pro", + "evaluation_name": "llm_stats.winogrande", + "source_data": { + "dataset_name": "Winogrande", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/mimo-v2.5-pro", + "https://llm-stats.com/benchmarks/winogrande", + "https://api.llm-stats.com/leaderboard/benchmarks/winogrande" + ], + "additional_details": { + "raw_benchmark_id": "winogrande", + "raw_model_id": "mimo-v2.5-pro", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "WinoGrande: An Adversarial Winograd Schema Challenge at Scale. A large-scale dataset of 44,000 pronoun resolution problems designed to test machine commonsense reasoning. Uses adversarial filtering to reduce spurious biases and provides a more robust evaluation of whether AI systems truly understand commonsense or exploit statistical shortcuts. Current best AI methods achieve 59.4-79.1% accuracy, significantly below human performance of 94.0%.", + "metric_id": "llm_stats.winogrande.score", + "metric_name": "Winogrande score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "winogrande", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Winogrande", + "raw_categories": "[\"language\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "22" + } + }, + "score_details": { + "score": 0.856, + "details": { + "raw_score": "0.856", + "raw_score_field": "score", + "raw_model_id": "mimo-v2.5-pro", + "raw_benchmark_id": "winogrande", + "source_urls_json": "[\"https://llm-stats.com/models/mimo-v2.5-pro\",\"https://llm-stats.com/benchmarks/winogrande\",\"https://api.llm-stats.com/leaderboard/benchmarks/winogrande\"]", + "raw_score_id": "winogrande::mimo-v2.5-pro", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/zai-org/glm-5.1/fe3c4d98-d4fd-4a6b-abea-d5d445b7ee27.json b/data/llm_stats/zai-org/glm-5.1/fe3c4d98-d4fd-4a6b-abea-d5d445b7ee27.json new file mode 100644 index 0000000000000000000000000000000000000000..778272dfc39f252bda611fb89dd445be63319949 --- /dev/null +++ b/data/llm_stats/zai-org/glm-5.1/fe3c4d98-d4fd-4a6b-abea-d5d445b7ee27.json @@ -0,0 +1,776 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/zai-org_glm-5.1/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.1", + "id": "zai-org/glm-5.1", + "developer": "zai-org", + "additional_details": { + "raw_id": "glm-5.1", + "raw_model_name": "glm-5.1", + "raw_display_name": "GLM-5.1", + "raw_organization_id": "zai-org", + "raw_organization_name": "Zhipu AI", + "raw_context_length": "200000", + "raw_input_modalities": "[\"text\"]", + "raw_output_modalities": "[\"text\"]", + "raw_input_price": "1.4", + "raw_output_price": "4.4", + "raw_model_id": "glm-5.1", + "raw_provider_slug": "zai-org", + "raw_provider_name": "ZeroEval" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aime-2026::aime-2026-glm-5.1", + "evaluation_name": "llm_stats.aime-2026", + "source_data": { + "dataset_name": "AIME 2026", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/glm-5.1", + "https://llm-stats.com/benchmarks/aime-2026", + "https://api.llm-stats.com/leaderboard/benchmarks/aime-2026" + ], + "additional_details": { + "raw_benchmark_id": "aime-2026", + "raw_model_id": "glm-5.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "All 30 problems from the 2026 American Invitational Mathematics Examination (AIME I and AIME II), testing olympiad-level mathematical reasoning with integer answers from 000-999. Used as an AI benchmark to evaluate large language models' ability to solve complex mathematical problems requiring multi-step logical deductions and structured symbolic reasoning.", + "metric_id": "llm_stats.aime-2026.score", + "metric_name": "AIME 2026 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "aime-2026", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "AIME 2026", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "17" + } + }, + "score_details": { + "score": 0.953, + "details": { + "raw_score": "0.953", + "raw_score_field": "score", + "raw_model_id": "glm-5.1", + "raw_benchmark_id": "aime-2026", + "source_urls_json": "[\"https://llm-stats.com/models/glm-5.1\",\"https://llm-stats.com/benchmarks/aime-2026\",\"https://api.llm-stats.com/leaderboard/benchmarks/aime-2026\"]", + "raw_score_id": "aime-2026::glm-5.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "browsecomp::browsecomp-glm-5.1", + "evaluation_name": "llm_stats.browsecomp", + "source_data": { + "dataset_name": "BrowseComp", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/glm-5.1", + "https://llm-stats.com/benchmarks/browsecomp", + "https://api.llm-stats.com/leaderboard/benchmarks/browsecomp" + ], + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_model_id": "glm-5.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BrowseComp is a benchmark comprising 1,266 questions that challenge AI agents to persistently navigate the internet in search of hard-to-find, entangled information. The benchmark measures agents' ability to exercise persistence in information gathering, demonstrate creativity in web navigation, and find concise, verifiable answers. Despite the difficulty of the questions, BrowseComp is simple and easy-to-use, as predicted answers are short and easily verifiable against reference answers.", + "metric_id": "llm_stats.browsecomp.score", + "metric_name": "BrowseComp score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BrowseComp", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "52" + } + }, + "score_details": { + "score": 0.793, + "details": { + "raw_score": "0.793", + "raw_score_field": "score", + "raw_model_id": "glm-5.1", + "raw_benchmark_id": "browsecomp", + "source_urls_json": "[\"https://llm-stats.com/models/glm-5.1\",\"https://llm-stats.com/benchmarks/browsecomp\",\"https://api.llm-stats.com/leaderboard/benchmarks/browsecomp\"]", + "raw_score_id": "browsecomp::glm-5.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "cybergym::cybergym-glm-5.1", + "evaluation_name": "llm_stats.cybergym", + "source_data": { + "dataset_name": "CyberGym", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/glm-5.1", + "https://llm-stats.com/benchmarks/cybergym", + "https://api.llm-stats.com/leaderboard/benchmarks/cybergym" + ], + "additional_details": { + "raw_benchmark_id": "cybergym", + "raw_model_id": "glm-5.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "CyberGym is a benchmark for evaluating AI agents on cybersecurity tasks, testing their ability to identify vulnerabilities, perform security analysis, and complete security-related challenges in a controlled environment.", + "metric_id": "llm_stats.cybergym.score", + "metric_name": "CyberGym score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "cybergym", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "CyberGym", + "raw_categories": "[\"safety\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "9" + } + }, + "score_details": { + "score": 0.687, + "details": { + "raw_score": "0.687", + "raw_score_field": "score", + "raw_model_id": "glm-5.1", + "raw_benchmark_id": "cybergym", + "source_urls_json": "[\"https://llm-stats.com/models/glm-5.1\",\"https://llm-stats.com/benchmarks/cybergym\",\"https://api.llm-stats.com/leaderboard/benchmarks/cybergym\"]", + "raw_score_id": "cybergym::glm-5.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hmmt-2025::hmmt-2025-glm-5.1", + "evaluation_name": "llm_stats.hmmt-2025", + "source_data": { + "dataset_name": "HMMT 2025", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/glm-5.1", + "https://llm-stats.com/benchmarks/hmmt-2025", + "https://api.llm-stats.com/leaderboard/benchmarks/hmmt-2025" + ], + "additional_details": { + "raw_benchmark_id": "hmmt-2025", + "raw_model_id": "glm-5.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Harvard-MIT Mathematics Tournament 2025 - A prestigious student-organized mathematics competition for high school students featuring two tournaments (November 2025 at MIT and February 2026 at Harvard) with individual tests, team rounds, and guts rounds", + "metric_id": "llm_stats.hmmt-2025.score", + "metric_name": "HMMT 2025 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hmmt-2025", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HMMT 2025", + "raw_categories": "[\"math\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "33" + } + }, + "score_details": { + "score": 0.94, + "details": { + "raw_score": "0.94", + "raw_score_field": "score", + "raw_model_id": "glm-5.1", + "raw_benchmark_id": "hmmt-2025", + "source_urls_json": "[\"https://llm-stats.com/models/glm-5.1\",\"https://llm-stats.com/benchmarks/hmmt-2025\",\"https://api.llm-stats.com/leaderboard/benchmarks/hmmt-2025\"]", + "raw_score_id": "hmmt-2025::glm-5.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "hmmt-feb-26::hmmt-feb-26-glm-5.1", + "evaluation_name": "llm_stats.hmmt-feb-26", + "source_data": { + "dataset_name": "HMMT Feb 26", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/glm-5.1", + "https://llm-stats.com/benchmarks/hmmt-feb-26", + "https://api.llm-stats.com/leaderboard/benchmarks/hmmt-feb-26" + ], + "additional_details": { + "raw_benchmark_id": "hmmt-feb-26", + "raw_model_id": "glm-5.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "HMMT February 2026 is a math competition benchmark based on problems from the Harvard-MIT Mathematics Tournament, testing advanced mathematical problem-solving and reasoning.", + "metric_id": "llm_stats.hmmt-feb-26.score", + "metric_name": "HMMT Feb 26 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "hmmt-feb-26", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "HMMT Feb 26", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.826, + "details": { + "raw_score": "0.826", + "raw_score_field": "score", + "raw_model_id": "glm-5.1", + "raw_benchmark_id": "hmmt-feb-26", + "source_urls_json": "[\"https://llm-stats.com/models/glm-5.1\",\"https://llm-stats.com/benchmarks/hmmt-feb-26\",\"https://api.llm-stats.com/leaderboard/benchmarks/hmmt-feb-26\"]", + "raw_score_id": "hmmt-feb-26::glm-5.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "humanity-s-last-exam::humanity-s-last-exam-glm-5.1", + "evaluation_name": "llm_stats.humanity-s-last-exam", + "source_data": { + "dataset_name": "Humanity's Last Exam", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/glm-5.1", + "https://llm-stats.com/benchmarks/humanity-s-last-exam", + "https://api.llm-stats.com/leaderboard/benchmarks/humanity's-last-exam" + ], + "additional_details": { + "raw_benchmark_id": "humanity's-last-exam", + "raw_model_id": "glm-5.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Humanity's Last Exam (HLE) is a multi-modal academic benchmark with 2,500 questions across mathematics, humanities, and natural sciences, designed to test LLM capabilities at the frontier of human knowledge with unambiguous, verifiable solutions", + "metric_id": "llm_stats.humanity-s-last-exam.score", + "metric_name": "Humanity's Last Exam score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "humanity's-last-exam", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Humanity's Last Exam", + "raw_categories": "[\"math\",\"reasoning\",\"vision\"]", + "raw_modality": "multimodal", + "raw_verified": "false", + "raw_model_count": "88" + } + }, + "score_details": { + "score": 0.523, + "details": { + "raw_score": "0.523", + "raw_score_field": "score", + "raw_model_id": "glm-5.1", + "raw_benchmark_id": "humanity's-last-exam", + "source_urls_json": "[\"https://llm-stats.com/models/glm-5.1\",\"https://llm-stats.com/benchmarks/humanity-s-last-exam\",\"https://api.llm-stats.com/leaderboard/benchmarks/humanity's-last-exam\"]", + "raw_score_id": "humanity's-last-exam::glm-5.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "imo-answerbench::imo-answerbench-glm-5.1", + "evaluation_name": "llm_stats.imo-answerbench", + "source_data": { + "dataset_name": "IMO-AnswerBench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/glm-5.1", + "https://llm-stats.com/benchmarks/imo-answerbench", + "https://api.llm-stats.com/leaderboard/benchmarks/imo-answerbench" + ], + "additional_details": { + "raw_benchmark_id": "imo-answerbench", + "raw_model_id": "glm-5.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "IMO-AnswerBench is a benchmark for evaluating mathematical reasoning capabilities on International Mathematical Olympiad (IMO) problems, focusing on answer generation and verification.", + "metric_id": "llm_stats.imo-answerbench.score", + "metric_name": "IMO-AnswerBench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "imo-answerbench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "IMO-AnswerBench", + "raw_categories": "[\"math\",\"reasoning\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "18" + } + }, + "score_details": { + "score": 0.838, + "details": { + "raw_score": "0.838", + "raw_score_field": "score", + "raw_model_id": "glm-5.1", + "raw_benchmark_id": "imo-answerbench", + "source_urls_json": "[\"https://llm-stats.com/models/glm-5.1\",\"https://llm-stats.com/benchmarks/imo-answerbench\",\"https://api.llm-stats.com/leaderboard/benchmarks/imo-answerbench\"]", + "raw_score_id": "imo-answerbench::glm-5.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mcp-atlas::mcp-atlas-glm-5.1", + "evaluation_name": "llm_stats.mcp-atlas", + "source_data": { + "dataset_name": "MCP Atlas", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/glm-5.1", + "https://llm-stats.com/benchmarks/mcp-atlas", + "https://api.llm-stats.com/leaderboard/benchmarks/mcp-atlas" + ], + "additional_details": { + "raw_benchmark_id": "mcp-atlas", + "raw_model_id": "glm-5.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MCP Atlas is a benchmark for evaluating AI models on scaled tool use capabilities, measuring how well models can coordinate and utilize multiple tools across complex multi-step tasks.", + "metric_id": "llm_stats.mcp-atlas.score", + "metric_name": "MCP Atlas score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mcp-atlas", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MCP Atlas", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.718, + "details": { + "raw_score": "0.718", + "raw_score_field": "score", + "raw_model_id": "glm-5.1", + "raw_benchmark_id": "mcp-atlas", + "source_urls_json": "[\"https://llm-stats.com/models/glm-5.1\",\"https://llm-stats.com/benchmarks/mcp-atlas\",\"https://api.llm-stats.com/leaderboard/benchmarks/mcp-atlas\"]", + "raw_score_id": "mcp-atlas::glm-5.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "nl2repo::nl2repo-glm-5.1", + "evaluation_name": "llm_stats.nl2repo", + "source_data": { + "dataset_name": "NL2Repo", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/glm-5.1", + "https://llm-stats.com/benchmarks/nl2repo", + "https://api.llm-stats.com/leaderboard/benchmarks/nl2repo" + ], + "additional_details": { + "raw_benchmark_id": "nl2repo", + "raw_model_id": "glm-5.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "NL2Repo evaluates long-horizon coding capabilities including repository-level understanding, where models must generate or modify code across entire repositories from natural language specifications.", + "metric_id": "llm_stats.nl2repo.score", + "metric_name": "NL2Repo score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "nl2repo", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "NL2Repo", + "raw_categories": "[\"agents\",\"coding\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "11" + } + }, + "score_details": { + "score": 0.427, + "details": { + "raw_score": "0.427", + "raw_score_field": "score", + "raw_model_id": "glm-5.1", + "raw_benchmark_id": "nl2repo", + "source_urls_json": "[\"https://llm-stats.com/models/glm-5.1\",\"https://llm-stats.com/benchmarks/nl2repo\",\"https://api.llm-stats.com/leaderboard/benchmarks/nl2repo\"]", + "raw_score_id": "nl2repo::glm-5.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "swe-bench-pro::swe-bench-pro-glm-5.1", + "evaluation_name": "llm_stats.swe-bench-pro", + "source_data": { + "dataset_name": "SWE-Bench Pro", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/glm-5.1", + "https://llm-stats.com/benchmarks/swe-bench-pro", + "https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-pro" + ], + "additional_details": { + "raw_benchmark_id": "swe-bench-pro", + "raw_model_id": "glm-5.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "SWE-Bench Pro is an advanced version of SWE-Bench that evaluates language models on complex, real-world software engineering tasks requiring extended reasoning and multi-step problem solving.", + "metric_id": "llm_stats.swe-bench-pro.score", + "metric_name": "SWE-Bench Pro score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "swe-bench-pro", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "SWE-Bench Pro", + "raw_categories": "[\"reasoning\",\"agents\",\"code\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "35" + } + }, + "score_details": { + "score": 0.584, + "details": { + "raw_score": "0.584", + "raw_score_field": "score", + "raw_model_id": "glm-5.1", + "raw_benchmark_id": "swe-bench-pro", + "source_urls_json": "[\"https://llm-stats.com/models/glm-5.1\",\"https://llm-stats.com/benchmarks/swe-bench-pro\",\"https://api.llm-stats.com/leaderboard/benchmarks/swe-bench-pro\"]", + "raw_score_id": "swe-bench-pro::glm-5.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "tau3-bench::tau3-bench-glm-5.1", + "evaluation_name": "llm_stats.tau3-bench", + "source_data": { + "dataset_name": "TAU3-Bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/glm-5.1", + "https://llm-stats.com/benchmarks/tau3-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/tau3-bench" + ], + "additional_details": { + "raw_benchmark_id": "tau3-bench", + "raw_model_id": "glm-5.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "TAU3-Bench is a benchmark for evaluating general-purpose agent capabilities, testing models on multi-turn interactions with simulated user models, retrieval, and complex decision-making scenarios.", + "metric_id": "llm_stats.tau3-bench.score", + "metric_name": "TAU3-Bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "tau3-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "TAU3-Bench", + "raw_categories": "[\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "5" + } + }, + "score_details": { + "score": 0.706, + "details": { + "raw_score": "0.706", + "raw_score_field": "score", + "raw_model_id": "glm-5.1", + "raw_benchmark_id": "tau3-bench", + "source_urls_json": "[\"https://llm-stats.com/models/glm-5.1\",\"https://llm-stats.com/benchmarks/tau3-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/tau3-bench\"]", + "raw_score_id": "tau3-bench::glm-5.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2::terminal-bench-2-glm-5.1", + "evaluation_name": "llm_stats.terminal-bench-2", + "source_data": { + "dataset_name": "Terminal-Bench 2.0", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/glm-5.1", + "https://llm-stats.com/benchmarks/terminal-bench-2", + "https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2" + ], + "additional_details": { + "raw_benchmark_id": "terminal-bench-2", + "raw_model_id": "glm-5.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Terminal-Bench 2.0 is an updated benchmark for testing AI agents' tool use ability to operate a computer via terminal. It evaluates how well models can handle real-world, end-to-end tasks autonomously, including compiling code, training models, setting up servers, system administration, security tasks, data science workflows, and cybersecurity vulnerabilities.", + "metric_id": "llm_stats.terminal-bench-2.score", + "metric_name": "Terminal-Bench 2.0 score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "terminal-bench-2", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Terminal-Bench 2.0", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "49" + } + }, + "score_details": { + "score": 0.69, + "details": { + "raw_score": "0.69", + "raw_score_field": "score", + "raw_model_id": "glm-5.1", + "raw_benchmark_id": "terminal-bench-2", + "source_urls_json": "[\"https://llm-stats.com/models/glm-5.1\",\"https://llm-stats.com/benchmarks/terminal-bench-2\",\"https://api.llm-stats.com/leaderboard/benchmarks/terminal-bench-2\"]", + "raw_score_id": "terminal-bench-2::glm-5.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "toolathlon::toolathlon-glm-5.1", + "evaluation_name": "llm_stats.toolathlon", + "source_data": { + "dataset_name": "Toolathlon", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/glm-5.1", + "https://llm-stats.com/benchmarks/toolathlon", + "https://api.llm-stats.com/leaderboard/benchmarks/toolathlon" + ], + "additional_details": { + "raw_benchmark_id": "toolathlon", + "raw_model_id": "glm-5.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Tool Decathlon is a comprehensive benchmark for evaluating AI agents' ability to use multiple tools across diverse task categories. It measures proficiency in tool selection, sequencing, and execution across ten different tool-use scenarios.", + "metric_id": "llm_stats.toolathlon.score", + "metric_name": "Toolathlon score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "toolathlon", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "Toolathlon", + "raw_categories": "[\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "24" + } + }, + "score_details": { + "score": 0.407, + "details": { + "raw_score": "0.407", + "raw_score_field": "score", + "raw_model_id": "glm-5.1", + "raw_benchmark_id": "toolathlon", + "source_urls_json": "[\"https://llm-stats.com/models/glm-5.1\",\"https://llm-stats.com/benchmarks/toolathlon\",\"https://api.llm-stats.com/leaderboard/benchmarks/toolathlon\"]", + "raw_score_id": "toolathlon::glm-5.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "vending-bench-2::vending-bench-2-glm-5.1", + "evaluation_name": "llm_stats.vending-bench-2", + "source_data": { + "dataset_name": "Vending-Bench 2", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/glm-5.1", + "https://llm-stats.com/benchmarks/vending-bench-2", + "https://api.llm-stats.com/leaderboard/benchmarks/vending-bench-2" + ], + "additional_details": { + "raw_benchmark_id": "vending-bench-2", + "raw_model_id": "glm-5.1", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "Vending-Bench 2 tests longer horizon planning capabilities by evaluating how well AI models can manage a simulated vending machine business over extended periods. The benchmark measures a model's ability to maintain consistent tool usage and decision-making for a full simulated year of operation, driving higher returns without drifting off task.", + "metric_id": "llm_stats.vending-bench-2.score", + "metric_name": "Vending-Bench 2 score", + "metric_kind": "benchmark_score", + "metric_unit": "points", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 5634.41, + "additional_details": { + "raw_benchmark_id": "vending-bench-2", + "raw_score_field": "score", + "bound_strategy": "observed_value", + "raw_name": "Vending-Bench 2", + "raw_categories": "[\"reasoning\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "4" + } + }, + "score_details": { + "score": 5634.41, + "details": { + "raw_score": "5634.41", + "raw_score_field": "score", + "raw_model_id": "glm-5.1", + "raw_benchmark_id": "vending-bench-2", + "source_urls_json": "[\"https://llm-stats.com/models/glm-5.1\",\"https://llm-stats.com/benchmarks/vending-bench-2\",\"https://api.llm-stats.com/leaderboard/benchmarks/vending-bench-2\"]", + "raw_score_id": "vending-bench-2::glm-5.1", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/llm_stats/zai-org/glm-5/04df6690-3328-4ce5-936c-975a411e4813.json b/data/llm_stats/zai-org/glm-5/04df6690-3328-4ce5-936c-975a411e4813.json new file mode 100644 index 0000000000000000000000000000000000000000..845bddaf0e03927c9afcb4cbdde2a70b3e3f077b --- /dev/null +++ b/data/llm_stats/zai-org/glm-5/04df6690-3328-4ce5-936c-975a411e4813.json @@ -0,0 +1,204 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "llm-stats/first_party/zai-org_glm-5/1783745564.116323", + "retrieved_timestamp": "1783745564.116323", + "source_metadata": { + "source_name": "LLM Stats API: first_party scores", + "source_type": "documentation", + "source_organization_name": "LLM Stats", + "source_organization_url": "https://llm-stats.com/", + "evaluator_relationship": "first_party", + "additional_details": { + "models_endpoint": "https://api.llm-stats.com/v1/models", + "benchmarks_endpoint": "https://api.llm-stats.com/leaderboard/benchmarks", + "scores_endpoint": "https://api.llm-stats.com/v1/scores", + "scores_endpoint_fallback": "https://api.llm-stats.com/leaderboard/benchmarks/{benchmark_id}", + "developer_page_url": "https://llm-stats.com/developer", + "attribution_url": "https://llm-stats.com/", + "attribution_required": "true", + "source_role": "aggregator" + } + }, + "eval_library": { + "name": "LLM Stats", + "version": "unknown" + }, + "model_info": { + "name": "glm-5", + "id": "zai-org/glm-5", + "developer": "zai-org", + "additional_details": { + "raw_id": "glm-5", + "raw_model_name": "glm-5", + "raw_display_name": "GLM-5", + "raw_organization_id": "zai-org", + "raw_organization_name": "Zhipu AI", + "raw_context_length": "200000", + "raw_input_modalities": "[\"text\"]", + "raw_output_modalities": "[\"text\"]", + "raw_input_price": "1.0", + "raw_output_price": "3.2", + "raw_model_id": "glm-5", + "raw_provider_slug": "zai-org", + "raw_provider_name": "ZeroEval" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "browsecomp::browsecomp-glm-5", + "evaluation_name": "llm_stats.browsecomp", + "source_data": { + "dataset_name": "BrowseComp", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/glm-5", + "https://llm-stats.com/benchmarks/browsecomp", + "https://api.llm-stats.com/leaderboard/benchmarks/browsecomp" + ], + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_model_id": "glm-5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "BrowseComp is a benchmark comprising 1,266 questions that challenge AI agents to persistently navigate the internet in search of hard-to-find, entangled information. The benchmark measures agents' ability to exercise persistence in information gathering, demonstrate creativity in web navigation, and find concise, verifiable answers. Despite the difficulty of the questions, BrowseComp is simple and easy-to-use, as predicted answers are short and easily verifiable against reference answers.", + "metric_id": "llm_stats.browsecomp.score", + "metric_name": "BrowseComp score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "browsecomp", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "BrowseComp", + "raw_categories": "[\"reasoning\",\"search\",\"agents\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "52" + } + }, + "score_details": { + "score": 0.759, + "details": { + "raw_score": "0.759", + "raw_score_field": "score", + "raw_model_id": "glm-5", + "raw_benchmark_id": "browsecomp", + "source_urls_json": "[\"https://llm-stats.com/models/glm-5\",\"https://llm-stats.com/benchmarks/browsecomp\",\"https://api.llm-stats.com/leaderboard/benchmarks/browsecomp\"]", + "raw_score_id": "browsecomp::glm-5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "mcp-atlas::mcp-atlas-glm-5", + "evaluation_name": "llm_stats.mcp-atlas", + "source_data": { + "dataset_name": "MCP Atlas", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/glm-5", + "https://llm-stats.com/benchmarks/mcp-atlas", + "https://api.llm-stats.com/leaderboard/benchmarks/mcp-atlas" + ], + "additional_details": { + "raw_benchmark_id": "mcp-atlas", + "raw_model_id": "glm-5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "MCP Atlas is a benchmark for evaluating AI models on scaled tool use capabilities, measuring how well models can coordinate and utilize multiple tools across complex multi-step tasks.", + "metric_id": "llm_stats.mcp-atlas.score", + "metric_name": "MCP Atlas score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "mcp-atlas", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "MCP Atlas", + "raw_categories": "[\"reasoning\",\"agents\",\"code\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "27" + } + }, + "score_details": { + "score": 0.678, + "details": { + "raw_score": "0.678", + "raw_score_field": "score", + "raw_model_id": "glm-5", + "raw_benchmark_id": "mcp-atlas", + "source_urls_json": "[\"https://llm-stats.com/models/glm-5\",\"https://llm-stats.com/benchmarks/mcp-atlas\",\"https://api.llm-stats.com/leaderboard/benchmarks/mcp-atlas\"]", + "raw_score_id": "mcp-atlas::glm-5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + }, + { + "evaluation_result_id": "t2-bench::t2-bench-glm-5", + "evaluation_name": "llm_stats.t2-bench", + "source_data": { + "dataset_name": "t2-bench", + "source_type": "url", + "url": [ + "https://llm-stats.com/models/glm-5", + "https://llm-stats.com/benchmarks/t2-bench", + "https://api.llm-stats.com/leaderboard/benchmarks/t2-bench" + ], + "additional_details": { + "raw_benchmark_id": "t2-bench", + "raw_model_id": "glm-5", + "source_role": "aggregator" + } + }, + "metric_config": { + "evaluation_description": "t2-bench is a benchmark for evaluating agentic tool use capabilities, measuring how well models can select, sequence, and utilize tools to solve complex tasks. It tests autonomous planning and execution in multi-step scenarios.", + "metric_id": "llm_stats.t2-bench.score", + "metric_name": "t2-bench score", + "metric_kind": "benchmark_score", + "metric_unit": "proportion", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 1.0, + "additional_details": { + "raw_benchmark_id": "t2-bench", + "raw_score_field": "score", + "bound_strategy": "inferred_proportion", + "raw_name": "t2-bench", + "raw_categories": "[\"reasoning\",\"agents\",\"tool_calling\"]", + "raw_modality": "text", + "raw_verified": "false", + "raw_model_count": "23" + } + }, + "score_details": { + "score": 0.897, + "details": { + "raw_score": "0.897", + "raw_score_field": "score", + "raw_model_id": "glm-5", + "raw_benchmark_id": "t2-bench", + "source_urls_json": "[\"https://llm-stats.com/models/glm-5\",\"https://llm-stats.com/benchmarks/t2-bench\",\"https://api.llm-stats.com/leaderboard/benchmarks/t2-bench\"]", + "raw_score_id": "t2-bench::glm-5", + "raw_provenance_label": "unknown", + "raw_verified": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/run_report.json b/data/run_report.json index 3f6ce50054ff57856cb906d64700d9cdd9d6643f..1903d2601149c648498f1b64b5612c75dcf4b8a3 100644 --- a/data/run_report.json +++ b/data/run_report.json @@ -1,31 +1,48 @@ { "adapters": { - "exgentic": { - "error": "^^^^^^^^^^\n File \"/home/runner/work/every_eval_ever/every_eval_ever/.venv/lib/python3.14/site-packages/dill/_dill.py\", line 1217, in save_module_dict\n StockPickler.save_dict(pickler, obj)\n ~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^\n File \"/opt/hostedtoolcache/Python/3.14.6/x64/lib/python3.14/pickle.py\", line 1058, in save_dict\n self._batch_setitems(obj.items(), obj)\n ~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^\nTypeError: Pickler._batch_setitems() takes 2 positional arguments but 3 were given\n", + "cocoabench": { + "error": " [--benchmark-version BENCHMARK_VERSION]\n [--eval-library-version EVAL_LIBRARY_VERSION]\n [--evaluation-timestamp EVALUATION_TIMESTAMP]\n [--public-source-url PUBLIC_SOURCE_URLS]\n [--benchmark-reference-url BENCHMARK_REFERENCE_URLS]\npython3 -m utils.cocoabench.adapter: error: the following arguments are required: --csv\n", "execution_failed": true }, - "helm": { + "hal": { "data_changed": false, "errors": [], "execution_failed": false, "failed_files": 0, - "valid_files": 68 + "valid_files": 246 }, - "livecodebenchpro": { - "error": ", in _run_module_as_main\n File \"\", line 88, in _run_code\n File \"/home/runner/work/every_eval_ever/every_eval_ever/utils/livecodebenchpro/adapter.py\", line 94, in \n main()\n ~~~~^^\n File \"/home/runner/work/every_eval_ever/every_eval_ever/utils/livecodebenchpro/adapter.py\", line 82, in main\n raise FileNotFoundError(f'No JSON files found in {DATA_DIR}')\nFileNotFoundError: No JSON files found in /home/runner/work/every_eval_ever/every_eval_ever/data/livecodebenchpro\n", - "execution_failed": true + "llm_stats": { + "data_changed": true, + "errors": [], + "execution_failed": false, + "failed_files": 0, + "valid_files": 324 }, - "mercor_eval": { - "error": "dule>\n main()\n ~~~~^^\n File \"/home/runner/work/every_eval_ever/every_eval_ever/utils/mercor_eval/adapter.py\", line 674, in main\n resolve_api_key(args.api_key),\n ~~~~~~~~~~~~~~~^^^^^^^^^^^^^^\n File \"/home/runner/work/every_eval_ever/every_eval_ever/utils/mercor_eval/adapter.py\", line 180, in resolve_api_key\n raise ValueError(\n f'Mercor API key required via --api-key or {API_KEY_ENV}.'\n )\nValueError: Mercor API key required via --api-key or MERCOR_EVAL_API_EVALEVAL_KEY.\n", + "mt_bench": { + "data_changed": false, + "errors": [], + "execution_failed": false, + "failed_files": 0, + "valid_files": 34 + }, + "swe_polybench": { + "error": "\n File \"/home/runner/work/every_eval_ever-cron/every_eval_ever-cron/.venv/lib/python3.14/site-packages/dill/_dill.py\", line 1217, in save_module_dict\n StockPickler.save_dict(pickler, obj)\n ~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^\n File \"/opt/hostedtoolcache/Python/3.14.6/x64/lib/python3.14/pickle.py\", line 1058, in save_dict\n self._batch_setitems(obj.items(), obj)\n ~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^\nTypeError: Pickler._batch_setitems() takes 2 positional arguments but 3 were given\n", "execution_failed": true }, - "sciarena": { + "terminal_bench_2": { + "data_changed": false, + "errors": [], + "execution_failed": false, + "failed_files": 0, + "valid_files": 115 + }, + "vals_ai": { "data_changed": true, "errors": [], "execution_failed": false, "failed_files": 0, - "valid_files": 38 + "valid_files": 2118 } }, - "date": "2026-07-10" + "date": "2026-07-11" } \ No newline at end of file diff --git a/data/vals_ai/Qwen/Qwen2.5-72B-Instruct-Turbo/b5c90249-d3da-43b6-8064-2b4f91b5e22d.json b/data/vals_ai/Qwen/Qwen2.5-72B-Instruct-Turbo/b5c90249-d3da-43b6-8064-2b4f91b5e22d.json new file mode 100644 index 0000000000000000000000000000000000000000..49867d2c19342f92662610e50eb3142f0eeb946b --- /dev/null +++ b/data/vals_ai/Qwen/Qwen2.5-72B-Instruct-Turbo/b5c90249-d3da-43b6-8064-2b4f91b5e22d.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/Qwen_Qwen2.5-72B-Instruct-Turbo/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Qwen2.5-72B-Instruct-Turbo", + "id": "Qwen/Qwen2.5-72B-Instruct-Turbo", + "developer": "Qwen", + "additional_details": { + "vals_model_id": "together/Qwen/Qwen2.5-72B-Instruct-Turbo", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:together/Qwen/Qwen2.5-72B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.012, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.012", + "raw_stderr": "1.026", + "latency": "0.513", + "temperature": "0.7", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.026, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:together/Qwen/Qwen2.5-72B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.576, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.576", + "raw_stderr": "0.2", + "latency": "0.592", + "temperature": "0.7", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.2, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:together/Qwen/Qwen2.5-72B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.193, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.193", + "raw_stderr": "0.678", + "latency": "0.638", + "temperature": "0.7", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.678, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:together/Qwen/Qwen2.5-72B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.403, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.403", + "raw_stderr": "0.66", + "latency": "0.831", + "temperature": "0.7", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.66, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:together/Qwen/Qwen2.5-72B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.36, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.36", + "raw_stderr": "1.173", + "latency": "0.545", + "temperature": "0.7", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.173, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:together/Qwen/Qwen2.5-72B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 74.874, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "74.874", + "raw_stderr": "2.819", + "latency": "1.868", + "temperature": "0.7", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.819, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/Qwen/Qwen2.5-7B-Instruct-Turbo/bf72ac51-ad2f-4154-8b0a-5f869d209ca3.json b/data/vals_ai/Qwen/Qwen2.5-7B-Instruct-Turbo/bf72ac51-ad2f-4154-8b0a-5f869d209ca3.json new file mode 100644 index 0000000000000000000000000000000000000000..4367ace7a7a38ee10fa70ad7fbaab97dae850db8 --- /dev/null +++ b/data/vals_ai/Qwen/Qwen2.5-7B-Instruct-Turbo/bf72ac51-ad2f-4154-8b0a-5f869d209ca3.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/Qwen_Qwen2.5-7B-Instruct-Turbo/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Qwen2.5-7B-Instruct-Turbo", + "id": "Qwen/Qwen2.5-7B-Instruct-Turbo", + "developer": "Qwen", + "additional_details": { + "vals_model_id": "together/Qwen/Qwen2.5-7B-Instruct-Turbo", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:together/Qwen/Qwen2.5-7B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 65.607, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "65.607", + "raw_stderr": "1.3", + "latency": "0.339", + "temperature": "0.7", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.3, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:together/Qwen/Qwen2.5-7B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 73.348, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "73.348", + "raw_stderr": "0.248", + "latency": "0.313", + "temperature": "0.7", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.248, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:together/Qwen/Qwen2.5-7B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 72.583, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "72.583", + "raw_stderr": "0.822", + "latency": "0.324", + "temperature": "0.7", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.822, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:together/Qwen/Qwen2.5-7B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 69.559, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "69.559", + "raw_stderr": "0.783", + "latency": "0.395", + "temperature": "0.7", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.783, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:together/Qwen/Qwen2.5-7B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 72.957, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "72.957", + "raw_stderr": "1.303", + "latency": "0.293", + "temperature": "0.7", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.303, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:together/Qwen/Qwen2.5-7B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 63.301, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "63.301", + "raw_stderr": "3.346", + "latency": "0.708", + "temperature": "0.7", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.346, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/ai21labs/jamba-1.5-large/3968f0da-960f-4cbc-9bcf-29a639590f5d.json b/data/vals_ai/ai21labs/jamba-1.5-large/3968f0da-960f-4cbc-9bcf-29a639590f5d.json new file mode 100644 index 0000000000000000000000000000000000000000..3f161c3d860284fd3f5d17a4b8640c283749d5de --- /dev/null +++ b/data/vals_ai/ai21labs/jamba-1.5-large/3968f0da-960f-4cbc-9bcf-29a639590f5d.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/ai21labs_jamba-1.5-large/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "jamba-1.5-large", + "id": "ai21labs/jamba-1.5-large", + "developer": "ai21labs", + "additional_details": { + "vals_model_id": "ai21labs/jamba-1.5-large", + "vals_provider": "AI21 Labs" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:ai21labs/jamba-1.5-large:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 45.455, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "45.455", + "raw_stderr": "1.7", + "latency": "3.138", + "cost_per_test": "0.003786", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.7, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:ai21labs/jamba-1.5-large:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 36.364, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "36.364", + "raw_stderr": "1.642", + "latency": "19.187", + "cost_per_test": "0.304754", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.642, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:ai21labs/jamba-1.5-large:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 39.433, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "39.433", + "raw_stderr": "0.959", + "latency": "10.765", + "cost_per_test": "0.149067", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 0.959, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:ai21labs/jamba-1.5-large:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 36.48, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "36.48", + "raw_stderr": "1.643", + "latency": "9.97", + "cost_per_test": "0.138661", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.643, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/ai21labs/jamba-1.5-large/6b353e8b-f20c-4cee-a1da-eec11d2a37ea.json b/data/vals_ai/ai21labs/jamba-1.5-large/6b353e8b-f20c-4cee-a1da-eec11d2a37ea.json new file mode 100644 index 0000000000000000000000000000000000000000..ac481b82c1fc3f8b03a04ee80ffc132cb715738f --- /dev/null +++ b/data/vals_ai/ai21labs/jamba-1.5-large/6b353e8b-f20c-4cee-a1da-eec11d2a37ea.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/ai21labs_jamba-1.5-large/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "jamba-1.5-large", + "id": "ai21labs/jamba-1.5-large", + "developer": "ai21labs", + "additional_details": { + "vals_model_id": "ai21labs/jamba-1.5-large", + "vals_provider": "AI21 Labs" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:ai21labs/jamba-1.5-large:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.671, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.671", + "raw_stderr": "1.06", + "latency": "0.522", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.06, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:ai21labs/jamba-1.5-large:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 74.878, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "74.878", + "raw_stderr": "0.217", + "latency": "0.693", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 0.217, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:ai21labs/jamba-1.5-large:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 72.783, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "72.783", + "raw_stderr": "0.77", + "latency": "1.011", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 0.77, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:ai21labs/jamba-1.5-large:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 74.16, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "74.16", + "raw_stderr": "0.639", + "latency": "0.808", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 0.639, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:ai21labs/jamba-1.5-large:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 73.026, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "73.026", + "raw_stderr": "1.292", + "latency": "0.714", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.292, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:ai21labs/jamba-1.5-large:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 70.442, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "70.442", + "raw_stderr": "2.605", + "latency": "1.1", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 2.605, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/ai21labs/jamba-1.5-large/c4dbd5d8-147f-48fd-8a5d-f62649a737fd.json b/data/vals_ai/ai21labs/jamba-1.5-large/c4dbd5d8-147f-48fd-8a5d-f62649a737fd.json new file mode 100644 index 0000000000000000000000000000000000000000..9ffabdc6ea9d373016ac32f15e940383b9df61b7 --- /dev/null +++ b/data/vals_ai/ai21labs/jamba-1.5-large/c4dbd5d8-147f-48fd-8a5d-f62649a737fd.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/ai21labs_jamba-1.5-large/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "jamba-1.5-large", + "id": "ai21labs/jamba-1.5-large", + "developer": "ai21labs", + "additional_details": { + "vals_model_id": "ai21labs/jamba-1.5-large", + "vals_provider": "AI21 Labs" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:ai21labs/jamba-1.5-large:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 46.77, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "46.77", + "raw_stderr": "1.427", + "latency": "20.028", + "cost_per_test": "0.005384", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.427, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:ai21labs/jamba-1.5-large:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 58.176, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.176", + "raw_stderr": "0.971", + "latency": "20.028", + "cost_per_test": "0.005384", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 0.971, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:ai21labs/jamba-1.5-large:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 69.583, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "69.583", + "raw_stderr": "1.316", + "latency": "20.028", + "cost_per_test": "0.005384", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.316, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/ai21labs/jamba-1.5-mini/28d058ad-c341-4cec-b0d4-73b182352286.json b/data/vals_ai/ai21labs/jamba-1.5-mini/28d058ad-c341-4cec-b0d4-73b182352286.json new file mode 100644 index 0000000000000000000000000000000000000000..d35966d27f4295003db68225e7fcda40894c1ae3 --- /dev/null +++ b/data/vals_ai/ai21labs/jamba-1.5-mini/28d058ad-c341-4cec-b0d4-73b182352286.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/ai21labs_jamba-1.5-mini/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "jamba-1.5-mini", + "id": "ai21labs/jamba-1.5-mini", + "developer": "ai21labs", + "additional_details": { + "vals_model_id": "ai21labs/jamba-1.5-mini", + "vals_provider": "AI21 Labs" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:ai21labs/jamba-1.5-mini:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 29.763, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "29.763", + "raw_stderr": "1.307", + "latency": "4.78", + "cost_per_test": "0.00028", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.307, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:ai21labs/jamba-1.5-mini:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 41.864, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "41.864", + "raw_stderr": "0.967", + "latency": "4.78", + "cost_per_test": "0.00028", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 0.967, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:ai21labs/jamba-1.5-mini:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 53.966, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "53.966", + "raw_stderr": "1.425", + "latency": "4.78", + "cost_per_test": "0.00028", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.425, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/ai21labs/jamba-1.5-mini/9fcc2a3d-233d-4c99-9056-2b3cae4bed69.json b/data/vals_ai/ai21labs/jamba-1.5-mini/9fcc2a3d-233d-4c99-9056-2b3cae4bed69.json new file mode 100644 index 0000000000000000000000000000000000000000..b3b90eafb141f093c793dd0e4dac86414fd1f140 --- /dev/null +++ b/data/vals_ai/ai21labs/jamba-1.5-mini/9fcc2a3d-233d-4c99-9056-2b3cae4bed69.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/ai21labs_jamba-1.5-mini/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "jamba-1.5-mini", + "id": "ai21labs/jamba-1.5-mini", + "developer": "ai21labs", + "additional_details": { + "vals_model_id": "ai21labs/jamba-1.5-mini", + "vals_provider": "AI21 Labs" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:ai21labs/jamba-1.5-mini:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 39.627, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "39.627", + "raw_stderr": "1.67", + "latency": "0.655", + "cost_per_test": "0.000336", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.67, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:ai21labs/jamba-1.5-mini:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 30.07, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "30.07", + "raw_stderr": "1.566", + "latency": "4.221", + "cost_per_test": "0.030431", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.566, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:ai21labs/jamba-1.5-mini:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 33.877, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "33.877", + "raw_stderr": "0.929", + "latency": "2.315", + "cost_per_test": "0.014865", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 0.929, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:ai21labs/jamba-1.5-mini:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 31.935, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "31.935", + "raw_stderr": "1.592", + "latency": "2.07", + "cost_per_test": "0.013827", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.592, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/ai21labs/jamba-1.5-mini/c5f4f576-845f-462a-8082-1249710d171f.json b/data/vals_ai/ai21labs/jamba-1.5-mini/c5f4f576-845f-462a-8082-1249710d171f.json new file mode 100644 index 0000000000000000000000000000000000000000..efad7b5f39ce059f3a8b419ad8978d3911a27ffd --- /dev/null +++ b/data/vals_ai/ai21labs/jamba-1.5-mini/c5f4f576-845f-462a-8082-1249710d171f.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/ai21labs_jamba-1.5-mini/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "jamba-1.5-mini", + "id": "ai21labs/jamba-1.5-mini", + "developer": "ai21labs", + "additional_details": { + "vals_model_id": "ai21labs/jamba-1.5-mini", + "vals_provider": "AI21 Labs" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:ai21labs/jamba-1.5-mini:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 68.941, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "68.941", + "raw_stderr": "1.138", + "latency": "0.238", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.138, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:ai21labs/jamba-1.5-mini:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 69.181, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "69.181", + "raw_stderr": "0.252", + "latency": "0.249", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 0.252, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:ai21labs/jamba-1.5-mini:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 61.185, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "61.185", + "raw_stderr": "0.912", + "latency": "0.279", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 0.912, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:ai21labs/jamba-1.5-mini:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 66.616, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "66.616", + "raw_stderr": "0.673", + "latency": "0.275", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 0.673, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:ai21labs/jamba-1.5-mini:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 64.682, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "64.682", + "raw_stderr": "1.548", + "latency": "0.253", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.548, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:ai21labs/jamba-1.5-mini:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 69.089, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "69.089", + "raw_stderr": "2.597", + "latency": "0.354", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 2.597, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/ai21labs/jamba-large-1.6/066f1ff6-d533-4b91-9abf-f8bd9f7b3dd2.json b/data/vals_ai/ai21labs/jamba-large-1.6/066f1ff6-d533-4b91-9abf-f8bd9f7b3dd2.json new file mode 100644 index 0000000000000000000000000000000000000000..2d7f39ee1ae5b4bf5281da12f14bbfc0cc7bae3a --- /dev/null +++ b/data/vals_ai/ai21labs/jamba-large-1.6/066f1ff6-d533-4b91-9abf-f8bd9f7b3dd2.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/ai21labs_jamba-large-1.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "jamba-large-1.6", + "id": "ai21labs/jamba-large-1.6", + "developer": "ai21labs", + "additional_details": { + "vals_model_id": "ai21labs/jamba-large-1.6", + "vals_provider": "AI21 Labs" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.13, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.13", + "raw_stderr": "1.692", + "latency": "9.478", + "cost_per_test": "0.004323", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.692, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 47.275, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "47.275", + "raw_stderr": "1.774", + "latency": "9.096", + "cost_per_test": "0.004523", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.774, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 50.265, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "50.265", + "raw_stderr": "1.486", + "latency": "18.791", + "cost_per_test": "0.007081", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.486, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 52.683, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "52.683", + "raw_stderr": "2.466", + "latency": "6.983", + "cost_per_test": "0.004989", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 2.466, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 55.45, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "55.45", + "raw_stderr": "1.713", + "latency": "5.624", + "cost_per_test": "0.004016", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.713, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 31.579, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "31.579", + "raw_stderr": "1.484", + "latency": "19.41", + "cost_per_test": "0.007021", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.484, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 56.846, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "56.846", + "raw_stderr": "1.732", + "latency": "7.652", + "cost_per_test": "0.003323", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.732, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 28.871, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "28.871", + "raw_stderr": "2.322", + "latency": "3.054", + "cost_per_test": "0.004932", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 2.322, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 35.695, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "35.695", + "raw_stderr": "1.443", + "latency": "7.047", + "cost_per_test": "0.004165", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.443, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 58.105, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "58.105", + "raw_stderr": "1.345", + "latency": "16.561", + "cost_per_test": "0.006397", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.345, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 44.048, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "44.048", + "raw_stderr": "1.629", + "latency": "4.044", + "cost_per_test": "0.002706", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.629, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 49.778, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "49.778", + "raw_stderr": "0.475", + "latency": "9.48", + "cost_per_test": "0.004628", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 0.475, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 47.295, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "47.295", + "raw_stderr": "2.235", + "latency": "4.55", + "cost_per_test": "0.003064", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 2.235, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 52.733, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "52.733", + "raw_stderr": "1.387", + "latency": "15.319", + "cost_per_test": "0.005276", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.387, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 64.912, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "64.912", + "raw_stderr": "1.689", + "latency": "5.11", + "cost_per_test": "0.00298", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.689, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/ai21labs/jamba-large-1.6/4df5d47e-e48e-4b11-aee1-ebad8e6ecbaa.json b/data/vals_ai/ai21labs/jamba-large-1.6/4df5d47e-e48e-4b11-aee1-ebad8e6ecbaa.json new file mode 100644 index 0000000000000000000000000000000000000000..553e76d84e242e7f7090e7c0af53289ef3e08177 --- /dev/null +++ b/data/vals_ai/ai21labs/jamba-large-1.6/4df5d47e-e48e-4b11-aee1-ebad8e6ecbaa.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/ai21labs_jamba-large-1.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "jamba-large-1.6", + "id": "ai21labs/jamba-large-1.6", + "developer": "ai21labs", + "additional_details": { + "vals_model_id": "ai21labs/jamba-large-1.6", + "vals_provider": "AI21 Labs" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 47.506, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "47.506", + "raw_stderr": "1.428", + "latency": "16.439", + "cost_per_test": "0.005162", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.428, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 60.875, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.875", + "raw_stderr": "0.949", + "latency": "16.439", + "cost_per_test": "0.005162", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 0.949, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 74.244, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "74.244", + "raw_stderr": "1.25", + "latency": "16.439", + "cost_per_test": "0.005162", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.25, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/ai21labs/jamba-large-1.6/53bbc007-8b3e-4452-97b1-b442b12bf0b6.json b/data/vals_ai/ai21labs/jamba-large-1.6/53bbc007-8b3e-4452-97b1-b442b12bf0b6.json new file mode 100644 index 0000000000000000000000000000000000000000..b2ba46f831b617a9edc024334b8b094e85e153d1 --- /dev/null +++ b/data/vals_ai/ai21labs/jamba-large-1.6/53bbc007-8b3e-4452-97b1-b442b12bf0b6.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/ai21labs_jamba-large-1.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "jamba-large-1.6", + "id": "ai21labs/jamba-large-1.6", + "developer": "ai21labs", + "additional_details": { + "vals_model_id": "ai21labs/jamba-large-1.6", + "vals_provider": "AI21 Labs" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 41.414, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "41.414", + "raw_stderr": "3.363", + "latency": "14.853", + "cost_per_test": "0.008222", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 3.363, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 30.808, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "30.808", + "raw_stderr": "2.513", + "latency": "16.089", + "cost_per_test": "0.005158", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 2.513, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 36.111, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "36.111", + "raw_stderr": "2.099", + "latency": "15.471", + "cost_per_test": "0.00669", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 2.099, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/ai21labs/jamba-large-1.6/681e5efe-60b5-43f6-a255-9cf84c90eb34.json b/data/vals_ai/ai21labs/jamba-large-1.6/681e5efe-60b5-43f6-a255-9cf84c90eb34.json new file mode 100644 index 0000000000000000000000000000000000000000..d04f7cc6d5f7a8d1d4da086d44787967e5dbcc2c --- /dev/null +++ b/data/vals_ai/ai21labs/jamba-large-1.6/681e5efe-60b5-43f6-a255-9cf84c90eb34.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/ai21labs_jamba-large-1.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "jamba-large-1.6", + "id": "ai21labs/jamba-large-1.6", + "developer": "ai21labs", + "additional_details": { + "vals_model_id": "ai21labs/jamba-large-1.6", + "vals_provider": "AI21 Labs" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 57.453, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "57.453", + "raw_stderr": "2.755", + "latency": "3.338", + "cost_per_test": "0.002158", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 2.755, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 1.429, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "1.429", + "raw_stderr": "0.634", + "latency": "6.075", + "cost_per_test": "0.004078", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 0.634, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 8.094, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "8.094", + "raw_stderr": "1.394", + "latency": "5.106", + "cost_per_test": "0.003471", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.394, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 22.325, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "22.325", + "raw_stderr": "1.051", + "latency": "4.84", + "cost_per_test": "0.003236", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.051, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/ai21labs/jamba-large-1.6/da6a71e3-8b7e-47b4-8f9e-91bd6a7ff898.json b/data/vals_ai/ai21labs/jamba-large-1.6/da6a71e3-8b7e-47b4-8f9e-91bd6a7ff898.json new file mode 100644 index 0000000000000000000000000000000000000000..6c9f54c49e45e1cea45ba79ce8250f40ac0b351b --- /dev/null +++ b/data/vals_ai/ai21labs/jamba-large-1.6/da6a71e3-8b7e-47b4-8f9e-91bd6a7ff898.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/ai21labs_jamba-large-1.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "jamba-large-1.6", + "id": "ai21labs/jamba-large-1.6", + "developer": "ai21labs", + "additional_details": { + "vals_model_id": "ai21labs/jamba-large-1.6", + "vals_provider": "AI21 Labs" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 50.117, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "50.117", + "raw_stderr": "1.707", + "latency": "7.516", + "cost_per_test": "0.0044", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.707, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 35.431, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "35.431", + "raw_stderr": "1.633", + "latency": "40.708", + "cost_per_test": "0.306016", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.633, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 41.531, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "41.531", + "raw_stderr": "0.963", + "latency": "27.38", + "cost_per_test": "0.150029", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 0.963, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 39.044, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "39.044", + "raw_stderr": "1.665", + "latency": "33.916", + "cost_per_test": "0.139671", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.665, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/ai21labs/jamba-large-1.6/e2eb02bd-5ee7-476b-a39d-0edf4ecd2c64.json b/data/vals_ai/ai21labs/jamba-large-1.6/e2eb02bd-5ee7-476b-a39d-0edf4ecd2c64.json new file mode 100644 index 0000000000000000000000000000000000000000..f0aa79caf67f76fcb1d2f211519809a660a19c3c --- /dev/null +++ b/data/vals_ai/ai21labs/jamba-large-1.6/e2eb02bd-5ee7-476b-a39d-0edf4ecd2c64.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/ai21labs_jamba-large-1.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "jamba-large-1.6", + "id": "ai21labs/jamba-large-1.6", + "developer": "ai21labs", + "additional_details": { + "vals_model_id": "ai21labs/jamba-large-1.6", + "vals_provider": "AI21 Labs" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.789, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.789", + "raw_stderr": "0.954", + "latency": "0.497", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 0.954, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 75.377, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "75.377", + "raw_stderr": "0.212", + "latency": "0.613", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 0.212, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 64.335, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "64.335", + "raw_stderr": "0.773", + "latency": "0.851", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 0.773, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 71.962, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "71.962", + "raw_stderr": "0.424", + "latency": "0.814", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 0.424, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 74.37, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "74.37", + "raw_stderr": "1.27", + "latency": "0.567", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.27, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:ai21labs/jamba-large-1.6:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 67.937, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "67.937", + "raw_stderr": "1.157", + "latency": "1.542", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.157, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/ai21labs/jamba-mini-1.6/4fea0227-a9d9-4285-a103-b2a0c79d515c.json b/data/vals_ai/ai21labs/jamba-mini-1.6/4fea0227-a9d9-4285-a103-b2a0c79d515c.json new file mode 100644 index 0000000000000000000000000000000000000000..eb6f79d132967aa5e1946566baafaff0e0d39ff7 --- /dev/null +++ b/data/vals_ai/ai21labs/jamba-mini-1.6/4fea0227-a9d9-4285-a103-b2a0c79d515c.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/ai21labs_jamba-mini-1.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "jamba-mini-1.6", + "id": "ai21labs/jamba-mini-1.6", + "developer": "ai21labs", + "additional_details": { + "vals_model_id": "ai21labs/jamba-mini-1.6", + "vals_provider": "AI21 Labs" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 25.941, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "25.941", + "raw_stderr": "1.637", + "latency": "2.025", + "cost_per_test": "0.000368", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.637, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 30.418, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "30.418", + "raw_stderr": "1.634", + "latency": "2.722", + "cost_per_test": "0.000353", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.634, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 22.085, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "22.085", + "raw_stderr": "1.231", + "latency": "4.486", + "cost_per_test": "0.000561", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.231, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 28.293, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "28.293", + "raw_stderr": "2.224", + "latency": "2.478", + "cost_per_test": "0.00044", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 2.224, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 29.621, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "29.621", + "raw_stderr": "1.57", + "latency": "1.406", + "cost_per_test": "0.00032", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.57, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 17.441, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "17.441", + "raw_stderr": "1.213", + "latency": "5.321", + "cost_per_test": "0.000545", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.213, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 43.399, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "43.399", + "raw_stderr": "1.733", + "latency": "1.08", + "cost_per_test": "0.000264", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.733, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 34.121, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "34.121", + "raw_stderr": "2.429", + "latency": "1.15", + "cost_per_test": "0.000471", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 2.429, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 27.157, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "27.157", + "raw_stderr": "1.34", + "latency": "1.471", + "cost_per_test": "0.000355", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.34, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 30.792, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "30.792", + "raw_stderr": "1.256", + "latency": "4.284", + "cost_per_test": "0.000583", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.256, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 41.991, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "41.991", + "raw_stderr": "1.624", + "latency": "1.232", + "cost_per_test": "0.00023", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.624, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 30.278, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "30.278", + "raw_stderr": "0.448", + "latency": "2.426", + "cost_per_test": "0.000391", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 0.448, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 35.671, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "35.671", + "raw_stderr": "2.144", + "latency": "1.912", + "cost_per_test": "0.000301", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 2.144, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 30.023, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "30.023", + "raw_stderr": "1.27", + "latency": "3.412", + "cost_per_test": "0.000426", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.27, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 26.942, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "26.942", + "raw_stderr": "1.571", + "latency": "0.978", + "cost_per_test": "0.000259", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.571, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/ai21labs/jamba-mini-1.6/772ceec7-b38e-4ea9-8762-a7436f594a13.json b/data/vals_ai/ai21labs/jamba-mini-1.6/772ceec7-b38e-4ea9-8762-a7436f594a13.json new file mode 100644 index 0000000000000000000000000000000000000000..298aada3051739a9b0dee7248c6f1fe01d207cf8 --- /dev/null +++ b/data/vals_ai/ai21labs/jamba-mini-1.6/772ceec7-b38e-4ea9-8762-a7436f594a13.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/ai21labs_jamba-mini-1.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "jamba-mini-1.6", + "id": "ai21labs/jamba-mini-1.6", + "developer": "ai21labs", + "additional_details": { + "vals_model_id": "ai21labs/jamba-mini-1.6", + "vals_provider": "AI21 Labs" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 44.988, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "44.988", + "raw_stderr": "1.698", + "latency": "1.553", + "cost_per_test": "0.000358", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.698, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 32.051, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "32.051", + "raw_stderr": "1.593", + "latency": "7.8", + "cost_per_test": "0.03048", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.593, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 38.034, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "38.034", + "raw_stderr": "0.951", + "latency": "4.248", + "cost_per_test": "0.014897", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 0.951, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 37.063, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "37.063", + "raw_stderr": "1.649", + "latency": "3.391", + "cost_per_test": "0.013854", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.649, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/ai21labs/jamba-mini-1.6/9473efa2-9ef7-4c33-9a5b-e273e912f27e.json b/data/vals_ai/ai21labs/jamba-mini-1.6/9473efa2-9ef7-4c33-9a5b-e273e912f27e.json new file mode 100644 index 0000000000000000000000000000000000000000..9d2dcb2ca7f63a0570003cb51e00d3b0e8b2e062 --- /dev/null +++ b/data/vals_ai/ai21labs/jamba-mini-1.6/9473efa2-9ef7-4c33-9a5b-e273e912f27e.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/ai21labs_jamba-mini-1.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "jamba-mini-1.6", + "id": "ai21labs/jamba-mini-1.6", + "developer": "ai21labs", + "additional_details": { + "vals_model_id": "ai21labs/jamba-mini-1.6", + "vals_provider": "AI21 Labs" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 30.335, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "30.335", + "raw_stderr": "1.315", + "latency": "4.337", + "cost_per_test": "0.00026", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.315, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 44.604, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "44.604", + "raw_stderr": "0.963", + "latency": "4.337", + "cost_per_test": "0.00026", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 0.963, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 58.872, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.872", + "raw_stderr": "1.407", + "latency": "4.337", + "cost_per_test": "0.00026", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.407, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/ai21labs/jamba-mini-1.6/9e76335d-f593-4eb2-84ad-3ad8854cee00.json b/data/vals_ai/ai21labs/jamba-mini-1.6/9e76335d-f593-4eb2-84ad-3ad8854cee00.json new file mode 100644 index 0000000000000000000000000000000000000000..1e49a9ef3bdfe0e0064afc1156cbe4ac57d48a43 --- /dev/null +++ b/data/vals_ai/ai21labs/jamba-mini-1.6/9e76335d-f593-4eb2-84ad-3ad8854cee00.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/ai21labs_jamba-mini-1.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "jamba-mini-1.6", + "id": "ai21labs/jamba-mini-1.6", + "developer": "ai21labs", + "additional_details": { + "vals_model_id": "ai21labs/jamba-mini-1.6", + "vals_provider": "AI21 Labs" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 73.114, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "73.114", + "raw_stderr": "1.122", + "latency": "0.248", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.122, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 69.282, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "69.282", + "raw_stderr": "0.247", + "latency": "0.258", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 0.247, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 71.986, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "71.986", + "raw_stderr": "0.81", + "latency": "0.279", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 0.81, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 69.726, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "69.726", + "raw_stderr": "0.515", + "latency": "0.305", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 0.515, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 65.935, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "65.935", + "raw_stderr": "1.471", + "latency": "0.259", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.471, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 68.312, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "68.312", + "raw_stderr": "1.581", + "latency": "0.481", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.581, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/ai21labs/jamba-mini-1.6/e3f6f1ca-88c3-45eb-8243-daffc3c3deeb.json b/data/vals_ai/ai21labs/jamba-mini-1.6/e3f6f1ca-88c3-45eb-8243-daffc3c3deeb.json new file mode 100644 index 0000000000000000000000000000000000000000..f353eb571c069e454fd8c53c6afe52117d6e8ac1 --- /dev/null +++ b/data/vals_ai/ai21labs/jamba-mini-1.6/e3f6f1ca-88c3-45eb-8243-daffc3c3deeb.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/ai21labs_jamba-mini-1.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "jamba-mini-1.6", + "id": "ai21labs/jamba-mini-1.6", + "developer": "ai21labs", + "additional_details": { + "vals_model_id": "ai21labs/jamba-mini-1.6", + "vals_provider": "AI21 Labs" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 27.64, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "27.64", + "raw_stderr": "2.492", + "latency": "0.814", + "cost_per_test": "0.000151", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 2.492, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 0.286, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0.286", + "raw_stderr": "0.285", + "latency": "1.298", + "cost_per_test": "0.000239", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 0.285, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 1.828, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "1.828", + "raw_stderr": "0.684", + "latency": "1.304", + "cost_per_test": "0.000207", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 0.684, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 9.918, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "9.918", + "raw_stderr": "0.867", + "latency": "1.139", + "cost_per_test": "0.000199", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 0.867, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/ai21labs/jamba-mini-1.6/fa4ed12d-b412-42c2-af0b-28500266fd3e.json b/data/vals_ai/ai21labs/jamba-mini-1.6/fa4ed12d-b412-42c2-af0b-28500266fd3e.json new file mode 100644 index 0000000000000000000000000000000000000000..fcc23a848b1f00e8e5d384d6de90145ab7c4a148 --- /dev/null +++ b/data/vals_ai/ai21labs/jamba-mini-1.6/fa4ed12d-b412-42c2-af0b-28500266fd3e.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/ai21labs_jamba-mini-1.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "jamba-mini-1.6", + "id": "ai21labs/jamba-mini-1.6", + "developer": "ai21labs", + "additional_details": { + "vals_model_id": "ai21labs/jamba-mini-1.6", + "vals_provider": "AI21 Labs" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "33.333", + "raw_stderr": "3.201", + "latency": "4.791", + "cost_per_test": "0.000652", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 3.201, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 30.808, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "30.808", + "raw_stderr": "2.36", + "latency": "5.115", + "cost_per_test": "0.000322", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 2.36, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:ai21labs/jamba-mini-1.6:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 32.07, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "32.07", + "raw_stderr": "1.988", + "latency": "4.953", + "cost_per_test": "0.000487", + "temperature": "0.4", + "max_output_tokens": "4096", + "provider": "AI21 Labs" + }, + "uncertainty": { + "standard_error": { + "value": 1.988, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.4, + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3-max-2026-01-23/0a8a8495-120d-4389-8295-f8766f71ea30.json b/data/vals_ai/alibaba/qwen3-max-2026-01-23/0a8a8495-120d-4389-8295-f8766f71ea30.json new file mode 100644 index 0000000000000000000000000000000000000000..88677e32b5407a8256cd91620ad2ba22111e3d58 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3-max-2026-01-23/0a8a8495-120d-4389-8295-f8766f71ea30.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/alibaba_qwen3-max-2026-01-23/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3-max-2026-01-23", + "id": "alibaba/qwen3-max-2026-01-23", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3-max-2026-01-23", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:alibaba/qwen3-max-2026-01-23:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 72.709, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "72.709", + "raw_stderr": "1.905", + "latency": "362.378", + "cost_per_test": "0.085327", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.905, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3-max-2026-01-23/2d408fef-9bb3-4087-a11c-87c3c64933b8.json b/data/vals_ai/alibaba/qwen3-max-2026-01-23/2d408fef-9bb3-4087-a11c-87c3c64933b8.json new file mode 100644 index 0000000000000000000000000000000000000000..1e952e45fb37eed812cb2fc31d44d423fddb0eb4 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3-max-2026-01-23/2d408fef-9bb3-4087-a11c-87c3c64933b8.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/alibaba_qwen3-max-2026-01-23/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3-max-2026-01-23", + "id": "alibaba/qwen3-max-2026-01-23", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3-max-2026-01-23", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:alibaba/qwen3-max-2026-01-23:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 12.333, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "12.333", + "raw_stderr": "5.124", + "latency": "3278.996", + "cost_per_test": "1.949747", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 5.124, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:alibaba/qwen3-max-2026-01-23:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 15.167, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "15.167", + "raw_stderr": "4.907", + "latency": "2627.47", + "cost_per_test": "1.853145", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 4.907, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:alibaba/qwen3-max-2026-01-23:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 13.75, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "13.75", + "raw_stderr": "3.547", + "latency": "2953.233", + "cost_per_test": "1.901446", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 3.547, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3-max-2026-01-23/6ba63c89-6fc2-4616-adc6-a2c447025b82.json b/data/vals_ai/alibaba/qwen3-max-2026-01-23/6ba63c89-6fc2-4616-adc6-a2c447025b82.json new file mode 100644 index 0000000000000000000000000000000000000000..0938b9bd96dd8361e49c286ad2275eed8ca83db1 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3-max-2026-01-23/6ba63c89-6fc2-4616-adc6-a2c447025b82.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/alibaba_qwen3-max-2026-01-23/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3-max-2026-01-23", + "id": "alibaba/qwen3-max-2026-01-23", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3-max-2026-01-23", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:alibaba/qwen3-max-2026-01-23:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.353, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.353", + "raw_stderr": "1.05", + "latency": "88.539", + "cost_per_test": "0.020461", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.05, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:alibaba/qwen3-max-2026-01-23:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.001, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.001", + "raw_stderr": "1.019", + "latency": "102.176", + "cost_per_test": "0.024141", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.019, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:alibaba/qwen3-max-2026-01-23:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.989, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.989", + "raw_stderr": "0.851", + "latency": "133.733", + "cost_per_test": "0.034771", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.851, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:alibaba/qwen3-max-2026-01-23:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.537, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.537", + "raw_stderr": "1.573", + "latency": "104.941", + "cost_per_test": "0.024617", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.573, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:alibaba/qwen3-max-2026-01-23:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.796, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.796", + "raw_stderr": "1.127", + "latency": "83.049", + "cost_per_test": "0.019045", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.127, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:alibaba/qwen3-max-2026-01-23:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.631, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.631", + "raw_stderr": "1.244", + "latency": "238.952", + "cost_per_test": "0.065445", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.244, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:alibaba/qwen3-max-2026-01-23:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.729, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.729", + "raw_stderr": "1.431", + "latency": "99.036", + "cost_per_test": "0.022536", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.431, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:alibaba/qwen3-max-2026-01-23:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.803, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.803", + "raw_stderr": "2.224", + "latency": "89.177", + "cost_per_test": "0.02017", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.224, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:alibaba/qwen3-max-2026-01-23:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 70.481, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.481", + "raw_stderr": "1.375", + "latency": "141.377", + "cost_per_test": "0.032196", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.375, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:alibaba/qwen3-max-2026-01-23:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 95.559, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "95.559", + "raw_stderr": "0.56", + "latency": "89.333", + "cost_per_test": "0.024008", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.56, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:alibaba/qwen3-max-2026-01-23:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.035, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.035", + "raw_stderr": "1.263", + "latency": "80.23", + "cost_per_test": "0.017908", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.263, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:alibaba/qwen3-max-2026-01-23:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.976, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.976", + "raw_stderr": "0.352", + "latency": "110.207", + "cost_per_test": "0.026557", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.352, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:alibaba/qwen3-max-2026-01-23:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.563, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.563", + "raw_stderr": "1.736", + "latency": "79.18", + "cost_per_test": "0.017179", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.736, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:alibaba/qwen3-max-2026-01-23:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.224, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.224", + "raw_stderr": "0.785", + "latency": "136.647", + "cost_per_test": "0.032391", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.785, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:alibaba/qwen3-max-2026-01-23:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.96, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.96", + "raw_stderr": "1.299", + "latency": "76.522", + "cost_per_test": "0.016928", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.299, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3-max-2026-01-23/aec9db52-e01a-4eaa-8aa8-7a15cab65cf4.json b/data/vals_ai/alibaba/qwen3-max-2026-01-23/aec9db52-e01a-4eaa-8aa8-7a15cab65cf4.json new file mode 100644 index 0000000000000000000000000000000000000000..b567e81cab53afea3120d5f7f922baae578ae6d8 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3-max-2026-01-23/aec9db52-e01a-4eaa-8aa8-7a15cab65cf4.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/alibaba_qwen3-max-2026-01-23/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3-max-2026-01-23", + "id": "alibaba/qwen3-max-2026-01-23", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3-max-2026-01-23", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:alibaba/qwen3-max-2026-01-23:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 64.918, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.918", + "raw_stderr": "1.629", + "latency": "76.495", + "cost_per_test": "0.018879", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.629, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:alibaba/qwen3-max-2026-01-23:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 70.629, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "70.629", + "raw_stderr": "1.555", + "latency": "198.506", + "cost_per_test": "0.565658", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.555, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:alibaba/qwen3-max-2026-01-23:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 68.026, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.026", + "raw_stderr": "0.918", + "latency": "121.317", + "cost_per_test": "0.261088", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.918, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:alibaba/qwen3-max-2026-01-23:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 68.531, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.531", + "raw_stderr": "1.585", + "latency": "88.951", + "cost_per_test": "0.198727", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.585, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3-max-2026-01-23/ce34094e-92be-4343-acef-5ac9ec11d8fb.json b/data/vals_ai/alibaba/qwen3-max-2026-01-23/ce34094e-92be-4343-acef-5ac9ec11d8fb.json new file mode 100644 index 0000000000000000000000000000000000000000..36e90bd444a9cf9fe5a4f69c11093271bbf6ae8e --- /dev/null +++ b/data/vals_ai/alibaba/qwen3-max-2026-01-23/ce34094e-92be-4343-acef-5ac9ec11d8fb.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/alibaba_qwen3-max-2026-01-23/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3-max-2026-01-23", + "id": "alibaba/qwen3-max-2026-01-23", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3-max-2026-01-23", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:alibaba/qwen3-max-2026-01-23:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 82.828, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.828", + "raw_stderr": "2.68", + "latency": "351.668", + "cost_per_test": "0.07229", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.68, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:alibaba/qwen3-max-2026-01-23:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 86.869, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.869", + "raw_stderr": "2.4", + "latency": "384.478", + "cost_per_test": "0.075911", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.4, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:alibaba/qwen3-max-2026-01-23:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 84.848, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.848", + "raw_stderr": "1.799", + "latency": "368.073", + "cost_per_test": "0.0741", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.799, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3-max-preview/190ead32-3bcf-4709-bbed-183b81ed38e3.json b/data/vals_ai/alibaba/qwen3-max-preview/190ead32-3bcf-4709-bbed-183b81ed38e3.json new file mode 100644 index 0000000000000000000000000000000000000000..07481ad1a4a4241a763182fe6bba06e68364f8df --- /dev/null +++ b/data/vals_ai/alibaba/qwen3-max-preview/190ead32-3bcf-4709-bbed-183b81ed38e3.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/alibaba_qwen3-max-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3-max-preview", + "id": "alibaba/qwen3-max-preview", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3-max-preview", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.447, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.447", + "raw_stderr": "0.689", + "latency": "26.702", + "cost_per_test": "0.003024", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.689, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 32.571, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "32.571", + "raw_stderr": "2.505", + "latency": "172.001", + "cost_per_test": "0.025552", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.505, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 69.713, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "69.713", + "raw_stderr": "2.348", + "latency": "90.038", + "cost_per_test": "0.012013", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.348, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 66.91, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66.91", + "raw_stderr": "1.167", + "latency": "96.247", + "cost_per_test": "0.01353", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.167, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3-max-preview/1ea8ea4e-64a4-44a5-82c5-977ccbe33ac4.json b/data/vals_ai/alibaba/qwen3-max-preview/1ea8ea4e-64a4-44a5-82c5-977ccbe33ac4.json new file mode 100644 index 0000000000000000000000000000000000000000..c00717b3f06f908a9b0c0805ec14d55f9688abc8 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3-max-preview/1ea8ea4e-64a4-44a5-82c5-977ccbe33ac4.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/alibaba_qwen3-max-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3-max-preview", + "id": "alibaba/qwen3-max-preview", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3-max-preview", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 11.667, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "11.667", + "raw_stderr": "5.728", + "latency": "1362.606", + "cost_per_test": "2.276975", + "temperature": "0.7", + "max_output_tokens": "32768", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 5.728, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 3.833, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "3.833", + "raw_stderr": "2.692", + "latency": "5381.491", + "cost_per_test": "0.521296", + "temperature": "0.7", + "max_output_tokens": "32768", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.692, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 7.75, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "7.75", + "raw_stderr": "3.165", + "latency": "3372.048", + "cost_per_test": "1.399136", + "temperature": "0.7", + "max_output_tokens": "32768", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 3.165, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3-max-preview/572cd247-9853-4176-9ea5-5a541b576168.json b/data/vals_ai/alibaba/qwen3-max-preview/572cd247-9853-4176-9ea5-5a541b576168.json new file mode 100644 index 0000000000000000000000000000000000000000..81a04809ca9aa5c107662e1b56ff74e354849c67 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3-max-preview/572cd247-9853-4176-9ea5-5a541b576168.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/alibaba_qwen3-max-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3-max-preview", + "id": "alibaba/qwen3-max-preview", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3-max-preview", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 59.907, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.907", + "raw_stderr": "1.673", + "latency": "5.561", + "cost_per_test": "0.003066", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.673, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 44.056, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "44.056", + "raw_stderr": "1.695", + "latency": "441.206", + "cost_per_test": "0.205935", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.695, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 52.953, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.953", + "raw_stderr": "0.975", + "latency": "227.451", + "cost_per_test": "0.09654", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.975, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 54.895, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "54.895", + "raw_stderr": "1.699", + "latency": "235.587", + "cost_per_test": "0.080618", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.699, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3-max-preview/85fdaeb8-e5db-4574-b875-6287a52e899b.json b/data/vals_ai/alibaba/qwen3-max-preview/85fdaeb8-e5db-4574-b875-6287a52e899b.json new file mode 100644 index 0000000000000000000000000000000000000000..c56b5ba62111eb02a5f8f36130fc09028924640d --- /dev/null +++ b/data/vals_ai/alibaba/qwen3-max-preview/85fdaeb8-e5db-4574-b875-6287a52e899b.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/alibaba_qwen3-max-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3-max-preview", + "id": "alibaba/qwen3-max-preview", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3-max-preview", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.329, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.329", + "raw_stderr": "0.994", + "latency": "35.357", + "cost_per_test": "0.005165", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.994, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.593, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.593", + "raw_stderr": "1.132", + "latency": "45.611", + "cost_per_test": "0.005816", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.132, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.929, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.929", + "raw_stderr": "0.894", + "latency": "48.823", + "cost_per_test": "0.008645", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.894, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.073, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.073", + "raw_stderr": "1.657", + "latency": "44.122", + "cost_per_test": "0.006141", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.657, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.322, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.322", + "raw_stderr": "1.145", + "latency": "32.749", + "cost_per_test": "0.004664", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.145, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.987, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.987", + "raw_stderr": "1.352", + "latency": "104.441", + "cost_per_test": "0.013442", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.352, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.851, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.851", + "raw_stderr": "1.428", + "latency": "36.405", + "cost_per_test": "0.00458", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.428, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.491, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.491", + "raw_stderr": "2.261", + "latency": "30.172", + "cost_per_test": "0.005582", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.261, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 68.029, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "68.029", + "raw_stderr": "1.406", + "latency": "44.386", + "cost_per_test": "0.006265", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.406, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.301, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.301", + "raw_stderr": "0.631", + "latency": "44.587", + "cost_per_test": "0.007185", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.631, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.844, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.844", + "raw_stderr": "1.295", + "latency": "28.347", + "cost_per_test": "0.003492", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.295, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.537, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.537", + "raw_stderr": "0.364", + "latency": "43.428", + "cost_per_test": "0.006194", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.364, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.557, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.557", + "raw_stderr": "1.837", + "latency": "34.792", + "cost_per_test": "0.004204", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.837, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.376, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.376", + "raw_stderr": "0.889", + "latency": "53.018", + "cost_per_test": "0.00767", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.889, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.837, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.837", + "raw_stderr": "1.27", + "latency": "25.18", + "cost_per_test": "0.003869", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.27, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3-max-preview/cd2f3ed3-bdb9-4e36-976a-255e4a4d8c62.json b/data/vals_ai/alibaba/qwen3-max-preview/cd2f3ed3-bdb9-4e36-976a-255e4a4d8c62.json new file mode 100644 index 0000000000000000000000000000000000000000..0810eb2db3e7ea143e54ebc5a6508f840b88cc49 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3-max-preview/cd2f3ed3-bdb9-4e36-976a-255e4a4d8c62.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/alibaba_qwen3-max-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3-max-preview", + "id": "alibaba/qwen3-max-preview", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3-max-preview", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 62.388, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.388", + "raw_stderr": "1.385", + "latency": "21.405", + "cost_per_test": "0.00671", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.385, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 73.958, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "73.958", + "raw_stderr": "0.856", + "latency": "21.405", + "cost_per_test": "0.00671", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.856, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 85.527, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "85.527", + "raw_stderr": "1.006", + "latency": "21.405", + "cost_per_test": "0.00671", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.006, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3-max-preview/dccdc4fa-4e6a-4a21-aafb-b28832a15cb7.json b/data/vals_ai/alibaba/qwen3-max-preview/dccdc4fa-4e6a-4a21-aafb-b28832a15cb7.json new file mode 100644 index 0000000000000000000000000000000000000000..57813a5e85670a78f1a2d01e3ca9d44580716a1f --- /dev/null +++ b/data/vals_ai/alibaba/qwen3-max-preview/dccdc4fa-4e6a-4a21-aafb-b28832a15cb7.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/alibaba_qwen3-max-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3-max-preview", + "id": "alibaba/qwen3-max-preview", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3-max-preview", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 84.999, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.999", + "raw_stderr": "0.861", + "latency": "1.451", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.861, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.058, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.058", + "raw_stderr": "0.205", + "latency": "1.54", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.205, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.509, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.509", + "raw_stderr": "0.662", + "latency": "1.558", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.662, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.333, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.333", + "raw_stderr": "0.397", + "latency": "1.886", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.397, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.229, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.229", + "raw_stderr": "0.83", + "latency": "1.569", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.83, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.87, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.87", + "raw_stderr": "1.425", + "latency": "3.31", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.425, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3-max-preview/e51d668a-d307-420a-8bdb-36260190f88c.json b/data/vals_ai/alibaba/qwen3-max-preview/e51d668a-d307-420a-8bdb-36260190f88c.json new file mode 100644 index 0000000000000000000000000000000000000000..79c843a66bda3cebd94a15ba9aa253ed4282e931 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3-max-preview/e51d668a-d307-420a-8bdb-36260190f88c.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/alibaba_qwen3-max-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3-max-preview", + "id": "alibaba/qwen3-max-preview", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3-max-preview", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 77.778, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.778", + "raw_stderr": "2.955", + "latency": "123.747", + "cost_per_test": "0.014168", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.955, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 77.778, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.778", + "raw_stderr": "2.955", + "latency": "89.82", + "cost_per_test": "0.012593", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.955, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:alibaba/qwen3-max-preview:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 77.778, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.778", + "raw_stderr": "2.09", + "latency": "106.784", + "cost_per_test": "0.01338", + "temperature": "0.7", + "max_output_tokens": "32000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.09, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3-max/1a92b810-dc1c-4d81-9af4-42f478b6cbad.json b/data/vals_ai/alibaba/qwen3-max/1a92b810-dc1c-4d81-9af4-42f478b6cbad.json new file mode 100644 index 0000000000000000000000000000000000000000..bbdf644e92ca9c73195bfbeb8d480bb338fb7d6d --- /dev/null +++ b/data/vals_ai/alibaba/qwen3-max/1a92b810-dc1c-4d81-9af4-42f478b6cbad.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/alibaba_qwen3-max/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3-max", + "id": "alibaba/qwen3-max", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3-max", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 3.506, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "3.506", + "raw_stderr": "1.292", + "latency": "3465.17", + "cost_per_test": "6.023105", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.292, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3-max/348ae29f-191b-4297-a10d-51d8f3d3fbd5.json b/data/vals_ai/alibaba/qwen3-max/348ae29f-191b-4297-a10d-51d8f3d3fbd5.json new file mode 100644 index 0000000000000000000000000000000000000000..3dd5fd51f8b13bbb11b4e9cc887e88bdb4a1bd53 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3-max/348ae29f-191b-4297-a10d-51d8f3d3fbd5.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/alibaba_qwen3-max/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3-max", + "id": "alibaba/qwen3-max", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3-max", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 61.733, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.733", + "raw_stderr": "1.39", + "latency": "29.296", + "cost_per_test": "0.00654", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.39, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 73.508, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "73.508", + "raw_stderr": "0.86", + "latency": "29.296", + "cost_per_test": "0.00654", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.86, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 85.282, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "85.282", + "raw_stderr": "1.013", + "latency": "29.296", + "cost_per_test": "0.00654", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.013, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3-max/4a1afd14-4494-434e-a47f-5198d481f3b7.json b/data/vals_ai/alibaba/qwen3-max/4a1afd14-4494-434e-a47f-5198d481f3b7.json new file mode 100644 index 0000000000000000000000000000000000000000..2dbc76e87c99c6fd8d64e9ee19c38a817443b93a --- /dev/null +++ b/data/vals_ai/alibaba/qwen3-max/4a1afd14-4494-434e-a47f-5198d481f3b7.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/alibaba_qwen3-max/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3-max", + "id": "alibaba/qwen3-max", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3-max", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 85.392, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.392", + "raw_stderr": "0.865", + "latency": "0", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.865, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.545, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.545", + "raw_stderr": "0.2", + "latency": "0", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.2, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 84.574, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.574", + "raw_stderr": "0.607", + "latency": "0", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.607, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.861, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.861", + "raw_stderr": "0.399", + "latency": "0", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.399, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.432, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.432", + "raw_stderr": "0.769", + "latency": "0", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.769, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.362, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.362", + "raw_stderr": "1.493", + "latency": "0", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.493, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3-max/5f64fef4-85d7-4215-8841-23324cacec00.json b/data/vals_ai/alibaba/qwen3-max/5f64fef4-85d7-4215-8841-23324cacec00.json new file mode 100644 index 0000000000000000000000000000000000000000..89b71bf7b3dbaf3ee37113653fedc5dcea8c0aef --- /dev/null +++ b/data/vals_ai/alibaba/qwen3-max/5f64fef4-85d7-4215-8841-23324cacec00.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/alibaba_qwen3-max/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3-max", + "id": "alibaba/qwen3-max", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3-max", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.748, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.748", + "raw_stderr": "0.969", + "latency": "0", + "cost_per_test": "0.005556", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.969, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.047, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.047", + "raw_stderr": "1.03", + "latency": "0", + "cost_per_test": "0.006394", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.03, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.283, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.283", + "raw_stderr": "0.88", + "latency": "0", + "cost_per_test": "0.008748", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.88, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.293, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.293", + "raw_stderr": "1.588", + "latency": "0", + "cost_per_test": "0.006896", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.588, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.322, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.322", + "raw_stderr": "1.145", + "latency": "0", + "cost_per_test": "0.00436", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.145, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.289, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.289", + "raw_stderr": "1.278", + "latency": "0", + "cost_per_test": "0.014173", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.278, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.484, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.484", + "raw_stderr": "1.437", + "latency": "0", + "cost_per_test": "0.004609", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.437, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.278, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.278", + "raw_stderr": "2.239", + "latency": "0", + "cost_per_test": "0.005804", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.239, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 69.755, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.755", + "raw_stderr": "1.384", + "latency": "0", + "cost_per_test": "0.00685", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.384, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 95.041, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "95.041", + "raw_stderr": "0.591", + "latency": "0", + "cost_per_test": "0.007666", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.591, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.519, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.519", + "raw_stderr": "1.303", + "latency": "0", + "cost_per_test": "0.003535", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.303, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.36, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.36", + "raw_stderr": "0.356", + "latency": "0", + "cost_per_test": "0.006426", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.356, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.557, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.557", + "raw_stderr": "1.837", + "latency": "0", + "cost_per_test": "0.003992", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.837, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.608, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.608", + "raw_stderr": "0.809", + "latency": "0", + "cost_per_test": "0.007626", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.809, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.812, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.812", + "raw_stderr": "1.251", + "latency": "0", + "cost_per_test": "0.00375", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.251, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3-max/82744829-26cb-4ba1-a61d-6d5e6e3e2e42.json b/data/vals_ai/alibaba/qwen3-max/82744829-26cb-4ba1-a61d-6d5e6e3e2e42.json new file mode 100644 index 0000000000000000000000000000000000000000..51eb89a6a7e305b511ab525cf45b01f13c1dff8d --- /dev/null +++ b/data/vals_ai/alibaba/qwen3-max/82744829-26cb-4ba1-a61d-6d5e6e3e2e42.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/alibaba_qwen3-max/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3-max", + "id": "alibaba/qwen3-max", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3-max", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 21.5, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "21.5", + "raw_stderr": "12.27", + "latency": "1069.464", + "cost_per_test": "0.415113", + "temperature": "0.7", + "max_output_tokens": "32768", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 12.27, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 9.833, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "9.833", + "raw_stderr": "3.295", + "latency": "3021.03", + "cost_per_test": "1.233933", + "temperature": "0.7", + "max_output_tokens": "32768", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 3.295, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 15.666, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "15.666", + "raw_stderr": "6.352", + "latency": "2045.247", + "cost_per_test": "0.824523", + "temperature": "0.7", + "max_output_tokens": "32768", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 6.352, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3-max/be70ca19-ebec-4fb9-83d6-b99092cebcd4.json b/data/vals_ai/alibaba/qwen3-max/be70ca19-ebec-4fb9-83d6-b99092cebcd4.json new file mode 100644 index 0000000000000000000000000000000000000000..b533150a7c5c51579131de0df4038500d759ea6e --- /dev/null +++ b/data/vals_ai/alibaba/qwen3-max/be70ca19-ebec-4fb9-83d6-b99092cebcd4.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/alibaba_qwen3-max/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3-max", + "id": "alibaba/qwen3-max", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3-max", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 99.068, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "99.068", + "raw_stderr": "0.535", + "latency": "24.441", + "cost_per_test": "0.005115", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 52.286, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "52.286", + "raw_stderr": "2.67", + "latency": "217.563", + "cost_per_test": "0.042105", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.67, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 83.29, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "83.29", + "raw_stderr": "1.906", + "latency": "81.531", + "cost_per_test": "0.016419", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.906, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 78.215, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "78.215", + "raw_stderr": "1.108", + "latency": "107.845", + "cost_per_test": "0.021213", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.108, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3-max/d8f6f3be-d70a-4ac7-a5c0-e1df2fb935b7.json b/data/vals_ai/alibaba/qwen3-max/d8f6f3be-d70a-4ac7-a5c0-e1df2fb935b7.json new file mode 100644 index 0000000000000000000000000000000000000000..5a77dd11ac698cf3c93f4bfcaf1a43fe223ca68a --- /dev/null +++ b/data/vals_ai/alibaba/qwen3-max/d8f6f3be-d70a-4ac7-a5c0-e1df2fb935b7.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/alibaba_qwen3-max/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3-max", + "id": "alibaba/qwen3-max", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3-max", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 78.788, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.788", + "raw_stderr": "2.905", + "latency": "68.635", + "cost_per_test": "0.017168", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.905, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 80.303, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.303", + "raw_stderr": "2.826", + "latency": "71.669", + "cost_per_test": "0.015765", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.826, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 79.546, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.546", + "raw_stderr": "2.026", + "latency": "70.152", + "cost_per_test": "0.016466", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.026, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3-max/f1c5fa5b-007a-4efe-b886-385a3277705d.json b/data/vals_ai/alibaba/qwen3-max/f1c5fa5b-007a-4efe-b886-385a3277705d.json new file mode 100644 index 0000000000000000000000000000000000000000..5ec5673c3d57ab04546112ee4d8a2a8e68d68b47 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3-max/f1c5fa5b-007a-4efe-b886-385a3277705d.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/alibaba_qwen3-max/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3-max", + "id": "alibaba/qwen3-max", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3-max", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 59.79, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.79", + "raw_stderr": "1.674", + "latency": "7.092", + "cost_per_test": "0.002796", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.674, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 52.681, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.681", + "raw_stderr": "1.705", + "latency": "116.245", + "cost_per_test": "0.177377", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.705, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 55.944, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "55.944", + "raw_stderr": "0.977", + "latency": "55.675", + "cost_per_test": "0.086893", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.977, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:alibaba/qwen3-max:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 55.361, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "55.361", + "raw_stderr": "1.697", + "latency": "43.689", + "cost_per_test": "0.080507", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.697, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3-vl-plus-2025-09-23/0ebe5569-d2ee-407a-9649-06eabd50fa38.json b/data/vals_ai/alibaba/qwen3-vl-plus-2025-09-23/0ebe5569-d2ee-407a-9649-06eabd50fa38.json new file mode 100644 index 0000000000000000000000000000000000000000..90af16acc4264ef6c44344920a46dc14262f3181 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3-vl-plus-2025-09-23/0ebe5569-d2ee-407a-9649-06eabd50fa38.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/alibaba_qwen3-vl-plus-2025-09-23/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3-vl-plus-2025-09-23", + "id": "alibaba/qwen3-vl-plus-2025-09-23", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3-vl-plus-2025-09-23", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:alibaba/qwen3-vl-plus-2025-09-23:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 77.129, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "77.129", + "raw_stderr": "1.916", + "latency": "71.346", + "cost_per_test": "0.02022", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.916, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3-vl-plus-2025-09-23/b88d95bf-96b2-450b-9f29-2363ae0f4bbc.json b/data/vals_ai/alibaba/qwen3-vl-plus-2025-09-23/b88d95bf-96b2-450b-9f29-2363ae0f4bbc.json new file mode 100644 index 0000000000000000000000000000000000000000..920bfcb734a366c392d1c0c3a082acbdab749409 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3-vl-plus-2025-09-23/b88d95bf-96b2-450b-9f29-2363ae0f4bbc.json @@ -0,0 +1,346 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/alibaba_qwen3-vl-plus-2025-09-23/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3-vl-plus-2025-09-23", + "id": "alibaba/qwen3-vl-plus-2025-09-23", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3-vl-plus-2025-09-23", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:alibaba/qwen3-vl-plus-2025-09-23:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 26.984, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "26.984", + "raw_stderr": "5.57", + "latency": "72.667", + "cost_per_test": "0.003717", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 5.57, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:alibaba/qwen3-vl-plus-2025-09-23:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 45.635, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "45.635", + "raw_stderr": "6.376", + "latency": "81.998", + "cost_per_test": "0.004252", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 6.376, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:alibaba/qwen3-vl-plus-2025-09-23:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 15.432, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "15.432", + "raw_stderr": "4.81", + "latency": "96.375", + "cost_per_test": "0.01699", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 4.81, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:alibaba/qwen3-vl-plus-2025-09-23:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 53.673, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "53.673", + "raw_stderr": "5.601", + "latency": "67.761", + "cost_per_test": "0.006561", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 5.601, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:alibaba/qwen3-vl-plus-2025-09-23:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 42.097, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "42.097", + "raw_stderr": "3.384", + "latency": "74.142", + "cost_per_test": "0.004843", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 3.384, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.5-flash/08a9a11e-212b-4b08-9aed-82c1a69a3c80.json b/data/vals_ai/alibaba/qwen3.5-flash/08a9a11e-212b-4b08-9aed-82c1a69a3c80.json new file mode 100644 index 0000000000000000000000000000000000000000..077c93026a6a56d2a4b94300eb610ee7be36f93a --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.5-flash/08a9a11e-212b-4b08-9aed-82c1a69a3c80.json @@ -0,0 +1,346 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/alibaba_qwen3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.5-flash", + "id": "alibaba/qwen3.5-flash", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.5-flash", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 39.683, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "39.683", + "raw_stderr": "6.148", + "latency": "91.316", + "cost_per_test": "0.005847", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 6.148, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 34.921, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "34.921", + "raw_stderr": "6.195", + "latency": "79.312", + "cost_per_test": "0.005093", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 6.195, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 44.444, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "44.444", + "raw_stderr": "6.225", + "latency": "138.842", + "cost_per_test": "0.011357", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 6.225, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 52.863, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "52.863", + "raw_stderr": "5.62", + "latency": "84.486", + "cost_per_test": "0.006619", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 5.62, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 42.489, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "42.489", + "raw_stderr": "3.46", + "latency": "85.038", + "cost_per_test": "0.005853", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 3.46, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.5-flash/1caee314-75b5-46d4-964b-1c33e7984529.json b/data/vals_ai/alibaba/qwen3.5-flash/1caee314-75b5-46d4-964b-1c33e7984529.json new file mode 100644 index 0000000000000000000000000000000000000000..9639e695ff89fb02eba5e5db58fe626ec52464ac --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.5-flash/1caee314-75b5-46d4-964b-1c33e7984529.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/alibaba_qwen3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.5-flash", + "id": "alibaba/qwen3.5-flash", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.5-flash", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 70.619, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "70.619", + "raw_stderr": "2.09", + "latency": "80.406", + "cost_per_test": "0.004425", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.09, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.5-flash/258d9875-1c6e-4088-aafe-f5bd4ecc2e04.json b/data/vals_ai/alibaba/qwen3.5-flash/258d9875-1c6e-4088-aafe-f5bd4ecc2e04.json new file mode 100644 index 0000000000000000000000000000000000000000..3cdeb4886c4f16f35418c71dc5a39931ed940877 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.5-flash/258d9875-1c6e-4088-aafe-f5bd4ecc2e04.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/alibaba_qwen3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.5-flash", + "id": "alibaba/qwen3.5-flash", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.5-flash", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 74.96, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "74.96", + "raw_stderr": "1.221", + "latency": "48.321", + "cost_per_test": "0.002844", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.221, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 67.368, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.368", + "raw_stderr": "0.922", + "latency": "36.822", + "cost_per_test": "0.002158", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.922, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 59.777, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.777", + "raw_stderr": "1.382", + "latency": "25.322", + "cost_per_test": "0.001473", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.382, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.5-flash/38f436ee-7d68-44b3-8110-5276037b28c1.json b/data/vals_ai/alibaba/qwen3.5-flash/38f436ee-7d68-44b3-8110-5276037b28c1.json new file mode 100644 index 0000000000000000000000000000000000000000..65cefb93bc8a4b1c5fc68c6202a1d4837447d5e1 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.5-flash/38f436ee-7d68-44b3-8110-5276037b28c1.json @@ -0,0 +1,101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/alibaba_qwen3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.5-flash", + "id": "alibaba/qwen3.5-flash", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.5-flash", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 81.908, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.908", + "raw_stderr": "0.926", + "latency": "45.836", + "cost_per_test": "0.002615", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.926, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.5-flash/5231cbbf-ec58-4f42-b895-299b9e2f1959.json b/data/vals_ai/alibaba/qwen3.5-flash/5231cbbf-ec58-4f42-b895-299b9e2f1959.json new file mode 100644 index 0000000000000000000000000000000000000000..65c7148d32235b2d9a66ed98f8da5c26b2b4cecf --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.5-flash/5231cbbf-ec58-4f42-b895-299b9e2f1959.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/alibaba_qwen3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.5-flash", + "id": "alibaba/qwen3.5-flash", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.5-flash", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.026, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.026", + "raw_stderr": "0.951", + "latency": "24.822", + "cost_per_test": "0.00137", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.951, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.354, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.354", + "raw_stderr": "1.098", + "latency": "26.597", + "cost_per_test": "0.001478", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.098, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.078, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.078", + "raw_stderr": "0.847", + "latency": "39.771", + "cost_per_test": "0.002165", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.847, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.561, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.561", + "raw_stderr": "1.63", + "latency": "32.653", + "cost_per_test": "0.001708", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.63, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.336, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.336", + "raw_stderr": "1.062", + "latency": "24.247", + "cost_per_test": "0.001311", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.062, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.456, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.456", + "raw_stderr": "1.222", + "latency": "56.165", + "cost_per_test": "0.002943", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.222, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.484, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.484", + "raw_stderr": "1.437", + "latency": "30.761", + "cost_per_test": "0.001703", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.437, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.391, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.391", + "raw_stderr": "2.315", + "latency": "36.51", + "cost_per_test": "0.001704", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.315, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 66.939, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.939", + "raw_stderr": "1.418", + "latency": "48.279", + "cost_per_test": "0.002347", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.418, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.893, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.893", + "raw_stderr": "0.599", + "latency": "27.458", + "cost_per_test": "0.001691", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.599, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.788, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.788", + "raw_stderr": "1.345", + "latency": "25.541", + "cost_per_test": "0.001274", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.345, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.061, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.061", + "raw_stderr": "0.36", + "latency": "33.303", + "cost_per_test": "0.001734", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.36, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.958, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.958", + "raw_stderr": "1.825", + "latency": "36.885", + "cost_per_test": "0.00145", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.825, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.378, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.378", + "raw_stderr": "0.779", + "latency": "31.736", + "cost_per_test": "0.001891", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.779, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.208, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.208", + "raw_stderr": "1.323", + "latency": "24.811", + "cost_per_test": "0.001243", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.323, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.5-flash/70af6be5-ee2d-44c4-8b81-e29bb5d7a474.json b/data/vals_ai/alibaba/qwen3.5-flash/70af6be5-ee2d-44c4-8b81-e29bb5d7a474.json new file mode 100644 index 0000000000000000000000000000000000000000..dc50e9ffb18c1dc664e17b81b73c388a9304d215 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.5-flash/70af6be5-ee2d-44c4-8b81-e29bb5d7a474.json @@ -0,0 +1,371 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/alibaba_qwen3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.5-flash", + "id": "alibaba/qwen3.5-flash", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.5-flash", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 14.286, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "14.286", + "raw_stderr": "5.465", + "latency": "540.997", + "cost_per_test": "0.421763", + "temperature": "0.6", + "top_p": "0.95", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 5.465, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 0.95, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 62.069, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "62.069", + "raw_stderr": "3.009", + "latency": "387.073", + "cost_per_test": "0.292058", + "temperature": "0.6", + "top_p": "0.95", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 3.009, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 0.95, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 78.866, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "78.866", + "raw_stderr": "2.939", + "latency": "291.905", + "cost_per_test": "0.189609", + "temperature": "0.6", + "top_p": "0.95", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.939, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 0.95, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "391.006", + "cost_per_test": "0.296017", + "temperature": "0.6", + "top_p": "0.95", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 0.95, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 64.4, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "64.4", + "raw_stderr": "2.143", + "latency": "363.101", + "cost_per_test": "0.263227", + "temperature": "0.6", + "top_p": "0.95", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.143, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 0.95, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.5-flash/77420e6e-cd4e-4de3-a2d1-2996e7614673.json b/data/vals_ai/alibaba/qwen3.5-flash/77420e6e-cd4e-4de3-a2d1-2996e7614673.json new file mode 100644 index 0000000000000000000000000000000000000000..d7118e636de65720a7ab0dd7507de6be744d367d --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.5-flash/77420e6e-cd4e-4de3-a2d1-2996e7614673.json @@ -0,0 +1,426 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/alibaba_qwen3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.5-flash", + "id": "alibaba/qwen3.5-flash", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.5-flash", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 88.542, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.542", + "raw_stderr": "0.815", + "latency": "14.198", + "cost_per_test": "0.000612", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.815, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.685, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.685", + "raw_stderr": "0.205", + "latency": "16.742", + "cost_per_test": "0.000755", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.205, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 90.869, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.869", + "raw_stderr": "0.547", + "latency": "17.011", + "cost_per_test": "0.000824", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.547, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 84.276, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.276", + "raw_stderr": "0.41", + "latency": "16.73", + "cost_per_test": "0.00076", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.41, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.564, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.564", + "raw_stderr": "0.883", + "latency": "16.071", + "cost_per_test": "0.00076", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.883, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.722, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.722", + "raw_stderr": "1.557", + "latency": "19.626", + "cost_per_test": "0.00085", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.557, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.5-flash/8cd41d93-bb2b-4dd6-b126-ec37c2d5e812.json b/data/vals_ai/alibaba/qwen3.5-flash/8cd41d93-bb2b-4dd6-b126-ec37c2d5e812.json new file mode 100644 index 0000000000000000000000000000000000000000..efe84fe80e8997053cd6e595d273abd580bb38da --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.5-flash/8cd41d93-bb2b-4dd6-b126-ec37c2d5e812.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/alibaba_qwen3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.5-flash", + "id": "alibaba/qwen3.5-flash", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.5-flash", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.758, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.758", + "raw_stderr": "0.617", + "latency": "86.637", + "cost_per_test": "0.003773", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.617, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 62.571, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "62.571", + "raw_stderr": "2.587", + "latency": "348.754", + "cost_per_test": "0.016007", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.587, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 88.512, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "88.512", + "raw_stderr": "1.629", + "latency": "192.49", + "cost_per_test": "0.008847", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.629, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 83.28, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "83.28", + "raw_stderr": "1.04", + "latency": "209.294", + "cost_per_test": "0.009542", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.04, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.5-flash/b4dd40d1-6862-4052-a89f-29c12f3c4a03.json b/data/vals_ai/alibaba/qwen3.5-flash/b4dd40d1-6862-4052-a89f-29c12f3c4a03.json new file mode 100644 index 0000000000000000000000000000000000000000..4cb51a075a8395db1218a402280b8a1ba6d6830e --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.5-flash/b4dd40d1-6862-4052-a89f-29c12f3c4a03.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/alibaba_qwen3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.5-flash", + "id": "alibaba/qwen3.5-flash", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.5-flash", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 82.323, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.323", + "raw_stderr": "2.711", + "latency": "77.318", + "cost_per_test": "0.003293", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.711, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 83.333, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.333", + "raw_stderr": "2.649", + "latency": "82.449", + "cost_per_test": "0.003533", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.649, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 82.828, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.828", + "raw_stderr": "1.895", + "latency": "79.884", + "cost_per_test": "0.003413", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.895, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.5-flash/c3f941a4-d4b7-432f-aa47-58e54f72cc46.json b/data/vals_ai/alibaba/qwen3.5-flash/c3f941a4-d4b7-432f-aa47-58e54f72cc46.json new file mode 100644 index 0000000000000000000000000000000000000000..4ff0ca47c3fe1624a4e42c8a4289a6e952b2a195 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.5-flash/c3f941a4-d4b7-432f-aa47-58e54f72cc46.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/alibaba_qwen3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.5-flash", + "id": "alibaba/qwen3.5-flash", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.5-flash", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 60.752, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.752", + "raw_stderr": "1.396", + "latency": "73.003", + "cost_per_test": "0.003099", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.396, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 72.158, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "72.158", + "raw_stderr": "0.876", + "latency": "73.003", + "cost_per_test": "0.003099", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.876, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 83.565, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "83.565", + "raw_stderr": "1.06", + "latency": "73.003", + "cost_per_test": "0.003099", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.06, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.5-flash/e2481486-0b8b-4e43-bfac-f9d1be6ffde8.json b/data/vals_ai/alibaba/qwen3.5-flash/e2481486-0b8b-4e43-bfac-f9d1be6ffde8.json new file mode 100644 index 0000000000000000000000000000000000000000..30a04ab11da9ee144915d715c8f88cc5fc8f385a --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.5-flash/e2481486-0b8b-4e43-bfac-f9d1be6ffde8.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/alibaba_qwen3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.5-flash", + "id": "alibaba/qwen3.5-flash", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.5-flash", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.723, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.723", + "raw_stderr": "1.667", + "latency": "13.893", + "cost_per_test": "0.00112", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.667, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 65.734, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.734", + "raw_stderr": "1.62", + "latency": "57.235", + "cost_per_test": "0.019801", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.62, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 63.559, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.559", + "raw_stderr": "0.947", + "latency": "35.198", + "cost_per_test": "0.009638", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.947, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:alibaba/qwen3.5-flash:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 64.219, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.219", + "raw_stderr": "1.636", + "latency": "34.465", + "cost_per_test": "0.007994", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.636, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.5-plus-thinking/1b7f1823-429e-482b-963b-62c836e15163.json b/data/vals_ai/alibaba/qwen3.5-plus-thinking/1b7f1823-429e-482b-963b-62c836e15163.json new file mode 100644 index 0000000000000000000000000000000000000000..321230371456545b085247c6a57c8bbd4b0b4311 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.5-plus-thinking/1b7f1823-429e-482b-963b-62c836e15163.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/alibaba_qwen3.5-plus-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.5-plus-thinking", + "id": "alibaba/qwen3.5-plus-thinking", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.5-plus-thinking", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 87.879, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.879", + "raw_stderr": "2.319", + "latency": "119.018", + "cost_per_test": "0.017962", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.319, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 86.869, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.869", + "raw_stderr": "2.4", + "latency": "88.527", + "cost_per_test": "0.019122", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.4, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 87.374, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.374", + "raw_stderr": "1.669", + "latency": "103.772", + "cost_per_test": "0.018542", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.669, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.5-plus-thinking/29d4016d-ea18-4c8a-820d-02675df2d133.json b/data/vals_ai/alibaba/qwen3.5-plus-thinking/29d4016d-ea18-4c8a-820d-02675df2d133.json new file mode 100644 index 0000000000000000000000000000000000000000..6d17315e930b207a5a39b0eb1e203c98e67c878b --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.5-plus-thinking/29d4016d-ea18-4c8a-820d-02675df2d133.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/alibaba_qwen3.5-plus-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.5-plus-thinking", + "id": "alibaba/qwen3.5-plus-thinking", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.5-plus-thinking", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 15.738, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "15.738", + "raw_stderr": "3.179", + "latency": "3015.618", + "cost_per_test": "3.803761", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 3.179, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.5-plus-thinking/3f874e2d-68ce-4ee4-890b-dad293f6bfcc.json b/data/vals_ai/alibaba/qwen3.5-plus-thinking/3f874e2d-68ce-4ee4-890b-dad293f6bfcc.json new file mode 100644 index 0000000000000000000000000000000000000000..4455c2cea03ae66e43f6f87ff220f36bb2bb99b3 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.5-plus-thinking/3f874e2d-68ce-4ee4-890b-dad293f6bfcc.json @@ -0,0 +1,426 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/alibaba_qwen3.5-plus-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.5-plus-thinking", + "id": "alibaba/qwen3.5-plus-thinking", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.5-plus-thinking", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 89.939, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.939", + "raw_stderr": "0.765", + "latency": "30.444", + "cost_per_test": "0.003255", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.765, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.869, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.869", + "raw_stderr": "0.198", + "latency": "32.807", + "cost_per_test": "0.004016", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.198, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 91.166, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "91.166", + "raw_stderr": "0.538", + "latency": "26.931", + "cost_per_test": "0.003432", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.538, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 85.104, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.104", + "raw_stderr": "0.408", + "latency": "30.422", + "cost_per_test": "0.003671", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.408, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.221, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.221", + "raw_stderr": "0.887", + "latency": "29.649", + "cost_per_test": "0.003689", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.887, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.325, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.325", + "raw_stderr": "1.571", + "latency": "32.28", + "cost_per_test": "0.003962", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.571, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.5-plus-thinking/4d2b3650-8339-4acc-8240-2c4164400e32.json b/data/vals_ai/alibaba/qwen3.5-plus-thinking/4d2b3650-8339-4acc-8240-2c4164400e32.json new file mode 100644 index 0000000000000000000000000000000000000000..c5a136d3496e8ea00d001d07722396184a29a175 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.5-plus-thinking/4d2b3650-8339-4acc-8240-2c4164400e32.json @@ -0,0 +1,361 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/alibaba_qwen3.5-plus-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.5-plus-thinking", + "id": "alibaba/qwen3.5-plus-thinking", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.5-plus-thinking", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 35.714, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "35.714", + "raw_stderr": "7.483", + "latency": "692.899", + "cost_per_test": "1.388493", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 7.483, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 67.816, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "67.816", + "raw_stderr": "2.897", + "latency": "562.405", + "cost_per_test": "1.157398", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.897, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 84.536, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "84.536", + "raw_stderr": "2.603", + "latency": "359.398", + "cost_per_test": "0.602475", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.603, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "836.564", + "cost_per_test": "1.499524", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 71.2, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "71.2", + "raw_stderr": "2.027", + "latency": "496.245", + "cost_per_test": "0.963553", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.027, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.5-plus-thinking/84d1df28-a1bf-4467-8e67-d3141c440ddb.json b/data/vals_ai/alibaba/qwen3.5-plus-thinking/84d1df28-a1bf-4467-8e67-d3141c440ddb.json new file mode 100644 index 0000000000000000000000000000000000000000..3adf4c8e4dcf33d0b52f3626c9826ff105463928 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.5-plus-thinking/84d1df28-a1bf-4467-8e67-d3141c440ddb.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/alibaba_qwen3.5-plus-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.5-plus-thinking", + "id": "alibaba/qwen3.5-plus-thinking", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.5-plus-thinking", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.142, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.142", + "raw_stderr": "0.877", + "latency": "55.706", + "cost_per_test": "0.006883", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.877, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.142, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.142", + "raw_stderr": "0.958", + "latency": "65.495", + "cost_per_test": "0.007925", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.958, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.226, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.226", + "raw_stderr": "0.796", + "latency": "92.131", + "cost_per_test": "0.011862", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.796, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.976, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.976", + "raw_stderr": "1.415", + "latency": "62.583", + "cost_per_test": "0.008357", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.415, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.692, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.692", + "raw_stderr": "1.047", + "latency": "65.942", + "cost_per_test": "0.006729", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.047, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.243, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.243", + "raw_stderr": "1.139", + "latency": "152.966", + "cost_per_test": "0.015775", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.139, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.073, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.073", + "raw_stderr": "1.397", + "latency": "65.695", + "cost_per_test": "0.008763", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.397, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.74, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.74", + "raw_stderr": "2.096", + "latency": "75.191", + "cost_per_test": "0.008337", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.096, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.933, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.933", + "raw_stderr": "1.323", + "latency": "98.798", + "cost_per_test": "0.011867", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.323, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 95.559, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "95.559", + "raw_stderr": "0.56", + "latency": "78.179", + "cost_per_test": "0.009015", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.56, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.468, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.468", + "raw_stderr": "1.251", + "latency": "62.772", + "cost_per_test": "0.0061", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.251, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.178, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.178", + "raw_stderr": "0.328", + "latency": "76.678", + "cost_per_test": "0.008912", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.328, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.571, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.571", + "raw_stderr": "1.573", + "latency": "60.923", + "cost_per_test": "0.00662", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.573, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.764, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.764", + "raw_stderr": "0.719", + "latency": "81.557", + "cost_per_test": "0.010281", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.719, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.967, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.967", + "raw_stderr": "1.192", + "latency": "55.554", + "cost_per_test": "0.006254", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.192, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.5-plus-thinking/86986cd0-49ac-4996-ba5d-c74062da6afe.json b/data/vals_ai/alibaba/qwen3.5-plus-thinking/86986cd0-49ac-4996-ba5d-c74062da6afe.json new file mode 100644 index 0000000000000000000000000000000000000000..eb4e2ba530ee38f19ab20f59618a55c2921fd27d --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.5-plus-thinking/86986cd0-49ac-4996-ba5d-c74062da6afe.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/alibaba_qwen3.5-plus-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.5-plus-thinking", + "id": "alibaba/qwen3.5-plus-thinking", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.5-plus-thinking", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.839, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.839", + "raw_stderr": "1.666", + "latency": "200.413", + "cost_per_test": "0.006714", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.666, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 67.249, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.249", + "raw_stderr": "1.602", + "latency": "226.284", + "cost_per_test": "0.133024", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.602, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 65.307, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.307", + "raw_stderr": "0.936", + "latency": "155.861", + "cost_per_test": "0.058073", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.936, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 67.832, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.832", + "raw_stderr": "1.595", + "latency": "40.886", + "cost_per_test": "0.034482", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.595, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.5-plus-thinking/b8536e15-e431-497a-918c-4c6ed0929ffb.json b/data/vals_ai/alibaba/qwen3.5-plus-thinking/b8536e15-e431-497a-918c-4c6ed0929ffb.json new file mode 100644 index 0000000000000000000000000000000000000000..a1ed42d0b98fdb3ce2b0f998adcb7eaf2c8fcf98 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.5-plus-thinking/b8536e15-e431-497a-918c-4c6ed0929ffb.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/alibaba_qwen3.5-plus-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.5-plus-thinking", + "id": "alibaba/qwen3.5-plus-thinking", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.5-plus-thinking", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.758, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.758", + "raw_stderr": "0.617", + "latency": "145.34", + "cost_per_test": "0.027081", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.617, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 67.143, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "67.143", + "raw_stderr": "2.511", + "latency": "355.842", + "cost_per_test": "0.07058", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.511, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 90.078, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "90.078", + "raw_stderr": "1.528", + "latency": "241.235", + "cost_per_test": "0.045918", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.528, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 85.326, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "85.326", + "raw_stderr": "1.001", + "latency": "247.472", + "cost_per_test": "0.04786", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.001, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.5-plus-thinking/c9f47094-4b77-46a7-9240-75ccde71516b.json b/data/vals_ai/alibaba/qwen3.5-plus-thinking/c9f47094-4b77-46a7-9240-75ccde71516b.json new file mode 100644 index 0000000000000000000000000000000000000000..01b6d8ad732fcbc12d04413d72cd7f540ef9470a --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.5-plus-thinking/c9f47094-4b77-46a7-9240-75ccde71516b.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/alibaba_qwen3.5-plus-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.5-plus-thinking", + "id": "alibaba/qwen3.5-plus-thinking", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.5-plus-thinking", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 62.878, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.878", + "raw_stderr": "1.362", + "latency": "83.537", + "cost_per_test": "0.01396", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.362, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 60.772, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.772", + "raw_stderr": "0.972", + "latency": "62.075", + "cost_per_test": "0.010812", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.972, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 58.665, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.665", + "raw_stderr": "1.388", + "latency": "40.613", + "cost_per_test": "0.007665", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.388, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.5-plus-thinking/f5028c1e-2de5-4ce2-94c4-3bf3b7d29efd.json b/data/vals_ai/alibaba/qwen3.5-plus-thinking/f5028c1e-2de5-4ce2-94c4-3bf3b7d29efd.json new file mode 100644 index 0000000000000000000000000000000000000000..c0a36007723b480174b3d2267cf62274654671b9 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.5-plus-thinking/f5028c1e-2de5-4ce2-94c4-3bf3b7d29efd.json @@ -0,0 +1,346 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/alibaba_qwen3.5-plus-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.5-plus-thinking", + "id": "alibaba/qwen3.5-plus-thinking", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.5-plus-thinking", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 19.355, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "19.355", + "raw_stderr": "5.013", + "latency": "341.733", + "cost_per_test": "0.040561", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 5.013, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 27.778, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "27.778", + "raw_stderr": "5.743", + "latency": "296.836", + "cost_per_test": "0.032038", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 5.743, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 13.272, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "13.272", + "raw_stderr": "4.785", + "latency": "345.421", + "cost_per_test": "0.051045", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 4.785, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 44.072, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "44.072", + "raw_stderr": "5.57", + "latency": "299.018", + "cost_per_test": "0.04062", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 5.57, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 30.402, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "30.402", + "raw_stderr": "3.147", + "latency": "312.529", + "cost_per_test": "0.03774", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 3.147, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.5-plus-thinking/f9293727-243d-446b-b72f-a6d5216c89ec.json b/data/vals_ai/alibaba/qwen3.5-plus-thinking/f9293727-243d-446b-b72f-a6d5216c89ec.json new file mode 100644 index 0000000000000000000000000000000000000000..2e8d49b143a5012052de42cdefccc613e7c5d9ce --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.5-plus-thinking/f9293727-243d-446b-b72f-a6d5216c89ec.json @@ -0,0 +1,101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/alibaba_qwen3.5-plus-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.5-plus-thinking", + "id": "alibaba/qwen3.5-plus-thinking", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.5-plus-thinking", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:alibaba/qwen3.5-plus-thinking:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 22.775, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "22.775", + "raw_stderr": "1.008", + "latency": "135.183", + "cost_per_test": "0.01513", + "temperature": "0.6", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.008, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.6-27b/082a41b7-10cf-4a26-8644-5244c8841625.json b/data/vals_ai/alibaba/qwen3.6-27b/082a41b7-10cf-4a26-8644-5244c8841625.json new file mode 100644 index 0000000000000000000000000000000000000000..8d93ccf775fadef72a6da1dbd6aed0e09510c1e7 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.6-27b/082a41b7-10cf-4a26-8644-5244c8841625.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/alibaba_qwen3.6-27b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.6-27b", + "id": "alibaba/qwen3.6-27b", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.6-27b", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:alibaba/qwen3.6-27b:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 11.941, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "11.941", + "raw_stderr": "3.414", + "latency": "9762.856", + "cost_per_test": "4.655127", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 3.414, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.6-27b/1fb4927a-8f80-4d0c-8106-aa71d676eda8.json b/data/vals_ai/alibaba/qwen3.6-27b/1fb4927a-8f80-4d0c-8106-aa71d676eda8.json new file mode 100644 index 0000000000000000000000000000000000000000..4eb15d5254e5915717db6dee135f4b01802efa7f --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.6-27b/1fb4927a-8f80-4d0c-8106-aa71d676eda8.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/alibaba_qwen3.6-27b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.6-27b", + "id": "alibaba/qwen3.6-27b", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.6-27b", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:alibaba/qwen3.6-27b:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 59.199, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.199", + "raw_stderr": "1.405", + "latency": "147.778", + "cost_per_test": "0.030256", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.405, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:alibaba/qwen3.6-27b:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 71.26, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.26", + "raw_stderr": "0.882", + "latency": "147.778", + "cost_per_test": "0.030256", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.882, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:alibaba/qwen3.6-27b:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 83.32, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "83.32", + "raw_stderr": "1.066", + "latency": "147.778", + "cost_per_test": "0.030256", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.066, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.6-27b/7a7044c3-0b70-4b42-ac4c-d3ce87cd50bd.json b/data/vals_ai/alibaba/qwen3.6-27b/7a7044c3-0b70-4b42-ac4c-d3ce87cd50bd.json new file mode 100644 index 0000000000000000000000000000000000000000..0dd76d503620eb456cd6ab1557dc707a0e2bfd81 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.6-27b/7a7044c3-0b70-4b42-ac4c-d3ce87cd50bd.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/alibaba_qwen3.6-27b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.6-27b", + "id": "alibaba/qwen3.6-27b", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.6-27b", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:alibaba/qwen3.6-27b:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 75.755, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "75.755", + "raw_stderr": "1.208", + "latency": "55.395", + "cost_per_test": "0.022666", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.208, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:alibaba/qwen3.6-27b:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 68.283, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.283", + "raw_stderr": "0.916", + "latency": "42.074", + "cost_per_test": "0.016906", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.916, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:alibaba/qwen3.6-27b:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 60.811, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.811", + "raw_stderr": "1.376", + "latency": "28.752", + "cost_per_test": "0.011147", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.376, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.6-27b/98c7ba99-3cdf-4df3-8748-dbae836549ad.json b/data/vals_ai/alibaba/qwen3.6-27b/98c7ba99-3cdf-4df3-8748-dbae836549ad.json new file mode 100644 index 0000000000000000000000000000000000000000..185bf94052cb45134ea748fe5d2a5f62dd43dad2 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.6-27b/98c7ba99-3cdf-4df3-8748-dbae836549ad.json @@ -0,0 +1,354 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/alibaba_qwen3.6-27b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.6-27b", + "id": "alibaba/qwen3.6-27b", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.6-27b", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:alibaba/qwen3.6-27b:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 41.27, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "41.27", + "raw_stderr": "6.185", + "latency": "188.201", + "cost_per_test": "0.05154", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 6.185, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:alibaba/qwen3.6-27b:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 35.714, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "35.714", + "raw_stderr": "6.073", + "latency": "174.695", + "cost_per_test": "0.047668", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 6.073, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:alibaba/qwen3.6-27b:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 18.519, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "18.519", + "raw_stderr": "5.315", + "latency": "125.145", + "cost_per_test": "0.072804", + "temperature": "1", + "top_p": "0.95", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 5.315, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:alibaba/qwen3.6-27b:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 59.746, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "59.746", + "raw_stderr": "5.531", + "latency": "114.854", + "cost_per_test": "0.051599", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 5.531, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:alibaba/qwen3.6-27b:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 45.577, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "45.577", + "raw_stderr": "3.427", + "latency": "159.25", + "cost_per_test": "0.050269", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 3.427, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.6-27b/b67f23e9-702a-4a81-b9a2-b21ab0d3619f.json b/data/vals_ai/alibaba/qwen3.6-27b/b67f23e9-702a-4a81-b9a2-b21ab0d3619f.json new file mode 100644 index 0000000000000000000000000000000000000000..56b7bc26d0f01e332115ac2866dfeb0209443a0a --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.6-27b/b67f23e9-702a-4a81-b9a2-b21ab0d3619f.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/alibaba_qwen3.6-27b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.6-27b", + "id": "alibaba/qwen3.6-27b", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.6-27b", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:alibaba/qwen3.6-27b:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 59.674, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.674", + "raw_stderr": "1.675", + "latency": "27.389", + "cost_per_test": "0.006658", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.675, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:alibaba/qwen3.6-27b:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 64.918, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.918", + "raw_stderr": "1.629", + "latency": "109.112", + "cost_per_test": "0.107726", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.629, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:alibaba/qwen3.6-27b:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 62.315, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.315", + "raw_stderr": "0.954", + "latency": "62.759", + "cost_per_test": "0.053549", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.954, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:alibaba/qwen3.6-27b:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 62.354, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.354", + "raw_stderr": "1.654", + "latency": "51.777", + "cost_per_test": "0.046263", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.654, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.6-27b/c2afdf4d-3cc5-43f6-b44d-3c120660d9a8.json b/data/vals_ai/alibaba/qwen3.6-27b/c2afdf4d-3cc5-43f6-b44d-3c120660d9a8.json new file mode 100644 index 0000000000000000000000000000000000000000..fd5cc7a62bc0f14a3a1fcd519323ba5ccc3238b8 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.6-27b/c2afdf4d-3cc5-43f6-b44d-3c120660d9a8.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/alibaba_qwen3.6-27b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.6-27b", + "id": "alibaba/qwen3.6-27b", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.6-27b", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:alibaba/qwen3.6-27b:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 8.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "8", + "raw_stderr": "2.713", + "latency": "695.339", + "cost_per_test": "0.512041", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.713, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.6-27b/ee27a536-e092-4cc9-bf28-e045224f8dee.json b/data/vals_ai/alibaba/qwen3.6-27b/ee27a536-e092-4cc9-bf28-e045224f8dee.json new file mode 100644 index 0000000000000000000000000000000000000000..9cf92e4fddb73db5553ad714ea01c638b8d4f43d --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.6-27b/ee27a536-e092-4cc9-bf28-e045224f8dee.json @@ -0,0 +1,371 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/alibaba_qwen3.6-27b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.6-27b", + "id": "alibaba/qwen3.6-27b", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.6-27b", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:alibaba/qwen3.6-27b:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "7.362", + "latency": "1758.448", + "cost_per_test": "1.757772", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 7.362, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:alibaba/qwen3.6-27b:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 68.199, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "68.199", + "raw_stderr": "2.888", + "latency": "1126.541", + "cost_per_test": "1.031067", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.888, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:alibaba/qwen3.6-27b:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 80.928, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "80.928", + "raw_stderr": "2.828", + "latency": "731.92", + "cost_per_test": "0.568419", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.828, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:alibaba/qwen3.6-27b:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "1497.429", + "cost_per_test": "1.287371", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:alibaba/qwen3.6-27b:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 70.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "70", + "raw_stderr": "2.051", + "latency": "1028.733", + "cost_per_test": "0.91414", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.051, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.6-max-preview/6a15f0a1-9e81-4f9e-b2d9-c85bad850d31.json b/data/vals_ai/alibaba/qwen3.6-max-preview/6a15f0a1-9e81-4f9e-b2d9-c85bad850d31.json new file mode 100644 index 0000000000000000000000000000000000000000..8a5cbbbc81248038175db0b8021969e7a138d090 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.6-max-preview/6a15f0a1-9e81-4f9e-b2d9-c85bad850d31.json @@ -0,0 +1,361 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/alibaba_qwen3.6-max-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.6-max-preview", + "id": "alibaba/qwen3.6-max-preview", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.6-max-preview", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:alibaba/qwen3.6-max-preview:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 30.952, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "30.952", + "raw_stderr": "7.22", + "latency": "1708.372", + "cost_per_test": "2.909365", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 7.22, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:alibaba/qwen3.6-max-preview:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 70.498, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "70.498", + "raw_stderr": "2.828", + "latency": "1209.658", + "cost_per_test": "1.884437", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.828, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:alibaba/qwen3.6-max-preview:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 85.567, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "85.567", + "raw_stderr": "2.53", + "latency": "759.47", + "cost_per_test": "0.826787", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.53, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:alibaba/qwen3.6-max-preview:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "1601.596", + "cost_per_test": "2.706236", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:alibaba/qwen3.6-max-preview:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 72.8, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "72.8", + "raw_stderr": "1.992", + "latency": "1079.228", + "cost_per_test": "1.565093", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.992, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.6-max-preview/f24d642f-f19a-46a0-ac58-5e1ca5e65172.json b/data/vals_ai/alibaba/qwen3.6-max-preview/f24d642f-f19a-46a0-ac58-5e1ca5e65172.json new file mode 100644 index 0000000000000000000000000000000000000000..8c3e7a0e40f748f60ac756842f384d350995d812 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.6-max-preview/f24d642f-f19a-46a0-ac58-5e1ca5e65172.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/alibaba_qwen3.6-max-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.6-max-preview", + "id": "alibaba/qwen3.6-max-preview", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.6-max-preview", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:alibaba/qwen3.6-max-preview:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 63.287, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.287", + "raw_stderr": "1.646", + "latency": "42.169", + "cost_per_test": "0.013772", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.646, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:alibaba/qwen3.6-max-preview:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 68.182, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.182", + "raw_stderr": "1.59", + "latency": "47.888", + "cost_per_test": "0.351906", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.59, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:alibaba/qwen3.6-max-preview:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 66.473, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.473", + "raw_stderr": "0.929", + "latency": "47.068", + "cost_per_test": "0.156029", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.929, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:alibaba/qwen3.6-max-preview:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 67.949, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.949", + "raw_stderr": "1.593", + "latency": "51.146", + "cost_per_test": "0.102408", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.593, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.6-plus/00093fd2-53ea-40e4-a4e6-5ee3de738d39.json b/data/vals_ai/alibaba/qwen3.6-plus/00093fd2-53ea-40e4-a4e6-5ee3de738d39.json new file mode 100644 index 0000000000000000000000000000000000000000..c3db3083e92bed3c99c537d866ae33e20bf4e247 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.6-plus/00093fd2-53ea-40e4-a4e6-5ee3de738d39.json @@ -0,0 +1,398 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/alibaba_qwen3.6-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.6-plus", + "id": "alibaba/qwen3.6-plus", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.6-plus", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 62.238, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "62.238", + "raw_stderr": "1.655", + "latency": "35.919", + "cost_per_test": "0.038254", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.655, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 40.88, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "40.88", + "raw_stderr": "0.729", + "latency": "585.203", + "cost_per_test": "0.630525", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.729, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:overall:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 48.887, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "48.887", + "raw_stderr": "1.455", + "latency": "835.329", + "cost_per_test": "1.443929", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.455, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 70.588, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "70.588", + "raw_stderr": "4.512", + "latency": "487.435", + "cost_per_test": "0.901327" + }, + "uncertainty": { + "standard_error": { + "value": 4.512, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 53.184, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "53.184", + "raw_stderr": "1.633", + "latency": "778.109", + "cost_per_test": "0.201902", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.633, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 28.253, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "28.253", + "raw_stderr": "6.126", + "latency": "2289.977", + "cost_per_test": "5.447636", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 6.126, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.6-plus/1e5d88fb-be5f-468f-8596-43d1fd121a60.json b/data/vals_ai/alibaba/qwen3.6-plus/1e5d88fb-be5f-468f-8596-43d1fd121a60.json new file mode 100644 index 0000000000000000000000000000000000000000..c8a3c8b4400781792f7f3cd04a93db4dcecb8dcb --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.6-plus/1e5d88fb-be5f-468f-8596-43d1fd121a60.json @@ -0,0 +1,522 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_multimodal_index/alibaba_qwen3.6-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Multimodal Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.6-plus", + "id": "alibaba/qwen3.6-plus", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.6-plus", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_multimodal_index:corp_fin_v2:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.vals_multimodal_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Multimodal Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (CorpFin v2).", + "metric_id": "vals_ai.vals_multimodal_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 62.238, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "62.238", + "raw_stderr": "1.655", + "latency": "35.919", + "cost_per_test": "0.038254", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.655, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:finance_agent:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.vals_multimodal_index.finance_agent", + "source_data": { + "dataset_name": "Vals Multimodal Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_multimodal_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 40.88, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "40.88", + "raw_stderr": "0.729", + "latency": "585.203", + "cost_per_test": "0.630525", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.729, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:mortgage_tax:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.vals_multimodal_index.mortgage_tax", + "source_data": { + "dataset_name": "Vals Multimodal Index - Mortgage Tax", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "mortgage_tax", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Mortgage Tax).", + "metric_id": "vals_ai.vals_multimodal_index.mortgage_tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 67.965, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "mortgage_tax", + "task_name": "Mortgage Tax", + "dataset_type": "private", + "industry": "index", + "raw_score": "67.965", + "raw_stderr": "0.908", + "latency": "78.18", + "cost_per_test": "0.013476", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.908, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:overall:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.vals_multimodal_index.overall", + "source_data": { + "dataset_name": "Vals Multimodal Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Overall).", + "metric_id": "vals_ai.vals_multimodal_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 51.516, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "51.516", + "raw_stderr": "1.327", + "latency": "640.584", + "cost_per_test": "1.039221", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.327, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:sage:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.vals_multimodal_index.sage", + "source_data": { + "dataset_name": "Vals Multimodal Index - SAGE", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "sage", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SAGE).", + "metric_id": "vals_ai.vals_multimodal_index.sage.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 44.86, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "sage", + "task_name": "SAGE", + "dataset_type": "private", + "industry": "index", + "raw_score": "44.86", + "raw_stderr": "3.433", + "latency": "229.267", + "cost_per_test": "0.041429", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 3.433, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:swebench:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.vals_multimodal_index.swebench", + "source_data": { + "dataset_name": "Vals Multimodal Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SWE-bench).", + "metric_id": "vals_ai.vals_multimodal_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 70.588, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "70.588", + "raw_stderr": "4.512", + "latency": "487.435", + "cost_per_test": "0.901327" + }, + "uncertainty": { + "standard_error": { + "value": 4.512, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:terminal_bench_2_1:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.vals_multimodal_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Multimodal Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_multimodal_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 53.184, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "53.184", + "raw_stderr": "1.633", + "latency": "778.109", + "cost_per_test": "0.201902", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.633, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:vibe_code_bench:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.vals_multimodal_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Multimodal Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_multimodal_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 28.253, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "28.253", + "raw_stderr": "6.126", + "latency": "2289.977", + "cost_per_test": "5.447636", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 6.126, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.6-plus/26440910-3d66-43a9-bddd-1f8cd89ffd37.json b/data/vals_ai/alibaba/qwen3.6-plus/26440910-3d66-43a9-bddd-1f8cd89ffd37.json new file mode 100644 index 0000000000000000000000000000000000000000..27045a565248cec00399f94b92e83c7ecdd95aaf --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.6-plus/26440910-3d66-43a9-bddd-1f8cd89ffd37.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/programbench/alibaba_qwen3.6-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProgramBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.6-plus", + "id": "alibaba/qwen3.6-plus", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.6-plus", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "programbench:almost:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.programbench.almost", + "source_data": { + "dataset_name": "ProgramBench - Almost Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "almost", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Almost Resolved).", + "metric_id": "vals_ai.programbench.almost.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 1.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "almost", + "task_name": "Almost Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "1", + "raw_stderr": "0.705", + "latency": "1981.11", + "cost_per_test": "4.018541", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.705, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "programbench:overall:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.programbench.overall", + "source_data": { + "dataset_name": "ProgramBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Overall).", + "metric_id": "vals_ai.programbench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1981.11", + "cost_per_test": "4.018541", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "programbench:partial:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.programbench.partial", + "source_data": { + "dataset_name": "ProgramBench - Raw Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "partial", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Raw Pass Rate).", + "metric_id": "vals_ai.programbench.partial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 28.561, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "partial", + "task_name": "Raw Pass Rate", + "dataset_type": "public", + "industry": "coding", + "raw_score": "28.561", + "raw_stderr": "1.974", + "latency": "1981.11", + "cost_per_test": "4.018541", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.974, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "programbench:strict:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.programbench.strict", + "source_data": { + "dataset_name": "ProgramBench - Fully Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "strict", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Fully Resolved).", + "metric_id": "vals_ai.programbench.strict.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "strict", + "task_name": "Fully Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1981.11", + "cost_per_test": "4.018541", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.6-plus/3f07b62c-5858-45ac-ab44-ee313cf21ead.json b/data/vals_ai/alibaba/qwen3.6-plus/3f07b62c-5858-45ac-ab44-ee313cf21ead.json new file mode 100644 index 0000000000000000000000000000000000000000..280d0c80bafc93069658cfe18d990ce2c8ef0601 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.6-plus/3f07b62c-5858-45ac-ab44-ee313cf21ead.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/code-migration/alibaba_qwen3.6-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Code Migration", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.6-plus", + "id": "alibaba/qwen3.6-plus", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.6-plus", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "code-migration:cli:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.code-migration.cli", + "source_data": { + "dataset_name": "Code Migration - CLI", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cli", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (CLI).", + "metric_id": "vals_ai.code-migration.cli.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 14.386, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cli", + "task_name": "CLI", + "dataset_type": "private", + "industry": "coding", + "raw_score": "14.386", + "raw_stderr": "1.7", + "latency": "1925.601", + "cost_per_test": "5.22785", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.7, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "code-migration:cobol:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.code-migration.cobol", + "source_data": { + "dataset_name": "Code Migration - COBOL", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cobol", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (COBOL).", + "metric_id": "vals_ai.code-migration.cobol.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 1.25, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cobol", + "task_name": "COBOL", + "dataset_type": "private", + "industry": "coding", + "raw_score": "1.25", + "raw_stderr": "1.25", + "latency": "1383.042", + "cost_per_test": "3.443127", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.25, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "code-migration:code_quality:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.code-migration.code_quality", + "source_data": { + "dataset_name": "Code Migration - Code Quality", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "code_quality", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Code Quality).", + "metric_id": "vals_ai.code-migration.code_quality.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 3.518, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "code_quality", + "task_name": "Code Quality", + "dataset_type": "private", + "industry": "coding", + "raw_score": "3.518", + "raw_stderr": "0.075", + "latency": "1789.961", + "cost_per_test": "4.781669", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.075, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "code-migration:overall:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.code-migration.overall", + "source_data": { + "dataset_name": "Code Migration - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Overall).", + "metric_id": "vals_ai.code-migration.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 11.102, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "11.102", + "raw_stderr": "1.312", + "latency": "1789.961", + "cost_per_test": "4.781669", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.312, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.6-plus/4efeba03-510c-43e5-aefe-a884c97c068d.json b/data/vals_ai/alibaba/qwen3.6-plus/4efeba03-510c-43e5-aefe-a884c97c068d.json new file mode 100644 index 0000000000000000000000000000000000000000..ad97c322bd561e057b21b817836f635cbb5a850f --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.6-plus/4efeba03-510c-43e5-aefe-a884c97c068d.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/alibaba_qwen3.6-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.6-plus", + "id": "alibaba/qwen3.6-plus", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.6-plus", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 76.963, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "76.963", + "raw_stderr": "1.917", + "latency": "173.545", + "cost_per_test": "0.029294", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.917, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.6-plus/67aa57e4-f76f-4cdf-952b-b49c79b2436b.json b/data/vals_ai/alibaba/qwen3.6-plus/67aa57e4-f76f-4cdf-952b-b49c79b2436b.json new file mode 100644 index 0000000000000000000000000000000000000000..ca593b80ff149ed5a26cda7a8d25e0f343fedf0b --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.6-plus/67aa57e4-f76f-4cdf-952b-b49c79b2436b.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/alibaba_qwen3.6-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.6-plus", + "id": "alibaba/qwen3.6-plus", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.6-plus", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 64.186, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.186", + "raw_stderr": "1.371", + "latency": "70.634", + "cost_per_test": "0.01145", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.371, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 74.734, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "74.734", + "raw_stderr": "0.852", + "latency": "70.634", + "cost_per_test": "0.01145", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.852, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 85.282, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "85.282", + "raw_stderr": "1.013", + "latency": "70.634", + "cost_per_test": "0.01145", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.013, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.6-plus/73e5b9e7-7e3f-4368-8583-fc974d57efb8.json b/data/vals_ai/alibaba/qwen3.6-plus/73e5b9e7-7e3f-4368-8583-fc974d57efb8.json new file mode 100644 index 0000000000000000000000000000000000000000..21987c28571bfed769d676a86d62f07382431152 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.6-plus/73e5b9e7-7e3f-4368-8583-fc974d57efb8.json @@ -0,0 +1,346 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/alibaba_qwen3.6-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.6-plus", + "id": "alibaba/qwen3.6-plus", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.6-plus", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 38.095, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "38.095", + "raw_stderr": "6.103", + "latency": "234.94", + "cost_per_test": "0.041626", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 6.103, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 36.508, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "36.508", + "raw_stderr": "6.182", + "latency": "247.912", + "cost_per_test": "0.037724", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 6.182, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 46.605, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "46.605", + "raw_stderr": "6.51", + "latency": "147.142", + "cost_per_test": "0.056504", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 6.51, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 59.977, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "59.977", + "raw_stderr": "5.53", + "latency": "204.949", + "cost_per_test": "0.044938", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 5.53, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 44.86, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "44.86", + "raw_stderr": "3.433", + "latency": "229.267", + "cost_per_test": "0.041429", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 3.433, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.6-plus/85cf5af1-b7d3-4f35-8c78-6454e6956cdd.json b/data/vals_ai/alibaba/qwen3.6-plus/85cf5af1-b7d3-4f35-8c78-6454e6956cdd.json new file mode 100644 index 0000000000000000000000000000000000000000..e8990740cf77f7904d8bb74794537581857a5466 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.6-plus/85cf5af1-b7d3-4f35-8c78-6454e6956cdd.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/alibaba_qwen3.6-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.6-plus", + "id": "alibaba/qwen3.6-plus", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.6-plus", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 78.14, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "78.14", + "raw_stderr": "1.165", + "latency": "106.003", + "cost_per_test": "0.017628", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.165, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 67.965, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.965", + "raw_stderr": "0.908", + "latency": "78.18", + "cost_per_test": "0.013476", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.908, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 57.79, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.79", + "raw_stderr": "1.392", + "latency": "50.358", + "cost_per_test": "0.009323", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.392, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.6-plus/866a03f6-0426-4477-877e-a8d2123484a0.json b/data/vals_ai/alibaba/qwen3.6-plus/866a03f6-0426-4477-877e-a8d2123484a0.json new file mode 100644 index 0000000000000000000000000000000000000000..21de68e74597458fa9defd0b55d1b8f3aae0fe76 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.6-plus/866a03f6-0426-4477-877e-a8d2123484a0.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/alibaba_qwen3.6-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.6-plus", + "id": "alibaba/qwen3.6-plus", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.6-plus", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 25.565, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "25.565", + "raw_stderr": "3.906", + "latency": "2289.977", + "cost_per_test": "5.447636", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 3.906, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.6-plus/901e9eb1-c466-485b-8e01-372e6c41329f.json b/data/vals_ai/alibaba/qwen3.6-plus/901e9eb1-c466-485b-8e01-372e6c41329f.json new file mode 100644 index 0000000000000000000000000000000000000000..84e3e3c71d6b3b8ac322315dae82ca74d23e16a6 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.6-plus/901e9eb1-c466-485b-8e01-372e6c41329f.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/alibaba_qwen3.6-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.6-plus", + "id": "alibaba/qwen3.6-plus", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.6-plus", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 75.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "75", + "raw_stderr": "0", + "latency": "17312.922", + "cost_per_test": "4.49233", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 31.111, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "31.111", + "raw_stderr": "1.111", + "latency": "2308.39", + "cost_per_test": "0.598977", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.111, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 63.636, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "63.636", + "raw_stderr": "2.099", + "latency": "1259.122", + "cost_per_test": "0.326715", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.099, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 53.184, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "53.184", + "raw_stderr": "1.633", + "latency": "778.109", + "cost_per_test": "0.201902", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.633, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.6-plus/aaea5cdf-75bc-4bc3-b6d7-5387cc2bf331.json b/data/vals_ai/alibaba/qwen3.6-plus/aaea5cdf-75bc-4bc3-b6d7-5387cc2bf331.json new file mode 100644 index 0000000000000000000000000000000000000000..a5b514554987e2b8a12f94bb95f4d9dd0e9ac3c5 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.6-plus/aaea5cdf-75bc-4bc3-b6d7-5387cc2bf331.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/alibaba_qwen3.6-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.6-plus", + "id": "alibaba/qwen3.6-plus", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.6-plus", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.606, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.606", + "raw_stderr": "1.668", + "latency": "28.533", + "cost_per_test": "0.00541", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.668, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 62.937, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.937", + "raw_stderr": "1.649", + "latency": "43.066", + "cost_per_test": "0.097946", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.649, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.927, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.927", + "raw_stderr": "0.957", + "latency": "35.839", + "cost_per_test": "0.047203", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.957, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 62.238, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.238", + "raw_stderr": "1.655", + "latency": "35.919", + "cost_per_test": "0.038254", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.655, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.6-plus/ac8f3ab4-a83c-46f7-848c-28771c44a1db.json b/data/vals_ai/alibaba/qwen3.6-plus/ac8f3ab4-a83c-46f7-848c-28771c44a1db.json new file mode 100644 index 0000000000000000000000000000000000000000..a717a036d510488ddb7400c568f124feaa22f242 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.6-plus/ac8f3ab4-a83c-46f7-848c-28771c44a1db.json @@ -0,0 +1,101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/alibaba_qwen3.6-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.6-plus", + "id": "alibaba/qwen3.6-plus", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.6-plus", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 84.162, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.162", + "raw_stderr": "0.878", + "latency": "111.943", + "cost_per_test": "0.017413", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.878, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.6-plus/b49e41fa-65c6-447f-8e5a-9974322969db.json b/data/vals_ai/alibaba/qwen3.6-plus/b49e41fa-65c6-447f-8e5a-9974322969db.json new file mode 100644 index 0000000000000000000000000000000000000000..a4071c6bb37e489e17629e2c9c4abf46a6a3ac35 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.6-plus/b49e41fa-65c6-447f-8e5a-9974322969db.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/alibaba_qwen3.6-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.6-plus", + "id": "alibaba/qwen3.6-plus", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.6-plus", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 88.384, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.384", + "raw_stderr": "2.277", + "latency": "98.941", + "cost_per_test": "0.017758", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.277, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 86.364, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.364", + "raw_stderr": "2.439", + "latency": "145.221", + "cost_per_test": "0.02817", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.439, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 87.374, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.374", + "raw_stderr": "1.668", + "latency": "122.081", + "cost_per_test": "0.022964", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.668, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.6-plus/bb5c97c0-208a-44d6-b043-9654506276fe.json b/data/vals_ai/alibaba/qwen3.6-plus/bb5c97c0-208a-44d6-b043-9654506276fe.json new file mode 100644 index 0000000000000000000000000000000000000000..ab55be394c867b60b4fe5b870baa3b05c692c5fc --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.6-plus/bb5c97c0-208a-44d6-b043-9654506276fe.json @@ -0,0 +1,426 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/alibaba_qwen3.6-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.6-plus", + "id": "alibaba/qwen3.6-plus", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.6-plus", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 89.177, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.177", + "raw_stderr": "0.669", + "latency": "20.039", + "cost_per_test": "0.003064", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.669, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.006, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.006", + "raw_stderr": "0.211", + "latency": "17.539", + "cost_per_test": "0.002808", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.211, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 89.727, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.727", + "raw_stderr": "0.567", + "latency": "13.911", + "cost_per_test": "0.002679", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.567, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 84.233, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.233", + "raw_stderr": "0.422", + "latency": "19.073", + "cost_per_test": "0.003042", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.422, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.033, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.033", + "raw_stderr": "0.878", + "latency": "16.284", + "cost_per_test": "0.002717", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.878, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.223, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.223", + "raw_stderr": "1.694", + "latency": "27.594", + "cost_per_test": "0.00394", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.694, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.6-plus/bcf3dda2-cb8e-4ff7-a123-50a1e1e018a9.json b/data/vals_ai/alibaba/qwen3.6-plus/bcf3dda2-cb8e-4ff7-a123-50a1e1e018a9.json new file mode 100644 index 0000000000000000000000000000000000000000..c86b6dbd561eb1d6c1c84e0c92cb98aef052ca9c --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.6-plus/bcf3dda2-cb8e-4ff7-a123-50a1e1e018a9.json @@ -0,0 +1,361 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/alibaba_qwen3.6-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.6-plus", + "id": "alibaba/qwen3.6-plus", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.6-plus", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 38.095, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "38.095", + "raw_stderr": "7.584", + "latency": "681.011", + "cost_per_test": "1.668146", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 7.584, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 70.881, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "70.881", + "raw_stderr": "2.818", + "latency": "461.254", + "cost_per_test": "0.977144", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.818, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 85.567, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "85.567", + "raw_stderr": "2.53", + "latency": "315.965", + "cost_per_test": "0.545146", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.53, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "599.614", + "cost_per_test": "1.250232", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 73.4, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "73.4", + "raw_stderr": "1.978", + "latency": "424.172", + "cost_per_test": "0.869211", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.978, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.6-plus/ca348305-11a9-4db7-b1ae-1855f271ea28.json b/data/vals_ai/alibaba/qwen3.6-plus/ca348305-11a9-4db7-b1ae-1855f271ea28.json new file mode 100644 index 0000000000000000000000000000000000000000..11d6872358c93af74a457dfe441586872b599ffb --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.6-plus/ca348305-11a9-4db7-b1ae-1855f271ea28.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/alibaba_qwen3.6-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.6-plus", + "id": "alibaba/qwen3.6-plus", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.6-plus", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 99.068, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "99.068", + "raw_stderr": "0.535", + "latency": "205.107", + "cost_per_test": "0.03336", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 67.143, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "67.143", + "raw_stderr": "2.511", + "latency": "764.464", + "cost_per_test": "0.124575", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.511, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 91.645, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "91.645", + "raw_stderr": "1.414", + "latency": "457.545", + "cost_per_test": "0.074949", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.414, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 85.952, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "85.952", + "raw_stderr": "0.977", + "latency": "475.705", + "cost_per_test": "0.077628", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.977, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.6-plus/cbde2eaf-a36c-4a16-bcda-520ee04e8500.json b/data/vals_ai/alibaba/qwen3.6-plus/cbde2eaf-a36c-4a16-bcda-520ee04e8500.json new file mode 100644 index 0000000000000000000000000000000000000000..d165b21f4998fe5c25bff4b4a1dcae2b3f8a1620 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.6-plus/cbde2eaf-a36c-4a16-bcda-520ee04e8500.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/alibaba_qwen3.6-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.6-plus", + "id": "alibaba/qwen3.6-plus", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.6-plus", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.724, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.724", + "raw_stderr": "0.906", + "latency": "35.767", + "cost_per_test": "0.007006", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.906, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.269, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.269", + "raw_stderr": "0.951", + "latency": "45.785", + "cost_per_test": "0.008359", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.951, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.431, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.431", + "raw_stderr": "0.832", + "latency": "66.874", + "cost_per_test": "0.013391", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.832, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.756, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.756", + "raw_stderr": "1.498", + "latency": "44.734", + "cost_per_test": "0.008378", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.498, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.692, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.692", + "raw_stderr": "1.047", + "latency": "35.519", + "cost_per_test": "0.006698", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.047, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.314, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.314", + "raw_stderr": "1.168", + "latency": "82.131", + "cost_per_test": "0.016678", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.168, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.807, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.807", + "raw_stderr": "1.377", + "latency": "35.864", + "cost_per_test": "0.00687", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.377, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.102, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.102", + "raw_stderr": "2.006", + "latency": "38.219", + "cost_per_test": "0.007206", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.006, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.658, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.658", + "raw_stderr": "1.293", + "latency": "49.942", + "cost_per_test": "0.009164", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.293, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 95.485, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "95.485", + "raw_stderr": "0.565", + "latency": "50.573", + "cost_per_test": "0.00986", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.565, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.714, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.714", + "raw_stderr": "1.151", + "latency": "34.123", + "cost_per_test": "0.00613", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.151, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.668, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.668", + "raw_stderr": "0.323", + "latency": "46.184", + "cost_per_test": "0.008813", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.323, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.174, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.174", + "raw_stderr": "1.497", + "latency": "38.14", + "cost_per_test": "0.006638", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.497, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.764, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.764", + "raw_stderr": "0.719", + "latency": "55.516", + "cost_per_test": "0.010983", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.719, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.469, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.469", + "raw_stderr": "1.172", + "latency": "33.387", + "cost_per_test": "0.006021", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.172, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.6-plus/e11e6126-e7bf-45e1-b05b-c138b7627c9e.json b/data/vals_ai/alibaba/qwen3.6-plus/e11e6126-e7bf-45e1-b05b-c138b7627c9e.json new file mode 100644 index 0000000000000000000000000000000000000000..ce50209f2ba2ed2d28971a2bf840389f506a4720 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.6-plus/e11e6126-e7bf-45e1-b05b-c138b7627c9e.json @@ -0,0 +1,718 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/alibaba_qwen3.6-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.6-plus", + "id": "alibaba/qwen3.6-plus", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.6-plus", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 35.673, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "35.673", + "raw_stderr": "2.924", + "latency": "674.622", + "cost_per_test": "0.666331", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.924, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 29.35, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "29.35", + "raw_stderr": "0.67", + "latency": "611.907", + "cost_per_test": "0.814379", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.67, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 43.683, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "43.683", + "raw_stderr": "0.904", + "latency": "734.01", + "cost_per_test": "0.937845", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.904, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 44.547, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "44.547", + "raw_stderr": "2.264", + "latency": "614.983", + "cost_per_test": "0.572353", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.264, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 52.398, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.398", + "raw_stderr": "0.998", + "latency": "440.575", + "cost_per_test": "0.344742", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.998, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 16.728, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "16.728", + "raw_stderr": "1.148", + "latency": "762.437", + "cost_per_test": "0.961778", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.148, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 51.94, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "51.94", + "raw_stderr": "1.346", + "latency": "608.006", + "cost_per_test": "1.906543", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.346, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 58.181, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.181", + "raw_stderr": "2.076", + "latency": "520.968", + "cost_per_test": "0.559083", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.076, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 47.324, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "47.324", + "raw_stderr": "1.192", + "latency": "355.663", + "cost_per_test": "0.373151", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.192, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:overall:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 40.846, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "40.846", + "raw_stderr": "0.131", + "latency": "611.907", + "cost_per_test": "0.814379", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.131, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:alibaba/qwen3.6-plus:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 17.137, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "17.137", + "raw_stderr": "1.22", + "latency": "821.432", + "cost_per_test": "1.054446", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.22, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.7-max/0dd3c99e-f67d-4b98-b205-6ab58b963976.json b/data/vals_ai/alibaba/qwen3.7-max/0dd3c99e-f67d-4b98-b205-6ab58b963976.json new file mode 100644 index 0000000000000000000000000000000000000000..9524e1ded239fe6a698b6db18adb2a6226651b6e --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.7-max/0dd3c99e-f67d-4b98-b205-6ab58b963976.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/alibaba_qwen3.7-max/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.7-max", + "id": "alibaba/qwen3.7-max", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.7-max", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 40.5, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "40.5", + "raw_stderr": "17.613", + "latency": "4033.028", + "cost_per_test": "4.69984", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 17.613, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 53.0, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "53", + "raw_stderr": "10.541", + "latency": "4727.561", + "cost_per_test": "5.019455", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 10.541, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 46.75, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "46.75", + "raw_stderr": "10.263", + "latency": "4380.294", + "cost_per_test": "4.859648", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 10.263, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.7-max/119024ea-4aeb-4039-a8e8-97809076ffb3.json b/data/vals_ai/alibaba/qwen3.7-max/119024ea-4aeb-4039-a8e8-97809076ffb3.json new file mode 100644 index 0000000000000000000000000000000000000000..9d36af4e304380ea1123addef61c8a3340232f23 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.7-max/119024ea-4aeb-4039-a8e8-97809076ffb3.json @@ -0,0 +1,1152 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_research/alibaba_qwen3.7-max/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Legal Research Bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.7-max", + "id": "alibaba/qwen3.7-max", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.7-max", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_research:administrative_regulatory:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.legal_research.administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 36.364, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "36.364", + "raw_stderr": "5.967", + "latency": "789.443", + "cost_per_test": "1.359392", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 5.967, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:business_commercial:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.legal_research.business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 20.93, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "20.93", + "raw_stderr": "6.277", + "latency": "913.289", + "cost_per_test": "1.493846", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 6.277, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:civil_litigation:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.legal_research.civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 25.926, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "25.926", + "raw_stderr": "8.594", + "latency": "876.732", + "cost_per_test": "1.504309", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 8.594, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:constitutional_civil_rights:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.legal_research.constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 23.684, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "23.684", + "raw_stderr": "6.989", + "latency": "782.498", + "cost_per_test": "1.136745", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 6.989, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:criminal:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.legal_research.criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 21.053, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "21.053", + "raw_stderr": "5.448", + "latency": "787.984", + "cost_per_test": "1.332257", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 5.448, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:family:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.legal_research.family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 9.091, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "9.091", + "raw_stderr": "6.273", + "latency": "900.57", + "cost_per_test": "1.420719", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 6.273, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:health:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.legal_research.health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 40.0, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "40", + "raw_stderr": "9.097", + "latency": "848.765", + "cost_per_test": "1.786099", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 9.097, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:immigration:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.legal_research.immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 18.182, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "18.182", + "raw_stderr": "12.197", + "latency": "705.891", + "cost_per_test": "1.495632", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 12.197, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:overall:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.legal_research.overall", + "source_data": { + "dataset_name": "Legal Research Bench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Overall).", + "metric_id": "vals_ai.legal_research.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 25.481, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "legal", + "raw_score": "25.481", + "raw_stderr": "3.029", + "latency": "808.802", + "cost_per_test": "1.400701", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 3.029, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass", + "source_data": { + "dataset_name": "Legal Research Bench - Weighted Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Weighted Pass).", + "metric_id": "vals_ai.legal_research.weighted_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 78.4, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass", + "task_name": "Weighted Pass", + "dataset_type": "private", + "industry": "legal", + "raw_score": "78.4", + "raw_stderr": "1.689", + "latency": "808.802", + "cost_per_test": "1.400701", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.689, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__administrative_regulatory:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.weighted_pass__administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 83.971, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "83.971", + "raw_stderr": "2.671", + "latency": "789.443", + "cost_per_test": "1.359392", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.671, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__business_commercial:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.weighted_pass__business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 80.966, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "80.966", + "raw_stderr": "3.37", + "latency": "913.289", + "cost_per_test": "1.493846", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 3.37, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__civil_litigation:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.weighted_pass__civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 84.566, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "84.566", + "raw_stderr": "3.007", + "latency": "876.732", + "cost_per_test": "1.504309", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 3.007, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__constitutional_civil_rights:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 82.248, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "82.248", + "raw_stderr": "3.483", + "latency": "782.498", + "cost_per_test": "1.136745", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 3.483, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__criminal:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.weighted_pass__criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 77.693, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "77.693", + "raw_stderr": "2.824", + "latency": "787.984", + "cost_per_test": "1.332257", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.824, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__family:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.weighted_pass__family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 59.268, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "59.268", + "raw_stderr": "6.671", + "latency": "900.57", + "cost_per_test": "1.420719", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 6.671, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__health:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.weighted_pass__health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 86.658, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "86.658", + "raw_stderr": "3.085", + "latency": "848.765", + "cost_per_test": "1.786099", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 3.085, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__immigration:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.weighted_pass__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 73.271, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "73.271", + "raw_stderr": "9.271", + "latency": "705.891", + "cost_per_test": "1.495632", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 9.271, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.7-max/55166557-2b9d-4014-b4a7-a67b1358cae8.json b/data/vals_ai/alibaba/qwen3.7-max/55166557-2b9d-4014-b4a7-a67b1358cae8.json new file mode 100644 index 0000000000000000000000000000000000000000..d23f79f5a45ccdaa492c59e40a3029c2817adf56 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.7-max/55166557-2b9d-4014-b4a7-a67b1358cae8.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/alibaba_qwen3.7-max/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.7-max", + "id": "alibaba/qwen3.7-max", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.7-max", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.447, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.447", + "raw_stderr": "0.689", + "latency": "21.851", + "cost_per_test": "0.008895", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.689, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 73.429, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "73.429", + "raw_stderr": "2.361", + "latency": "439.27", + "cost_per_test": "0.20172", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.361, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 89.295, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "89.295", + "raw_stderr": "1.58", + "latency": "115.405", + "cost_per_test": "0.049639", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.58, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 87.057, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "87.057", + "raw_stderr": "0.974", + "latency": "192.175", + "cost_per_test": "0.086751", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.974, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.7-max/6622a04c-0db8-471d-b72e-3eaebd528736.json b/data/vals_ai/alibaba/qwen3.7-max/6622a04c-0db8-471d-b72e-3eaebd528736.json new file mode 100644 index 0000000000000000000000000000000000000000..927df11adcc23047671a0cf3a26c12858e80ac8b --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.7-max/6622a04c-0db8-471d-b72e-3eaebd528736.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/alibaba_qwen3.7-max/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.7-max", + "id": "alibaba/qwen3.7-max", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.7-max", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 95.119, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "95.119", + "raw_stderr": "0.805", + "latency": "21.48", + "cost_per_test": "0.012963", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.805, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.536, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.536", + "raw_stderr": "0.875", + "latency": "25.757", + "cost_per_test": "0.014848", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.875, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.314, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.314", + "raw_stderr": "0.792", + "latency": "40.784", + "cost_per_test": "0.024737", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.792, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.22, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.22", + "raw_stderr": "1.398", + "latency": "28.331", + "cost_per_test": "0.016489", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.398, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.706, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.706", + "raw_stderr": "0.949", + "latency": "20.211", + "cost_per_test": "0.011905", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.949, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.759, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.759", + "raw_stderr": "1.123", + "latency": "61.801", + "cost_per_test": "0.035533", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.123, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.352, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.352", + "raw_stderr": "1.27", + "latency": "25.191", + "cost_per_test": "0.013281", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.27, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.152, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.152", + "raw_stderr": "1.962", + "latency": "24.783", + "cost_per_test": "0.015035", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.962, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.473, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.473", + "raw_stderr": "1.217", + "latency": "32.261", + "cost_per_test": "0.017397", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.217, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 96.003, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "96.003", + "raw_stderr": "0.533", + "latency": "29.213", + "cost_per_test": "0.019017", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.533, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.039, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.039", + "raw_stderr": "1.14", + "latency": "19.902", + "cost_per_test": "0.010569", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.14, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.311, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.311", + "raw_stderr": "0.305", + "latency": "29.15", + "cost_per_test": "0.01678", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.305, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.178, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.178", + "raw_stderr": "1.391", + "latency": "21.703", + "cost_per_test": "0.01142", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.391, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.533, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.533", + "raw_stderr": "0.682", + "latency": "37.31", + "cost_per_test": "0.020998", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.682, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.975, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.975", + "raw_stderr": "1.063", + "latency": "19.369", + "cost_per_test": "0.010724", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.063, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.7-max/6b1594ca-4b69-449f-bfee-ca4730b15b6f.json b/data/vals_ai/alibaba/qwen3.7-max/6b1594ca-4b69-449f-bfee-ca4730b15b6f.json new file mode 100644 index 0000000000000000000000000000000000000000..acc7eecd8a8387facdcd5299c022bdedce018479 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.7-max/6b1594ca-4b69-449f-bfee-ca4730b15b6f.json @@ -0,0 +1,426 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/alibaba_qwen3.7-max/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.7-max", + "id": "alibaba/qwen3.7-max", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.7-max", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 89.763, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.763", + "raw_stderr": "0.639", + "latency": "11.735", + "cost_per_test": "0.005228", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.639, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.268, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.268", + "raw_stderr": "0.232", + "latency": "8.421", + "cost_per_test": "0.004529", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.232, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 89.731, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.731", + "raw_stderr": "0.571", + "latency": "7.485", + "cost_per_test": "0.005988", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.571, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 84.913, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.913", + "raw_stderr": "0.413", + "latency": "11.853", + "cost_per_test": "0.005967", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.413, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.65, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.65", + "raw_stderr": "0.87", + "latency": "9.665", + "cost_per_test": "0.005424", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.87, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.154, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.154", + "raw_stderr": "1.646", + "latency": "21.957", + "cost_per_test": "0.008665", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.646, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.7-max/6d19cf5b-719c-40f3-93eb-75fd91faeee7.json b/data/vals_ai/alibaba/qwen3.7-max/6d19cf5b-719c-40f3-93eb-75fd91faeee7.json new file mode 100644 index 0000000000000000000000000000000000000000..f34c48f8b42790248c7e9fa59814bf05b3f6be51 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.7-max/6d19cf5b-719c-40f3-93eb-75fd91faeee7.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/alibaba_qwen3.7-max/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.7-max", + "id": "alibaba/qwen3.7-max", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.7-max", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 79.396, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "79.396", + "raw_stderr": "1.907", + "latency": "108.046", + "cost_per_test": "0.06907", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.907, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.7-max/8125b35a-5cc8-475c-85cf-c9aade755244.json b/data/vals_ai/alibaba/qwen3.7-max/8125b35a-5cc8-475c-85cf-c9aade755244.json new file mode 100644 index 0000000000000000000000000000000000000000..a31c095ca366b6939598a8f782415ca0814064de --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.7-max/8125b35a-5cc8-475c-85cf-c9aade755244.json @@ -0,0 +1,361 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/alibaba_qwen3.7-max/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.7-max", + "id": "alibaba/qwen3.7-max", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.7-max", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 38.095, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "38.095", + "raw_stderr": "7.584", + "latency": "1199.386", + "cost_per_test": "4.340344", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 7.584, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 64.368, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "64.368", + "raw_stderr": "2.97", + "latency": "1002.811", + "cost_per_test": "3.605562", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.97, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 82.474, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "82.474", + "raw_stderr": "2.737", + "latency": "887.884", + "cost_per_test": "2.450485", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.737, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1236.1", + "cost_per_test": "4.954436", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 68.8, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "68.8", + "raw_stderr": "2.074", + "latency": "976.131", + "cost_per_test": "3.227207", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.074, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.7-max/8b08b0a7-4a8d-49d2-bc4f-84ce3b87abea.json b/data/vals_ai/alibaba/qwen3.7-max/8b08b0a7-4a8d-49d2-bc4f-84ce3b87abea.json new file mode 100644 index 0000000000000000000000000000000000000000..575ff1b86dd74de13f9feba2afdf6fbb22cafbd1 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.7-max/8b08b0a7-4a8d-49d2-bc4f-84ce3b87abea.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/alibaba_qwen3.7-max/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.7-max", + "id": "alibaba/qwen3.7-max", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.7-max", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.023, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.023", + "raw_stderr": "1.672", + "latency": "15.358", + "cost_per_test": "0.01138", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.672, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 65.734, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.734", + "raw_stderr": "1.62", + "latency": "26.109", + "cost_per_test": "0.473648", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.62, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 63.714, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.714", + "raw_stderr": "0.946", + "latency": "19.934", + "cost_per_test": "0.219883", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.946, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 65.385, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.385", + "raw_stderr": "1.624", + "latency": "18.336", + "cost_per_test": "0.17462", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.624, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.7-max/90526f71-fdc2-4c0a-813f-dfb2a5ecd0da.json b/data/vals_ai/alibaba/qwen3.7-max/90526f71-fdc2-4c0a-813f-dfb2a5ecd0da.json new file mode 100644 index 0000000000000000000000000000000000000000..b7c02a521ac6ed082203a2b157d5e232c804b184 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.7-max/90526f71-fdc2-4c0a-813f-dfb2a5ecd0da.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/alibaba_qwen3.7-max/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.7-max", + "id": "alibaba/qwen3.7-max", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.7-max", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 26.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "26", + "raw_stderr": "4.408", + "latency": "771.408", + "cost_per_test": "2.462069", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 4.408, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.7-max/a5785840-a6b3-478e-b6fb-425bc7e5fcad.json b/data/vals_ai/alibaba/qwen3.7-max/a5785840-a6b3-478e-b6fb-425bc7e5fcad.json new file mode 100644 index 0000000000000000000000000000000000000000..e2ed89bf0342844c7f50f6a38c115cedea845007 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.7-max/a5785840-a6b3-478e-b6fb-425bc7e5fcad.json @@ -0,0 +1,398 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/alibaba_qwen3.7-max/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.7-max", + "id": "alibaba/qwen3.7-max", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.7-max", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 65.385, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "65.385", + "raw_stderr": "1.624", + "latency": "18.336", + "cost_per_test": "0.17462", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.624, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 48.353, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "48.353", + "raw_stderr": "0.477", + "latency": "483.912", + "cost_per_test": "1.971421", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.477, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:overall:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 57.487, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "57.487", + "raw_stderr": "1.572", + "latency": "1831.709", + "cost_per_test": "5.223348", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.572, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "66.667", + "raw_stderr": "4.668", + "latency": "1037.61", + "cost_per_test": "3.384593" + }, + "uncertainty": { + "standard_error": { + "value": 4.668, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 61.049, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "61.049", + "raw_stderr": "0.375", + "latency": "935.196", + "cost_per_test": "0.33233", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.375, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 52.88, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "52.88", + "raw_stderr": "6.854", + "latency": "6683.491", + "cost_per_test": "20.253775", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 6.854, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.7-max/ae513ea3-80a4-4795-bdb2-25b063e10ebb.json b/data/vals_ai/alibaba/qwen3.7-max/ae513ea3-80a4-4795-bdb2-25b063e10ebb.json new file mode 100644 index 0000000000000000000000000000000000000000..d602894dd7590498f1217a608266564d17575af8 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.7-max/ae513ea3-80a4-4795-bdb2-25b063e10ebb.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/alibaba_qwen3.7-max/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.7-max", + "id": "alibaba/qwen3.7-max", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.7-max", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 63.532, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.532", + "raw_stderr": "1.376", + "latency": "57.677", + "cost_per_test": "0.028122", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.376, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 75.306, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "75.306", + "raw_stderr": "0.839", + "latency": "57.677", + "cost_per_test": "0.028122", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.839, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 87.081, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "87.081", + "raw_stderr": "0.959", + "latency": "57.677", + "cost_per_test": "0.028122", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.959, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.7-max/af716465-bce6-497b-a24e-b7ea986748be.json b/data/vals_ai/alibaba/qwen3.7-max/af716465-bce6-497b-a24e-b7ea986748be.json new file mode 100644 index 0000000000000000000000000000000000000000..0976594ad782be4a213b55839e0b7081c68408e8 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.7-max/af716465-bce6-497b-a24e-b7ea986748be.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/alibaba_qwen3.7-max/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.7-max", + "id": "alibaba/qwen3.7-max", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.7-max", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 89.899, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.899", + "raw_stderr": "2.142", + "latency": "82.709", + "cost_per_test": "0.044792", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.142, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 90.404, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.404", + "raw_stderr": "2.093", + "latency": "170.513", + "cost_per_test": "0.078189", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.093, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 90.152, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.152", + "raw_stderr": "1.497", + "latency": "126.611", + "cost_per_test": "0.06149", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.497, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.7-max/be4d7d33-9fe7-4af6-a0d0-1670ce466727.json b/data/vals_ai/alibaba/qwen3.7-max/be4d7d33-9fe7-4af6-a0d0-1670ce466727.json new file mode 100644 index 0000000000000000000000000000000000000000..37f841230e6304223362313d5751215726886a1e --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.7-max/be4d7d33-9fe7-4af6-a0d0-1670ce466727.json @@ -0,0 +1,656 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/emb/alibaba_qwen3.7-max/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - EMB", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.7-max", + "id": "alibaba/qwen3.7-max", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.7-max", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "emb:comps:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.emb.comps", + "source_data": { + "dataset_name": "EMB - Comps", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "comps", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Comps).", + "metric_id": "vals_ai.emb.comps.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 50.921, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "comps", + "task_name": "Comps", + "dataset_type": "private", + "industry": "finance", + "raw_score": "50.921", + "raw_stderr": "7.018", + "latency": "2323.758", + "cost_per_test": "3.397738", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 7.018, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:dataroom:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.emb.dataroom", + "source_data": { + "dataset_name": "EMB - Dataroom", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dataroom", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Dataroom).", + "metric_id": "vals_ai.emb.dataroom.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 80.699, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dataroom", + "task_name": "Dataroom", + "dataset_type": "private", + "industry": "finance", + "raw_score": "80.699", + "raw_stderr": "5.942", + "latency": "1359.288", + "cost_per_test": "1.776778", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 5.942, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:dcf:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.emb.dcf", + "source_data": { + "dataset_name": "EMB - DCF", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dcf", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (DCF).", + "metric_id": "vals_ai.emb.dcf.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 43.53, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dcf", + "task_name": "DCF", + "dataset_type": "private", + "industry": "finance", + "raw_score": "43.53", + "raw_stderr": "7.777", + "latency": "1771.106", + "cost_per_test": "2.896375", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 7.777, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:lbo:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.emb.lbo", + "source_data": { + "dataset_name": "EMB - LBO", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "lbo", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (LBO).", + "metric_id": "vals_ai.emb.lbo.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 47.18, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "lbo", + "task_name": "LBO", + "dataset_type": "private", + "industry": "finance", + "raw_score": "47.18", + "raw_stderr": "4.739", + "latency": "2382.054", + "cost_per_test": "4.717455", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 4.739, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:ma:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.emb.ma", + "source_data": { + "dataset_name": "EMB - M&A", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "ma", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (M&A).", + "metric_id": "vals_ai.emb.ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 67.104, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "ma", + "task_name": "M&A", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.104", + "raw_stderr": "6.819", + "latency": "1958.74", + "cost_per_test": "2.277121", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 6.819, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:misc:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.emb.misc", + "source_data": { + "dataset_name": "EMB - Misc", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "misc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Misc).", + "metric_id": "vals_ai.emb.misc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 57.746, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "misc", + "task_name": "Misc", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.746", + "raw_stderr": "6.394", + "latency": "1987.43", + "cost_per_test": "1.909493", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 6.394, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:operating:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.emb.operating", + "source_data": { + "dataset_name": "EMB - Operating", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "operating", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Operating).", + "metric_id": "vals_ai.emb.operating.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 52.105, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "operating", + "task_name": "Operating", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.105", + "raw_stderr": "8.48", + "latency": "1827.548", + "cost_per_test": "2.586275", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 8.48, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:overall:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.emb.overall", + "source_data": { + "dataset_name": "EMB - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Overall).", + "metric_id": "vals_ai.emb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 57.041, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.041", + "raw_stderr": "2.853", + "latency": "1957.165", + "cost_per_test": "2.857267", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.853, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:scratch:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.emb.scratch", + "source_data": { + "dataset_name": "EMB - Scratch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "scratch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Scratch).", + "metric_id": "vals_ai.emb.scratch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 61.447, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "scratch", + "task_name": "Scratch", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.447", + "raw_stderr": "2.75", + "latency": "1962.552", + "cost_per_test": "2.74412", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.75, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:template:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.emb.template", + "source_data": { + "dataset_name": "EMB - Template", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "template", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Template).", + "metric_id": "vals_ai.emb.template.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 51.787, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "template", + "task_name": "Template", + "dataset_type": "private", + "industry": "finance", + "raw_score": "51.787", + "raw_stderr": "4.788", + "latency": "1951.777", + "cost_per_test": "2.970414", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 4.788, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.7-max/d6270fea-ed00-40ee-9107-9a9f7aca9cf3.json b/data/vals_ai/alibaba/qwen3.7-max/d6270fea-ed00-40ee-9107-9a9f7aca9cf3.json new file mode 100644 index 0000000000000000000000000000000000000000..b92fa023e3fae45bdf87b252b5ae75304b1cb2de --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.7-max/d6270fea-ed00-40ee-9107-9a9f7aca9cf3.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/alibaba_qwen3.7-max/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.7-max", + "id": "alibaba/qwen3.7-max", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.7-max", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 83.333, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "83.333", + "raw_stderr": "8.333", + "latency": "20808.115", + "cost_per_test": "7.39434", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 8.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 44.444, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "44.444", + "raw_stderr": "1.111", + "latency": "2774.415", + "cost_per_test": "0.985912", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.111, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 68.485, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "68.485", + "raw_stderr": "1.212", + "latency": "1513.317", + "cost_per_test": "0.53777", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.212, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 61.049, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "61.049", + "raw_stderr": "0.375", + "latency": "935.196", + "cost_per_test": "0.33233", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.375, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.7-max/e93c4b08-6735-4bfd-b701-ef9d01dbf1dc.json b/data/vals_ai/alibaba/qwen3.7-max/e93c4b08-6735-4bfd-b701-ef9d01dbf1dc.json new file mode 100644 index 0000000000000000000000000000000000000000..f3f6fb28cad827a5ac2b5400ee74277e899ca6c7 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.7-max/e93c4b08-6735-4bfd-b701-ef9d01dbf1dc.json @@ -0,0 +1,3286 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/hlab/alibaba_qwen3.7-max/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Harvey's Legal Agent Benchmark", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.7-max", + "id": "alibaba/qwen3.7-max", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.7-max", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "hlab:antitrust_competition:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "2246.054", + "cost_per_test": "2.392755", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:arbitration_international_dispute_resolution:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1798.273", + "cost_per_test": "1.869712", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:banking_finance:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1633.278", + "cost_per_test": "1.681982", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:bankruptcy_restructuring:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "2252.381", + "cost_per_test": "2.436706", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:capital_markets:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1598.598", + "cost_per_test": "1.415374", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:corporate_governance:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1980.137", + "cost_per_test": "1.221399", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:corporate_ma:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1496.277", + "cost_per_test": "1.496984", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Criteria Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Criteria Pass Rate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 83.467, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate", + "task_name": "Criteria Pass Rate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.467", + "raw_stderr": "0.449", + "latency": "1923.268", + "cost_per_test": "1.57592", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.449, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__antitrust_competition:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 85.145, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.145", + "raw_stderr": "2.145", + "latency": "2246.054", + "cost_per_test": "2.392755", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.145, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__arbitration_international_dispute_resolution:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 78.853, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.853", + "raw_stderr": "2.449", + "latency": "1798.273", + "cost_per_test": "1.869712", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.449, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__banking_finance:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 85.874, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.874", + "raw_stderr": "2.128", + "latency": "1633.278", + "cost_per_test": "1.681982", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.128, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__bankruptcy_restructuring:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 78.201, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.201", + "raw_stderr": "2.433", + "latency": "2252.381", + "cost_per_test": "2.436706", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.433, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__capital_markets:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 84.706, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.706", + "raw_stderr": "2.258", + "latency": "1598.598", + "cost_per_test": "1.415374", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.258, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_governance:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 85.959, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.959", + "raw_stderr": "2.037", + "latency": "1980.137", + "cost_per_test": "1.221399", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.037, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_ma:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 88.782, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.782", + "raw_stderr": "1.79", + "latency": "1496.277", + "cost_per_test": "1.496984", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.79, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__data_privacy_cybersecurity:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 88.339, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.339", + "raw_stderr": "1.911", + "latency": "1847.052", + "cost_per_test": "1.368981", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.911, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__emerging_companies_venture_capital:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 85.324, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.324", + "raw_stderr": "2.071", + "latency": "1939.613", + "cost_per_test": "1.696531", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.071, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__employment_labor:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 76.779, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "76.779", + "raw_stderr": "2.589", + "latency": "1382.561", + "cost_per_test": "1.421157", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.589, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__energy_natural_resources:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 88.449, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.449", + "raw_stderr": "1.839", + "latency": "1484.992", + "cost_per_test": "1.559743", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.839, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__environmental_esg:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 81.004, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.004", + "raw_stderr": "2.353", + "latency": "1902.607", + "cost_per_test": "1.529383", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.353, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__funds_asset_management:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 80.758, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.758", + "raw_stderr": "2.132", + "latency": "2408.753", + "cost_per_test": "1.625214", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.132, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__healthcare_life_sciences:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 84.818, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.818", + "raw_stderr": "2.065", + "latency": "2182.825", + "cost_per_test": "1.715622", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.065, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__immigration:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 80.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80", + "raw_stderr": "2.416", + "latency": "2243.916", + "cost_per_test": "1.390074", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.416, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__insurance:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 82.051, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.051", + "raw_stderr": "2.327", + "latency": "2064.254", + "cost_per_test": "1.401873", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.327, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__intellectual_property:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 91.846, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "91.846", + "raw_stderr": "1.797", + "latency": "1483.879", + "cost_per_test": "1.757206", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.797, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__international_trade_sanctions:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 79.365, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.365", + "raw_stderr": "2.554", + "latency": "2245.603", + "cost_per_test": "1.091719", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.554, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__litigation_dispute_resolution:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 86.891, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.891", + "raw_stderr": "2.069", + "latency": "1498.176", + "cost_per_test": "1.41634", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.069, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__real_estate:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 90.088, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.088", + "raw_stderr": "1.616", + "latency": "1886.135", + "cost_per_test": "1.775065", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.616, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__structured_finance_securitization:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 79.365, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.365", + "raw_stderr": "2.284", + "latency": "1812.583", + "cost_per_test": "1.250902", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.284, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__tax:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 77.592, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.592", + "raw_stderr": "2.415", + "latency": "2473.173", + "cost_per_test": "1.620848", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.415, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__trusts_estates_private_client:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 82.061, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.061", + "raw_stderr": "2.375", + "latency": "2344.86", + "cost_per_test": "1.287155", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.375, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__white_collar_defense_investigations:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 80.808, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.808", + "raw_stderr": "2.289", + "latency": "1864.579", + "cost_per_test": "1.435625", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.289, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:data_privacy_cybersecurity:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1847.052", + "cost_per_test": "1.368981", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:emerging_companies_venture_capital:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1939.613", + "cost_per_test": "1.696531", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:employment_labor:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1382.561", + "cost_per_test": "1.421157", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:energy_natural_resources:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1484.992", + "cost_per_test": "1.559743", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:environmental_esg:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1902.607", + "cost_per_test": "1.529383", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:funds_asset_management:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "2408.753", + "cost_per_test": "1.625214", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:healthcare_life_sciences:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "2182.825", + "cost_per_test": "1.715622", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:immigration:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "2243.916", + "cost_per_test": "1.390074", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:insurance:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "2064.254", + "cost_per_test": "1.401873", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:intellectual_property:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1483.879", + "cost_per_test": "1.757206", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:international_trade_sanctions:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "2245.603", + "cost_per_test": "1.091719", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:litigation_dispute_resolution:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1498.176", + "cost_per_test": "1.41634", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:overall:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.overall", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Overall).", + "metric_id": "vals_ai.hlab.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 1.667, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "1.667", + "raw_stderr": "0.84", + "latency": "1923.268", + "cost_per_test": "1.57592", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.84, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:real_estate:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1886.135", + "cost_per_test": "1.775065", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:structured_finance_securitization:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1812.583", + "cost_per_test": "1.250902", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:tax:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "2473.173", + "cost_per_test": "1.620848", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:trusts_estates_private_client:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "2344.86", + "cost_per_test": "1.287155", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:white_collar_defense_investigations:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.hlab.white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1864.579", + "cost_per_test": "1.435625", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.7-max/f293fa1d-ea4d-4469-be2c-829711b53f4b.json b/data/vals_ai/alibaba/qwen3.7-max/f293fa1d-ea4d-4469-be2c-829711b53f4b.json new file mode 100644 index 0000000000000000000000000000000000000000..8126ae59d5fc1130b65f9d5decb7479232e0c648 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.7-max/f293fa1d-ea4d-4469-be2c-829711b53f4b.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/alibaba_qwen3.7-max/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.7-max", + "id": "alibaba/qwen3.7-max", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.7-max", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:alibaba/qwen3.7-max:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 47.671, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "47.671", + "raw_stderr": "4.63", + "latency": "4689.627", + "cost_per_test": "11.400322", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 4.63, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.7-plus/3013059b-fa1a-4119-80f3-ed5ce840796d.json b/data/vals_ai/alibaba/qwen3.7-plus/3013059b-fa1a-4119-80f3-ed5ce840796d.json new file mode 100644 index 0000000000000000000000000000000000000000..e2a28ffd2eb1b89312da27670b61e644781414a2 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.7-plus/3013059b-fa1a-4119-80f3-ed5ce840796d.json @@ -0,0 +1,398 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/alibaba_qwen3.7-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.7-plus", + "id": "alibaba/qwen3.7-plus", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.7-plus", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 62.005, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "62.005", + "raw_stderr": "1.657", + "latency": "34.056", + "cost_per_test": "0.028532", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.657, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 41.245, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "41.245", + "raw_stderr": "1.036", + "latency": "639.699", + "cost_per_test": "0.342801", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.036, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:overall:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 52.331, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "52.331", + "raw_stderr": "1.628", + "latency": "832.609", + "cost_per_test": "0.351425", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.628, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "66.667", + "raw_stderr": "4.668", + "latency": "602.556", + "cost_per_test": "0.207597" + }, + "uncertainty": { + "standard_error": { + "value": 4.668, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 52.809, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "52.809", + "raw_stderr": "0.649", + "latency": "654.983", + "cost_per_test": "0.095849", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.649, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 46.94, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "46.94", + "raw_stderr": "7.011", + "latency": "2231.75", + "cost_per_test": "1.082348", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 7.011, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.7-plus/32ad3751-a214-417a-b4de-521108759520.json b/data/vals_ai/alibaba/qwen3.7-plus/32ad3751-a214-417a-b4de-521108759520.json new file mode 100644 index 0000000000000000000000000000000000000000..c044aa801d36606659e68fe6315590b72aa637ad --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.7-plus/32ad3751-a214-417a-b4de-521108759520.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/alibaba_qwen3.7-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.7-plus", + "id": "alibaba/qwen3.7-plus", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.7-plus", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 75.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "75", + "raw_stderr": "14.434", + "latency": "14573.382", + "cost_per_test": "2.13264", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 14.434, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 23.333, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "23.333", + "raw_stderr": "3.849", + "latency": "1943.118", + "cost_per_test": "0.284352", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 3.849, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 67.273, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "67.273", + "raw_stderr": "1.818", + "latency": "1059.882", + "cost_per_test": "0.155101", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.818, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 52.809, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "52.809", + "raw_stderr": "0.649", + "latency": "654.983", + "cost_per_test": "0.095849", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.649, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.7-plus/371deb2c-14e4-4a5a-b243-0c24fe257de1.json b/data/vals_ai/alibaba/qwen3.7-plus/371deb2c-14e4-4a5a-b243-0c24fe257de1.json new file mode 100644 index 0000000000000000000000000000000000000000..46c0595aeea874ab235c6c3d271bf3739a02ab8c --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.7-plus/371deb2c-14e4-4a5a-b243-0c24fe257de1.json @@ -0,0 +1,718 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/alibaba_qwen3.7-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.7-plus", + "id": "alibaba/qwen3.7-plus", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.7-plus", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 35.538, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "35.538", + "raw_stderr": "2.485", + "latency": "733.145", + "cost_per_test": "0.424503", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.485, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 25.816, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "25.816", + "raw_stderr": "1.318", + "latency": "542.045", + "cost_per_test": "0.357661", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.318, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 38.848, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "38.848", + "raw_stderr": "2.279", + "latency": "1023.789", + "cost_per_test": "0.486576", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.279, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 44.441, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "44.441", + "raw_stderr": "0.797", + "latency": "704.588", + "cost_per_test": "0.364243", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.797, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 50.339, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "50.339", + "raw_stderr": "2.073", + "latency": "616.533", + "cost_per_test": "0.143864", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.073, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 10.057, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "10.057", + "raw_stderr": "0.985", + "latency": "802.253", + "cost_per_test": "0.361189", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.985, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 45.979, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "45.979", + "raw_stderr": "1.143", + "latency": "535.085", + "cost_per_test": "0.336957", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.143, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 52.24, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.24", + "raw_stderr": "2.78", + "latency": "690.858", + "cost_per_test": "0.429169", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.78, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 48.47, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "48.47", + "raw_stderr": "1.552", + "latency": "506.411", + "cost_per_test": "0.148204", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.552, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:overall:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 38.22, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "38.22", + "raw_stderr": "1.04", + "latency": "542.045", + "cost_per_test": "0.357661", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.04, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 18.067, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "18.067", + "raw_stderr": "2.655", + "latency": "1002.639", + "cost_per_test": "0.525511", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.655, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.7-plus/4f480a21-437d-4bed-9794-8f1f855b537f.json b/data/vals_ai/alibaba/qwen3.7-plus/4f480a21-437d-4bed-9794-8f1f855b537f.json new file mode 100644 index 0000000000000000000000000000000000000000..c0bea01390aecd692fb63d1808e4beb6c202cf30 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.7-plus/4f480a21-437d-4bed-9794-8f1f855b537f.json @@ -0,0 +1,656 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/emb/alibaba_qwen3.7-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - EMB", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.7-plus", + "id": "alibaba/qwen3.7-plus", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.7-plus", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "emb:comps:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.emb.comps", + "source_data": { + "dataset_name": "EMB - Comps", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "comps", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Comps).", + "metric_id": "vals_ai.emb.comps.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 39.126, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "comps", + "task_name": "Comps", + "dataset_type": "private", + "industry": "finance", + "raw_score": "39.126", + "raw_stderr": "7.846", + "latency": "1981.624", + "cost_per_test": "0.878993", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 7.846, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:dataroom:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.emb.dataroom", + "source_data": { + "dataset_name": "EMB - Dataroom", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dataroom", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Dataroom).", + "metric_id": "vals_ai.emb.dataroom.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 71.712, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dataroom", + "task_name": "Dataroom", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.712", + "raw_stderr": "7.809", + "latency": "1653.009", + "cost_per_test": "0.41702", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 7.809, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:dcf:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.emb.dcf", + "source_data": { + "dataset_name": "EMB - DCF", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dcf", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (DCF).", + "metric_id": "vals_ai.emb.dcf.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 37.242, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dcf", + "task_name": "DCF", + "dataset_type": "private", + "industry": "finance", + "raw_score": "37.242", + "raw_stderr": "8.715", + "latency": "2177.255", + "cost_per_test": "0.436574", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 8.715, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:lbo:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.emb.lbo", + "source_data": { + "dataset_name": "EMB - LBO", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "lbo", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (LBO).", + "metric_id": "vals_ai.emb.lbo.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 37.986, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "lbo", + "task_name": "LBO", + "dataset_type": "private", + "industry": "finance", + "raw_score": "37.986", + "raw_stderr": "6.186", + "latency": "3113.006", + "cost_per_test": "0.985801", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 6.186, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:ma:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.emb.ma", + "source_data": { + "dataset_name": "EMB - M&A", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "ma", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (M&A).", + "metric_id": "vals_ai.emb.ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 59.022, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "ma", + "task_name": "M&A", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.022", + "raw_stderr": "7.824", + "latency": "2155.373", + "cost_per_test": "0.42061", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 7.824, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:misc:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.emb.misc", + "source_data": { + "dataset_name": "EMB - Misc", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "misc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Misc).", + "metric_id": "vals_ai.emb.misc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 42.31, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "misc", + "task_name": "Misc", + "dataset_type": "private", + "industry": "finance", + "raw_score": "42.31", + "raw_stderr": "8.382", + "latency": "2195.876", + "cost_per_test": "0.468188", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 8.382, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:operating:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.emb.operating", + "source_data": { + "dataset_name": "EMB - Operating", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "operating", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Operating).", + "metric_id": "vals_ai.emb.operating.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 58.018, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "operating", + "task_name": "Operating", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.018", + "raw_stderr": "7.224", + "latency": "2104.952", + "cost_per_test": "0.479629", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 7.224, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:overall:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.emb.overall", + "source_data": { + "dataset_name": "EMB - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Overall).", + "metric_id": "vals_ai.emb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 49.345, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "49.345", + "raw_stderr": "3.181", + "latency": "2209.243", + "cost_per_test": "0.597724", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 3.181, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:scratch:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.emb.scratch", + "source_data": { + "dataset_name": "EMB - Scratch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "scratch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Scratch).", + "metric_id": "vals_ai.emb.scratch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 56.962, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "scratch", + "task_name": "Scratch", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.962", + "raw_stderr": "2.874", + "latency": "2346.663", + "cost_per_test": "0.652699", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.874, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:template:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.emb.template", + "source_data": { + "dataset_name": "EMB - Template", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "template", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Template).", + "metric_id": "vals_ai.emb.template.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 41.498, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "template", + "task_name": "Template", + "dataset_type": "private", + "industry": "finance", + "raw_score": "41.498", + "raw_stderr": "5.26", + "latency": "2071.822", + "cost_per_test": "0.54275", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 5.26, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.7-plus/524688b4-949f-40ef-a9b2-9be98e44e60b.json b/data/vals_ai/alibaba/qwen3.7-plus/524688b4-949f-40ef-a9b2-9be98e44e60b.json new file mode 100644 index 0000000000000000000000000000000000000000..cc971dbc35697778785cc0ee38096d443ce0b332 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.7-plus/524688b4-949f-40ef-a9b2-9be98e44e60b.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/alibaba_qwen3.7-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.7-plus", + "id": "alibaba/qwen3.7-plus", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.7-plus", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 74.801, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "74.801", + "raw_stderr": "1.224", + "latency": "79.471", + "cost_per_test": "0.009322", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.224, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 66.176, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.176", + "raw_stderr": "0.928", + "latency": "60.548", + "cost_per_test": "0.007588", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.928, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 57.552, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.552", + "raw_stderr": "1.394", + "latency": "41.625", + "cost_per_test": "0.005853", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.394, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.7-plus/97737c89-32d6-4860-a74b-6662e147d741.json b/data/vals_ai/alibaba/qwen3.7-plus/97737c89-32d6-4860-a74b-6662e147d741.json new file mode 100644 index 0000000000000000000000000000000000000000..beced3c3df9e4a7d1f6e63f14fbeb7e67dc13051 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.7-plus/97737c89-32d6-4860-a74b-6662e147d741.json @@ -0,0 +1,194 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/cyber/alibaba_qwen3.7-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CyberBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "beta", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.7-plus", + "id": "alibaba/qwen3.7-plus", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.7-plus", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "cyber:overall:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.cyber.overall", + "source_data": { + "dataset_name": "CyberBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (Overall).", + "metric_id": "vals_ai.cyber.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 37.5, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "beta", + "raw_score": "37.5", + "raw_stderr": "3.002", + "latency": "454.079", + "cost_per_test": "0.0999", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 3.002, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "cyber:patch:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.cyber.patch", + "source_data": { + "dataset_name": "CyberBench - Patch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "patch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (Patch).", + "metric_id": "vals_ai.cyber.patch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 75.0, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "patch", + "task_name": "Patch", + "dataset_type": "private", + "industry": "beta", + "raw_score": "75", + "raw_stderr": "6.005", + "latency": "908.157", + "cost_per_test": "0.0999", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 6.005, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "cyber:poc:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.cyber.poc", + "source_data": { + "dataset_name": "CyberBench - PoC", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "poc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (PoC).", + "metric_id": "vals_ai.cyber.poc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "poc", + "task_name": "PoC", + "dataset_type": "private", + "industry": "beta", + "raw_score": "0", + "raw_stderr": "0", + "latency": "0", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.7-plus/a1500863-34bf-4367-8ecb-28670d1d76b0.json b/data/vals_ai/alibaba/qwen3.7-plus/a1500863-34bf-4367-8ecb-28670d1d76b0.json new file mode 100644 index 0000000000000000000000000000000000000000..b452a800fbe2e507d5043afb82b9f687bfde9130 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.7-plus/a1500863-34bf-4367-8ecb-28670d1d76b0.json @@ -0,0 +1,522 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_multimodal_index/alibaba_qwen3.7-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Multimodal Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.7-plus", + "id": "alibaba/qwen3.7-plus", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.7-plus", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_multimodal_index:corp_fin_v2:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.vals_multimodal_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Multimodal Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (CorpFin v2).", + "metric_id": "vals_ai.vals_multimodal_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 62.005, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "62.005", + "raw_stderr": "1.657", + "latency": "34.056", + "cost_per_test": "0.028532", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.657, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:finance_agent:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.vals_multimodal_index.finance_agent", + "source_data": { + "dataset_name": "Vals Multimodal Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_multimodal_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 41.245, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "41.245", + "raw_stderr": "1.036", + "latency": "639.699", + "cost_per_test": "0.342801", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.036, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:mortgage_tax:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.vals_multimodal_index.mortgage_tax", + "source_data": { + "dataset_name": "Vals Multimodal Index - Mortgage Tax", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "mortgage_tax", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Mortgage Tax).", + "metric_id": "vals_ai.vals_multimodal_index.mortgage_tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 66.176, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "mortgage_tax", + "task_name": "Mortgage Tax", + "dataset_type": "private", + "industry": "index", + "raw_score": "66.176", + "raw_stderr": "0.928", + "latency": "60.548", + "cost_per_test": "0.007588", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.928, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:overall:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.vals_multimodal_index.overall", + "source_data": { + "dataset_name": "Vals Multimodal Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Overall).", + "metric_id": "vals_ai.vals_multimodal_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 53.892, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "53.892", + "raw_stderr": "1.478", + "latency": "628.662", + "cost_per_test": "0.255213", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 1.478, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:sage:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.vals_multimodal_index.sage", + "source_data": { + "dataset_name": "Vals Multimodal Index - SAGE", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "sage", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SAGE).", + "metric_id": "vals_ai.vals_multimodal_index.sage.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 39.251, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "sage", + "task_name": "SAGE", + "dataset_type": "private", + "industry": "index", + "raw_score": "39.251", + "raw_stderr": "3.384", + "latency": "177.044", + "cost_per_test": "0.021774", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 3.384, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:swebench:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.vals_multimodal_index.swebench", + "source_data": { + "dataset_name": "Vals Multimodal Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SWE-bench).", + "metric_id": "vals_ai.vals_multimodal_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "66.667", + "raw_stderr": "4.668", + "latency": "602.556", + "cost_per_test": "0.207597" + }, + "uncertainty": { + "standard_error": { + "value": 4.668, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:terminal_bench_2_1:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.vals_multimodal_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Multimodal Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_multimodal_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 52.809, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "52.809", + "raw_stderr": "0.649", + "latency": "654.983", + "cost_per_test": "0.095849", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.649, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:vibe_code_bench:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.vals_multimodal_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Multimodal Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_multimodal_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 46.94, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "46.94", + "raw_stderr": "7.011", + "latency": "2231.75", + "cost_per_test": "1.082348", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 7.011, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.7-plus/a96e62a1-9571-4b17-bc6e-426159d99048.json b/data/vals_ai/alibaba/qwen3.7-plus/a96e62a1-9571-4b17-bc6e-426159d99048.json new file mode 100644 index 0000000000000000000000000000000000000000..299912176b906002d56918823ecdabc0cb00fb70 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.7-plus/a96e62a1-9571-4b17-bc6e-426159d99048.json @@ -0,0 +1,3286 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/hlab/alibaba_qwen3.7-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Harvey's Legal Agent Benchmark", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.7-plus", + "id": "alibaba/qwen3.7-plus", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.7-plus", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "hlab:antitrust_competition:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1000.301", + "cost_per_test": "0.706227", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:arbitration_international_dispute_resolution:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "664.978", + "cost_per_test": "0.232631", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:banking_finance:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "766.156", + "cost_per_test": "0.171268", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:bankruptcy_restructuring:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "461.048", + "cost_per_test": "0.269306", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:capital_markets:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "255.419", + "cost_per_test": "0.119538", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:corporate_governance:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "364.932", + "cost_per_test": "0.133065", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:corporate_ma:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "457.836", + "cost_per_test": "0.17506", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Criteria Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Criteria Pass Rate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 62.694, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate", + "task_name": "Criteria Pass Rate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "62.694", + "raw_stderr": "0.587", + "latency": "529.209", + "cost_per_test": "0.234226", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.587, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__antitrust_competition:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 60.145, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "60.145", + "raw_stderr": "2.952", + "latency": "1000.301", + "cost_per_test": "0.706227", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.952, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__arbitration_international_dispute_resolution:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 58.423, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "58.423", + "raw_stderr": "2.956", + "latency": "664.978", + "cost_per_test": "0.232631", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.956, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__banking_finance:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 64.312, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "64.312", + "raw_stderr": "2.926", + "latency": "766.156", + "cost_per_test": "0.171268", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.926, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__bankruptcy_restructuring:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 55.017, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "55.017", + "raw_stderr": "2.931", + "latency": "461.048", + "cost_per_test": "0.269306", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.931, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__capital_markets:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 72.941, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "72.941", + "raw_stderr": "2.788", + "latency": "255.419", + "cost_per_test": "0.119538", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.788, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_governance:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 64.384, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "64.384", + "raw_stderr": "2.807", + "latency": "364.932", + "cost_per_test": "0.133065", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.807, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_ma:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 61.218, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "61.218", + "raw_stderr": "2.763", + "latency": "457.836", + "cost_per_test": "0.17506", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.763, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__data_privacy_cybersecurity:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 63.958, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "63.958", + "raw_stderr": "2.859", + "latency": "329.543", + "cost_per_test": "0.12102", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.859, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__emerging_companies_venture_capital:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 59.044, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "59.044", + "raw_stderr": "2.878", + "latency": "870", + "cost_per_test": "0.309345", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.878, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__employment_labor:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 56.18, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "56.18", + "raw_stderr": "3.042", + "latency": "421.742", + "cost_per_test": "0.169864", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 3.042, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__energy_natural_resources:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 74.257, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "74.257", + "raw_stderr": "2.516", + "latency": "368.878", + "cost_per_test": "0.181632", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.516, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__environmental_esg:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 55.502, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "55.502", + "raw_stderr": "3.446", + "latency": "1173.584", + "cost_per_test": "0.395489", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 3.446, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__funds_asset_management:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 53.353, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "53.353", + "raw_stderr": "2.698", + "latency": "344.292", + "cost_per_test": "0.132025", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.698, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__healthcare_life_sciences:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 69.307, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "69.307", + "raw_stderr": "2.654", + "latency": "496.704", + "cost_per_test": "0.183866", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.654, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__immigration:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 62.546, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "62.546", + "raw_stderr": "2.924", + "latency": "309.231", + "cost_per_test": "0.127284", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.924, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__insurance:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 57.509, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "57.509", + "raw_stderr": "2.997", + "latency": "879.71", + "cost_per_test": "0.471893", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.997, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__intellectual_property:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 79.821, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.821", + "raw_stderr": "2.694", + "latency": "1112.95", + "cost_per_test": "0.39665", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.694, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__international_trade_sanctions:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 71.032, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "71.032", + "raw_stderr": "2.863", + "latency": "438.76", + "cost_per_test": "0.359741", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.863, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__litigation_dispute_resolution:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 65.543, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "65.543", + "raw_stderr": "2.914", + "latency": "361.003", + "cost_per_test": "0.145792", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.914, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__real_estate:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 65.598, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "65.598", + "raw_stderr": "2.569", + "latency": "415.567", + "cost_per_test": "0.175367", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.569, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__structured_finance_securitization:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 60.318, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "60.318", + "raw_stderr": "2.761", + "latency": "353.192", + "cost_per_test": "0.195145", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.761, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__tax:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 60.87, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "60.87", + "raw_stderr": "2.827", + "latency": "404.452", + "cost_per_test": "0.177002", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.827, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__trusts_estates_private_client:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 72.519, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "72.519", + "raw_stderr": "2.763", + "latency": "372.309", + "cost_per_test": "0.188846", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.763, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__white_collar_defense_investigations:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 46.465, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "46.465", + "raw_stderr": "2.899", + "latency": "324.062", + "cost_per_test": "0.148109", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 2.899, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:data_privacy_cybersecurity:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "329.543", + "cost_per_test": "0.12102", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:emerging_companies_venture_capital:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "870", + "cost_per_test": "0.309345", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:employment_labor:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "421.742", + "cost_per_test": "0.169864", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:energy_natural_resources:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "368.878", + "cost_per_test": "0.181632", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:environmental_esg:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1173.584", + "cost_per_test": "0.395489", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:funds_asset_management:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "344.292", + "cost_per_test": "0.132025", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:healthcare_life_sciences:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "496.704", + "cost_per_test": "0.183866", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:immigration:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "309.231", + "cost_per_test": "0.127284", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:insurance:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "879.71", + "cost_per_test": "0.471893", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:intellectual_property:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1112.95", + "cost_per_test": "0.39665", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:international_trade_sanctions:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "438.76", + "cost_per_test": "0.359741", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:litigation_dispute_resolution:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "361.003", + "cost_per_test": "0.145792", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:overall:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.overall", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Overall).", + "metric_id": "vals_ai.hlab.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "529.209", + "cost_per_test": "0.234226", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:real_estate:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "415.567", + "cost_per_test": "0.175367", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:structured_finance_securitization:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "353.192", + "cost_per_test": "0.195145", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:tax:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "404.452", + "cost_per_test": "0.177002", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:trusts_estates_private_client:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "372.309", + "cost_per_test": "0.188846", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:white_collar_defense_investigations:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.hlab.white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "324.062", + "cost_per_test": "0.148109", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.7-plus/b4877a9a-400e-4959-b20c-a73435cde33b.json b/data/vals_ai/alibaba/qwen3.7-plus/b4877a9a-400e-4959-b20c-a73435cde33b.json new file mode 100644 index 0000000000000000000000000000000000000000..bb1b94041e20a2c1b3efd5b69c2f432dc36a879d --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.7-plus/b4877a9a-400e-4959-b20c-a73435cde33b.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/alibaba_qwen3.7-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.7-plus", + "id": "alibaba/qwen3.7-plus", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.7-plus", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 46.391, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "46.391", + "raw_stderr": "4.614", + "latency": "2231.75", + "cost_per_test": "1.082348", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 4.614, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/alibaba/qwen3.7-plus/d5e27204-9068-41c3-9d52-890d728abfb7.json b/data/vals_ai/alibaba/qwen3.7-plus/d5e27204-9068-41c3-9d52-890d728abfb7.json new file mode 100644 index 0000000000000000000000000000000000000000..85781f8da8a9a6b0070f2a7485682de790cda6d0 --- /dev/null +++ b/data/vals_ai/alibaba/qwen3.7-plus/d5e27204-9068-41c3-9d52-890d728abfb7.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/alibaba_qwen3.7-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3.7-plus", + "id": "alibaba/qwen3.7-plus", + "developer": "alibaba", + "additional_details": { + "vals_model_id": "alibaba/qwen3.7-plus", + "vals_provider": "Alibaba" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "33.333", + "raw_stderr": "5.927", + "latency": "189.296", + "cost_per_test": "0.020906", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 5.927, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 32.54, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "32.54", + "raw_stderr": "6.027", + "latency": "165.302", + "cost_per_test": "0.017998", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 6.027, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 51.88, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "51.88", + "raw_stderr": "5.623", + "latency": "176.535", + "cost_per_test": "0.026418", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 5.623, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:alibaba/qwen3.7-plus:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 39.251, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "39.251", + "raw_stderr": "3.384", + "latency": "177.044", + "cost_per_test": "0.021774", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Alibaba" + }, + "uncertainty": { + "standard_error": { + "value": 3.384, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-3-5-haiku-20241022/345929c8-7c6c-4e89-89d8-886ca9e5d6d3.json b/data/vals_ai/anthropic/claude-3-5-haiku-20241022/345929c8-7c6c-4e89-89d8-886ca9e5d6d3.json new file mode 100644 index 0000000000000000000000000000000000000000..b51ff4fb9f34fa0374cd4dbc5e384e84aafe39ba --- /dev/null +++ b/data/vals_ai/anthropic/claude-3-5-haiku-20241022/345929c8-7c6c-4e89-89d8-886ca9e5d6d3.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/anthropic_claude-3-5-haiku-20241022/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-3-5-haiku-20241022", + "id": "anthropic/claude-3-5-haiku-20241022", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-3-5-haiku-20241022", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 76.226, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "76.226", + "raw_stderr": "0.976", + "latency": "0.89", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.976, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 74.919, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "74.919", + "raw_stderr": "0.217", + "latency": "0.992", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.217, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 58.652, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "58.652", + "raw_stderr": "0.593", + "latency": "0.999", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.593, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 70.331, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "70.331", + "raw_stderr": "0.64", + "latency": "1.066", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.64, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 73.208, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "73.208", + "raw_stderr": "1.439", + "latency": "0.954", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.439, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 68.65, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "68.65", + "raw_stderr": "2.614", + "latency": "1.493", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.614, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-3-5-haiku-20241022/6ed72ceb-3f1e-450b-95ff-e4ad47a4e027.json b/data/vals_ai/anthropic/claude-3-5-haiku-20241022/6ed72ceb-3f1e-450b-95ff-e4ad47a4e027.json new file mode 100644 index 0000000000000000000000000000000000000000..1a4a9eb09aa4bf1b76c402298bc5e810b7a9c7c0 --- /dev/null +++ b/data/vals_ai/anthropic/claude-3-5-haiku-20241022/6ed72ceb-3f1e-450b-95ff-e4ad47a4e027.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/anthropic_claude-3-5-haiku-20241022/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-3-5-haiku-20241022", + "id": "anthropic/claude-3-5-haiku-20241022", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-3-5-haiku-20241022", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 85.714, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "85.714", + "raw_stderr": "1.95", + "latency": "8.556", + "cost_per_test": "0.002431", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.95, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 7.143, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "7.143", + "raw_stderr": "1.377", + "latency": "12.326", + "cost_per_test": "0.003605", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.377, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 32.898, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "32.898", + "raw_stderr": "2.401", + "latency": "10.86", + "cost_per_test": "0.003155", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.401, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 41.918, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "41.918", + "raw_stderr": "1.129", + "latency": "10.581", + "cost_per_test": "0.003064", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.129, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-3-5-haiku-20241022/7d8c5471-b7ae-436c-a580-1040519df45d.json b/data/vals_ai/anthropic/claude-3-5-haiku-20241022/7d8c5471-b7ae-436c-a580-1040519df45d.json new file mode 100644 index 0000000000000000000000000000000000000000..bc291a1d4cb3d9209d6f6e9390bb810d22d820f8 --- /dev/null +++ b/data/vals_ai/anthropic/claude-3-5-haiku-20241022/7d8c5471-b7ae-436c-a580-1040519df45d.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/anthropic_claude-3-5-haiku-20241022/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-3-5-haiku-20241022", + "id": "anthropic/claude-3-5-haiku-20241022", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-3-5-haiku-20241022", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 46.525, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "46.525", + "raw_stderr": "1.426", + "latency": "4.987", + "cost_per_test": "0.001317", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.426, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 57.359, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.359", + "raw_stderr": "0.976", + "latency": "4.987", + "cost_per_test": "0.001317", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.976, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 68.193, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.193", + "raw_stderr": "1.332", + "latency": "4.987", + "cost_per_test": "0.001317", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.332, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-3-5-haiku-20241022/8ddaa408-7232-468d-b045-c187978ac5a1.json b/data/vals_ai/anthropic/claude-3-5-haiku-20241022/8ddaa408-7232-468d-b045-c187978ac5a1.json new file mode 100644 index 0000000000000000000000000000000000000000..66357c29c72a1e41adec1541a8990d6a89b1ecc6 --- /dev/null +++ b/data/vals_ai/anthropic/claude-3-5-haiku-20241022/8ddaa408-7232-468d-b045-c187978ac5a1.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/anthropic_claude-3-5-haiku-20241022/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-3-5-haiku-20241022", + "id": "anthropic/claude-3-5-haiku-20241022", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-3-5-haiku-20241022", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 56.41, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.41", + "raw_stderr": "1.693", + "latency": "0.062", + "cost_per_test": "0.001851", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.693, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 44.755, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "44.755", + "raw_stderr": "1.698", + "latency": "0.009", + "cost_per_test": "0.129482", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.698, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 50.816, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "50.816", + "raw_stderr": "0.981", + "latency": "0.039", + "cost_per_test": "0.063315", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.981, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 51.282, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "51.282", + "raw_stderr": "1.706", + "latency": "0.046", + "cost_per_test": "0.058611", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.706, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-3-5-haiku-20241022/ac21ef9e-9b95-4b54-b1f4-a95c138d00b5.json b/data/vals_ai/anthropic/claude-3-5-haiku-20241022/ac21ef9e-9b95-4b54-b1f4-a95c138d00b5.json new file mode 100644 index 0000000000000000000000000000000000000000..29509a4f353e626dbbd8c9546bbdf3baf089b396 --- /dev/null +++ b/data/vals_ai/anthropic/claude-3-5-haiku-20241022/ac21ef9e-9b95-4b54-b1f4-a95c138d00b5.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/anthropic_claude-3-5-haiku-20241022/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-3-5-haiku-20241022", + "id": "anthropic/claude-3-5-haiku-20241022", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-3-5-haiku-20241022", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 36.364, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "36.364", + "raw_stderr": "3.419", + "latency": "6.978", + "cost_per_test": "0.003004", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.419, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 39.394, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "39.394", + "raw_stderr": "3.472", + "latency": "6.143", + "cost_per_test": "0.001437", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.472, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 37.879, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "37.879", + "raw_stderr": "2.436", + "latency": "6.56", + "cost_per_test": "0.00222", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.436, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-3-5-haiku-20241022/e4c1ace5-0c4f-42c4-83fa-7170f1540250.json b/data/vals_ai/anthropic/claude-3-5-haiku-20241022/e4c1ace5-0c4f-42c4-83fa-7170f1540250.json new file mode 100644 index 0000000000000000000000000000000000000000..e0cfd94686036eda3e2c85face5293ca3ee761b3 --- /dev/null +++ b/data/vals_ai/anthropic/claude-3-5-haiku-20241022/e4c1ace5-0c4f-42c4-83fa-7170f1540250.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/anthropic_claude-3-5-haiku-20241022/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-3-5-haiku-20241022", + "id": "anthropic/claude-3-5-haiku-20241022", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-3-5-haiku-20241022", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.869, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.869", + "raw_stderr": "1.439", + "latency": "5.974", + "cost_per_test": "0.002267", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.439, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 69.835, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.835", + "raw_stderr": "1.634", + "latency": "5.48", + "cost_per_test": "0.001917", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.634, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 61.219, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "61.219", + "raw_stderr": "1.448", + "latency": "5.68", + "cost_per_test": "0.002622", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.448, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 65.61, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "65.61", + "raw_stderr": "2.346", + "latency": "5.583", + "cost_per_test": "0.002407", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.346, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.697, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.697", + "raw_stderr": "1.516", + "latency": "5.701", + "cost_per_test": "0.002098", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.516, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 45.924, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "45.924", + "raw_stderr": "1.601", + "latency": "6.225", + "cost_per_test": "0.002363", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.601, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 66.748, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.748", + "raw_stderr": "1.647", + "latency": "5.851", + "cost_per_test": "0.001951", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.647, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 63.255, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "63.255", + "raw_stderr": "2.47", + "latency": "5.903", + "cost_per_test": "0.00281", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.47, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 44.142, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "44.142", + "raw_stderr": "1.496", + "latency": "6.261", + "cost_per_test": "0.002355", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.496, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 66.173, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.173", + "raw_stderr": "1.287", + "latency": "5.628", + "cost_per_test": "0.002685", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.287, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 66.991, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.991", + "raw_stderr": "1.547", + "latency": "5.582", + "cost_per_test": "0.001749", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.547, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 64.117, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "64.117", + "raw_stderr": "0.46", + "latency": "5.792", + "cost_per_test": "0.002224", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.46, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 58.317, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "58.317", + "raw_stderr": "2.207", + "latency": "5.723", + "cost_per_test": "0.001843", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.207, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 61.432, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "61.432", + "raw_stderr": "1.351", + "latency": "5.691", + "cost_per_test": "0.00214", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.351, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:anthropic/claude-3-5-haiku-20241022:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 72.431, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.431", + "raw_stderr": "1.582", + "latency": "5.807", + "cost_per_test": "0.001922", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.582, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-3-5-sonnet-20241022/1a1fc5fc-1adc-4096-9ee5-09ca6ecc1f39.json b/data/vals_ai/anthropic/claude-3-5-sonnet-20241022/1a1fc5fc-1adc-4096-9ee5-09ca6ecc1f39.json new file mode 100644 index 0000000000000000000000000000000000000000..f36d2212b8cdc810b96ab04561c15584f0cc9642 --- /dev/null +++ b/data/vals_ai/anthropic/claude-3-5-sonnet-20241022/1a1fc5fc-1adc-4096-9ee5-09ca6ecc1f39.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/anthropic_claude-3-5-sonnet-20241022/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-3-5-sonnet-20241022", + "id": "anthropic/claude-3-5-sonnet-20241022", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-3-5-sonnet-20241022", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 59.853, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.853", + "raw_stderr": "1.402", + "latency": "5.943", + "cost_per_test": "0.006006", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.402, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 70.156, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "70.156", + "raw_stderr": "0.902", + "latency": "5.943", + "cost_per_test": "0.006006", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.902, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 80.458, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "80.458", + "raw_stderr": "1.134", + "latency": "5.943", + "cost_per_test": "0.006006", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.134, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-3-5-sonnet-20241022/2b0d34cb-667a-4124-9e90-f4fc129d891d.json b/data/vals_ai/anthropic/claude-3-5-sonnet-20241022/2b0d34cb-667a-4124-9e90-f4fc129d891d.json new file mode 100644 index 0000000000000000000000000000000000000000..46b014877af8f214168f2fdec69d01fd2e9e3484 --- /dev/null +++ b/data/vals_ai/anthropic/claude-3-5-sonnet-20241022/2b0d34cb-667a-4124-9e90-f4fc129d891d.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/anthropic_claude-3-5-sonnet-20241022/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-3-5-sonnet-20241022", + "id": "anthropic/claude-3-5-sonnet-20241022", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-3-5-sonnet-20241022", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 57.809, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.809", + "raw_stderr": "1.686", + "latency": "0.047", + "cost_per_test": "0.007112", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.686, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 53.263, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "53.263", + "raw_stderr": "1.703", + "latency": "0.049", + "cost_per_test": "0.486092", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.703, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 53.613, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "53.613", + "raw_stderr": "0.981", + "latency": "0.049", + "cost_per_test": "0.23772", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.981, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 49.767, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "49.767", + "raw_stderr": "1.707", + "latency": "0.052", + "cost_per_test": "0.219957", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.707, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-3-5-sonnet-20241022/4715abbe-da8b-430e-b8e7-341efaae7f2f.json b/data/vals_ai/anthropic/claude-3-5-sonnet-20241022/4715abbe-da8b-430e-b8e7-341efaae7f2f.json new file mode 100644 index 0000000000000000000000000000000000000000..6fb2d983fcb67f612cd32947c1aa10f36d8fe93c --- /dev/null +++ b/data/vals_ai/anthropic/claude-3-5-sonnet-20241022/4715abbe-da8b-430e-b8e7-341efaae7f2f.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/anthropic_claude-3-5-sonnet-20241022/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-3-5-sonnet-20241022", + "id": "anthropic/claude-3-5-sonnet-20241022", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-3-5-sonnet-20241022", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 76.868, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "76.868", + "raw_stderr": "2.328", + "latency": "4.079", + "cost_per_test": "0.01475", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.328, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 64.07, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.07", + "raw_stderr": "1.805", + "latency": "4.224", + "cost_per_test": "0.015076", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.805, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 51.272, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "51.272", + "raw_stderr": "2.758", + "latency": "4.369", + "cost_per_test": "0.015402", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.758, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-3-5-sonnet-20241022/6b82ede6-ed28-4687-b93e-8ebcedd050bd.json b/data/vals_ai/anthropic/claude-3-5-sonnet-20241022/6b82ede6-ed28-4687-b93e-8ebcedd050bd.json new file mode 100644 index 0000000000000000000000000000000000000000..5880427a0d2889d79ac0e562d2752b1d49ee45c0 --- /dev/null +++ b/data/vals_ai/anthropic/claude-3-5-sonnet-20241022/6b82ede6-ed28-4687-b93e-8ebcedd050bd.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/anthropic_claude-3-5-sonnet-20241022/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-3-5-sonnet-20241022", + "id": "anthropic/claude-3-5-sonnet-20241022", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-3-5-sonnet-20241022", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 92.236, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "92.236", + "raw_stderr": "1.491", + "latency": "8.78", + "cost_per_test": "0.008858", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.491, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 12.0, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "12", + "raw_stderr": "1.737", + "latency": "12.952", + "cost_per_test": "0.013989", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.737, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 44.648, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "44.648", + "raw_stderr": "2.54", + "latency": "11.374", + "cost_per_test": "0.012137", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.54, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 49.628, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "49.628", + "raw_stderr": "1.14", + "latency": "11.035", + "cost_per_test": "0.011661", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.14, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-3-5-sonnet-20241022/b131527d-e10d-47ed-a2c4-1749a16d195c.json b/data/vals_ai/anthropic/claude-3-5-sonnet-20241022/b131527d-e10d-47ed-a2c4-1749a16d195c.json new file mode 100644 index 0000000000000000000000000000000000000000..db5129f7def9ab21c4324b61666506d4007f2590 --- /dev/null +++ b/data/vals_ai/anthropic/claude-3-5-sonnet-20241022/b131527d-e10d-47ed-a2c4-1749a16d195c.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/anthropic_claude-3-5-sonnet-20241022/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-3-5-sonnet-20241022", + "id": "anthropic/claude-3-5-sonnet-20241022", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-3-5-sonnet-20241022", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 56.566, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "56.566", + "raw_stderr": "3.523", + "latency": "7.711", + "cost_per_test": "0.011928", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.523, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 62.121, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "62.121", + "raw_stderr": "3.456", + "latency": "6.771", + "cost_per_test": "0.006229", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.456, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 59.344, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "59.344", + "raw_stderr": "2.468", + "latency": "7.241", + "cost_per_test": "0.009078", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.468, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-3-5-sonnet-20241022/d7cedd62-9215-4f6a-9222-b9aa69f6aa57.json b/data/vals_ai/anthropic/claude-3-5-sonnet-20241022/d7cedd62-9215-4f6a-9222-b9aa69f6aa57.json new file mode 100644 index 0000000000000000000000000000000000000000..5e09f6251d969fc45cbfae5049425883ca2efa34 --- /dev/null +++ b/data/vals_ai/anthropic/claude-3-5-sonnet-20241022/d7cedd62-9215-4f6a-9222-b9aa69f6aa57.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/anthropic_claude-3-5-sonnet-20241022/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-3-5-sonnet-20241022", + "id": "anthropic/claude-3-5-sonnet-20241022", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-3-5-sonnet-20241022", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.819, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.819", + "raw_stderr": "1.129", + "latency": "6.303", + "cost_per_test": "0.008748", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.129, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.27, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.27", + "raw_stderr": "1.329", + "latency": "5.569", + "cost_per_test": "0.007379", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.329, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.77, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.77", + "raw_stderr": "1.194", + "latency": "6.262", + "cost_per_test": "0.010358", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.194, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.22, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.22", + "raw_stderr": "1.929", + "latency": "6.402", + "cost_per_test": "0.009551", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.929, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.123, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.123", + "raw_stderr": "1.258", + "latency": "5.825", + "cost_per_test": "0.007887", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.258, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 61.197, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "61.197", + "raw_stderr": "1.565", + "latency": "6.929", + "cost_per_test": "0.009441", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.565, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.973, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.973", + "raw_stderr": "1.425", + "latency": "6.52", + "cost_per_test": "0.007583", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.425, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.328, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.328", + "raw_stderr": "2.209", + "latency": "6.513", + "cost_per_test": "0.010784", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.209, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 67.666, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "67.666", + "raw_stderr": "1.41", + "latency": "6.734", + "cost_per_test": "0.009054", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.41, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.051, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.051", + "raw_stderr": "1.066", + "latency": "6.366", + "cost_per_test": "0.010386", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.066, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.545, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.545", + "raw_stderr": "1.327", + "latency": "5.76", + "cost_per_test": "0.006445", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.327, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.404, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.404", + "raw_stderr": "0.397", + "latency": "6.288", + "cost_per_test": "0.00859", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.397, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.152, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.152", + "raw_stderr": "1.908", + "latency": "6.094", + "cost_per_test": "0.007086", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.908, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.214, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.214", + "raw_stderr": "1.145", + "latency": "6.246", + "cost_per_test": "0.008225", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.145, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.328, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.328", + "raw_stderr": "1.379", + "latency": "6.511", + "cost_per_test": "0.00733", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.379, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-3-5-sonnet-20241022/d93004e9-5465-4baa-a9f6-bf840c70b1b0.json b/data/vals_ai/anthropic/claude-3-5-sonnet-20241022/d93004e9-5465-4baa-a9f6-bf840c70b1b0.json new file mode 100644 index 0000000000000000000000000000000000000000..6c6a7883f4f1546bac4972392eda236359f75b18 --- /dev/null +++ b/data/vals_ai/anthropic/claude-3-5-sonnet-20241022/d93004e9-5465-4baa-a9f6-bf840c70b1b0.json @@ -0,0 +1,101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/anthropic_claude-3-5-sonnet-20241022/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-3-5-sonnet-20241022", + "id": "anthropic/claude-3-5-sonnet-20241022", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-3-5-sonnet-20241022", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:anthropic/claude-3-5-sonnet-20241022:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 68.804, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "68.804", + "raw_stderr": "1.114", + "latency": "46.011", + "cost_per_test": "0.010237", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.114, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-3-7-sonnet-20250219-thinking/16517448-4cfc-46f4-a435-3c36bff6f04b.json b/data/vals_ai/anthropic/claude-3-7-sonnet-20250219-thinking/16517448-4cfc-46f4-a435-3c36bff6f04b.json new file mode 100644 index 0000000000000000000000000000000000000000..d1de1187dabf1f1eba4f22eab3e455d87d1b0b10 --- /dev/null +++ b/data/vals_ai/anthropic/claude-3-7-sonnet-20250219-thinking/16517448-4cfc-46f4-a435-3c36bff6f04b.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/anthropic_claude-3-7-sonnet-20250219-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-3-7-sonnet-20250219-thinking", + "id": "anthropic/claude-3-7-sonnet-20250219-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-3-7-sonnet-20250219-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 63.287, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.287", + "raw_stderr": "1.378", + "latency": "43.314", + "cost_per_test": "0.050764", + "temperature": "1", + "max_output_tokens": "18384", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.378, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 18384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 74.039, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "74.039", + "raw_stderr": "0.859", + "latency": "43.314", + "cost_per_test": "0.050764", + "temperature": "1", + "max_output_tokens": "18384", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.859, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 18384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 84.791, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "84.791", + "raw_stderr": "1.027", + "latency": "43.314", + "cost_per_test": "0.050764", + "temperature": "1", + "max_output_tokens": "18384", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.027, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 18384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-3-7-sonnet-20250219-thinking/1b2f1df7-f6d9-49b4-b026-686ee864d881.json b/data/vals_ai/anthropic/claude-3-7-sonnet-20250219-thinking/1b2f1df7-f6d9-49b4-b026-686ee864d881.json new file mode 100644 index 0000000000000000000000000000000000000000..4e345fc088ce157e0225fe21173144dece25652f --- /dev/null +++ b/data/vals_ai/anthropic/claude-3-7-sonnet-20250219-thinking/1b2f1df7-f6d9-49b4-b026-686ee864d881.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/anthropic_claude-3-7-sonnet-20250219-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-3-7-sonnet-20250219-thinking", + "id": "anthropic/claude-3-7-sonnet-20250219-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-3-7-sonnet-20250219-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.587, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.587", + "raw_stderr": "0.8", + "latency": "6.595", + "temperature": "1", + "max_output_tokens": "5096", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.8, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 5096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.261, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.261", + "raw_stderr": "0.2", + "latency": "6.686", + "temperature": "1", + "max_output_tokens": "5096", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.2, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 5096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 86.619, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.619", + "raw_stderr": "0.588", + "latency": "6.353", + "temperature": "1", + "max_output_tokens": "5096", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.588, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 5096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.519, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.519", + "raw_stderr": "0.53", + "latency": "6.536", + "temperature": "1", + "max_output_tokens": "5096", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.53, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 5096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.366, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.366", + "raw_stderr": "1.252", + "latency": "5.542", + "temperature": "1", + "max_output_tokens": "5096", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.252, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 5096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.762, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.762", + "raw_stderr": "2.107", + "latency": "7.506", + "temperature": "1", + "max_output_tokens": "5096", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.107, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 5096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-3-7-sonnet-20250219-thinking/57d0e904-044d-423a-8953-af4ad79e7258.json b/data/vals_ai/anthropic/claude-3-7-sonnet-20250219-thinking/57d0e904-044d-423a-8953-af4ad79e7258.json new file mode 100644 index 0000000000000000000000000000000000000000..bbcab1de2d656a516ce04af20936b106ef9c4c78 --- /dev/null +++ b/data/vals_ai/anthropic/claude-3-7-sonnet-20250219-thinking/57d0e904-044d-423a-8953-af4ad79e7258.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/anthropic_claude-3-7-sonnet-20250219-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-3-7-sonnet-20250219-thinking", + "id": "anthropic/claude-3-7-sonnet-20250219-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-3-7-sonnet-20250219-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 75.755, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "75.755", + "raw_stderr": "2.366", + "latency": "14.885", + "cost_per_test": "0.024887", + "temperature": "1", + "max_output_tokens": "4096", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.366, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 66.852, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.852", + "raw_stderr": "1.804", + "latency": "13.585", + "cost_per_test": "0.02365", + "temperature": "1", + "max_output_tokens": "4096", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.804, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 57.949, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.949", + "raw_stderr": "2.724", + "latency": "12.285", + "cost_per_test": "0.022414", + "temperature": "1", + "max_output_tokens": "4096", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.724, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-3-7-sonnet-20250219-thinking/74383842-cda6-45c4-a6f1-61eaecbd2678.json b/data/vals_ai/anthropic/claude-3-7-sonnet-20250219-thinking/74383842-cda6-45c4-a6f1-61eaecbd2678.json new file mode 100644 index 0000000000000000000000000000000000000000..ee116d9c33f50782469da18254b2fc9f03e8296a --- /dev/null +++ b/data/vals_ai/anthropic/claude-3-7-sonnet-20250219-thinking/74383842-cda6-45c4-a6f1-61eaecbd2678.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/anthropic_claude-3-7-sonnet-20250219-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-3-7-sonnet-20250219-thinking", + "id": "anthropic/claude-3-7-sonnet-20250219-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-3-7-sonnet-20250219-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 59.79, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.79", + "raw_stderr": "1.674", + "latency": "10.947", + "cost_per_test": "0.01504", + "temperature": "1", + "max_output_tokens": "18384", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.674, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 18384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 59.207, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.207", + "raw_stderr": "1.678", + "latency": "32.135", + "cost_per_test": "0.452639", + "temperature": "1", + "max_output_tokens": "18384", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.678, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 18384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.412, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.412", + "raw_stderr": "0.964", + "latency": "22.163", + "cost_per_test": "0.232197", + "temperature": "1", + "max_output_tokens": "18384", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.964, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 18384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 62.238, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.238", + "raw_stderr": "1.655", + "latency": "23.407", + "cost_per_test": "0.228911", + "temperature": "1", + "max_output_tokens": "18384", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.655, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 18384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-3-7-sonnet-20250219-thinking/8aaa7e6c-2843-4810-a79a-3c4011790fa9.json b/data/vals_ai/anthropic/claude-3-7-sonnet-20250219-thinking/8aaa7e6c-2843-4810-a79a-3c4011790fa9.json new file mode 100644 index 0000000000000000000000000000000000000000..922f728ae115cd4dcc1759df49287955b1df49d8 --- /dev/null +++ b/data/vals_ai/anthropic/claude-3-7-sonnet-20250219-thinking/8aaa7e6c-2843-4810-a79a-3c4011790fa9.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/anthropic_claude-3-7-sonnet-20250219-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-3-7-sonnet-20250219-thinking", + "id": "anthropic/claude-3-7-sonnet-20250219-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-3-7-sonnet-20250219-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 75.253, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.253", + "raw_stderr": "3.067", + "latency": "139.726", + "cost_per_test": "0.172742", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.067, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 75.253, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.253", + "raw_stderr": "3.067", + "latency": "170.315", + "cost_per_test": "0.204759", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.067, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 75.253, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.253", + "raw_stderr": "2.169", + "latency": "155.02", + "cost_per_test": "0.18875", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.169, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-3-7-sonnet-20250219-thinking/c3497b0c-dae7-4b9a-8207-d249cc7de151.json b/data/vals_ai/anthropic/claude-3-7-sonnet-20250219-thinking/c3497b0c-dae7-4b9a-8207-d249cc7de151.json new file mode 100644 index 0000000000000000000000000000000000000000..9dc3cb32fb5255dfc0d7629a6a284ab0480b1df2 --- /dev/null +++ b/data/vals_ai/anthropic/claude-3-7-sonnet-20250219-thinking/c3497b0c-dae7-4b9a-8207-d249cc7de151.json @@ -0,0 +1,101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/anthropic_claude-3-7-sonnet-20250219-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-3-7-sonnet-20250219-thinking", + "id": "anthropic/claude-3-7-sonnet-20250219-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-3-7-sonnet-20250219-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 75.101, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.101", + "raw_stderr": "1.039", + "latency": "69.581", + "cost_per_test": "0.051028", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.039, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-3-7-sonnet-20250219-thinking/de890687-bdb0-43b2-a0b7-db23b445d4a8.json b/data/vals_ai/anthropic/claude-3-7-sonnet-20250219-thinking/de890687-bdb0-43b2-a0b7-db23b445d4a8.json new file mode 100644 index 0000000000000000000000000000000000000000..8e30bdfaf709eb423dbbf0b7b5a226b6c95fb313 --- /dev/null +++ b/data/vals_ai/anthropic/claude-3-7-sonnet-20250219-thinking/de890687-bdb0-43b2-a0b7-db23b445d4a8.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/anthropic_claude-3-7-sonnet-20250219-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-3-7-sonnet-20250219-thinking", + "id": "anthropic/claude-3-7-sonnet-20250219-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-3-7-sonnet-20250219-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.447, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.447", + "raw_stderr": "0.689", + "latency": "56.858", + "cost_per_test": "0.072221", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.689, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 22.286, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "22.286", + "raw_stderr": "2.224", + "latency": "128.992", + "cost_per_test": "0.166043", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.224, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 60.574, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "60.574", + "raw_stderr": "2.497", + "latency": "103.562", + "cost_per_test": "0.135941", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.497, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 60.436, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "60.436", + "raw_stderr": "1.138", + "latency": "96.471", + "cost_per_test": "0.124735", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.138, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-3-7-sonnet-20250219-thinking/f5e12801-f17e-45de-a182-eb19282a42f4.json b/data/vals_ai/anthropic/claude-3-7-sonnet-20250219-thinking/f5e12801-f17e-45de-a182-eb19282a42f4.json new file mode 100644 index 0000000000000000000000000000000000000000..d93942c5014969ef82c27d5f0cacd095ff64330c --- /dev/null +++ b/data/vals_ai/anthropic/claude-3-7-sonnet-20250219-thinking/f5e12801-f17e-45de-a182-eb19282a42f4.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/anthropic_claude-3-7-sonnet-20250219-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-3-7-sonnet-20250219-thinking", + "id": "anthropic/claude-3-7-sonnet-20250219-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-3-7-sonnet-20250219-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.237, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.237", + "raw_stderr": "1.108", + "latency": "24.131", + "cost_per_test": "0.030189", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.108, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.847, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.847", + "raw_stderr": "1.121", + "latency": "27.822", + "cost_per_test": "0.035756", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.121, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.516, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.516", + "raw_stderr": "0.948", + "latency": "44.105", + "cost_per_test": "0.060384", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.948, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.878, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.878", + "raw_stderr": "1.769", + "latency": "34.525", + "cost_per_test": "0.044185", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.769, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.493, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.493", + "raw_stderr": "1.177", + "latency": "21.682", + "cost_per_test": "0.026329", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.177, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.787, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.787", + "raw_stderr": "1.413", + "latency": "67.263", + "cost_per_test": "0.087638", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.413, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.34, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.34", + "raw_stderr": "1.416", + "latency": "24.979", + "cost_per_test": "0.028831", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.416, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.278, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.278", + "raw_stderr": "2.239", + "latency": "19.362", + "cost_per_test": "0.024612", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.239, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 67.575, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "67.575", + "raw_stderr": "1.411", + "latency": "32.762", + "cost_per_test": "0.036944", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.411, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.414, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.414", + "raw_stderr": "0.762", + "latency": "40.166", + "cost_per_test": "0.05626", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.762, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.225, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.225", + "raw_stderr": "1.229", + "latency": "19.935", + "cost_per_test": "0.022639", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.229, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.728, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.728", + "raw_stderr": "0.37", + "latency": "31.744", + "cost_per_test": "0.040182", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.37, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.76, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.76", + "raw_stderr": "1.799", + "latency": "24.985", + "cost_per_test": "0.029247", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.799, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.76, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.76", + "raw_stderr": "0.909", + "latency": "44.271", + "cost_per_test": "0.058227", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.909, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:anthropic/claude-3-7-sonnet-20250219-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.08, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.08", + "raw_stderr": "1.358", + "latency": "18.433", + "cost_per_test": "0.021302", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.358, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-3-7-sonnet-20250219/3bd5ce94-a423-482b-8f37-b21fa8ef9e4c.json b/data/vals_ai/anthropic/claude-3-7-sonnet-20250219/3bd5ce94-a423-482b-8f37-b21fa8ef9e4c.json new file mode 100644 index 0000000000000000000000000000000000000000..e0704c5864b4651638cc0e3b901cf413c0120cf3 --- /dev/null +++ b/data/vals_ai/anthropic/claude-3-7-sonnet-20250219/3bd5ce94-a423-482b-8f37-b21fa8ef9e4c.json @@ -0,0 +1,101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/anthropic_claude-3-7-sonnet-20250219/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-3-7-sonnet-20250219", + "id": "anthropic/claude-3-7-sonnet-20250219", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-3-7-sonnet-20250219", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 71.519, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.519", + "raw_stderr": "1.085", + "latency": "8.323", + "cost_per_test": "0.012358", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.085, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-3-7-sonnet-20250219/449d5096-126b-44b7-896a-2f0e6241b520.json b/data/vals_ai/anthropic/claude-3-7-sonnet-20250219/449d5096-126b-44b7-896a-2f0e6241b520.json new file mode 100644 index 0000000000000000000000000000000000000000..d728ef54d21f391388f51eb5863c515eef6f6c17 --- /dev/null +++ b/data/vals_ai/anthropic/claude-3-7-sonnet-20250219/449d5096-126b-44b7-896a-2f0e6241b520.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/anthropic_claude-3-7-sonnet-20250219/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-3-7-sonnet-20250219", + "id": "anthropic/claude-3-7-sonnet-20250219", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-3-7-sonnet-20250219", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.703, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.703", + "raw_stderr": "1.182", + "latency": "6.711", + "cost_per_test": "0.009595", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.182, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.199, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.199", + "raw_stderr": "1.189", + "latency": "5.263", + "cost_per_test": "0.008106", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.189, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.686, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.686", + "raw_stderr": "1.125", + "latency": "6.869", + "cost_per_test": "0.012016", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.125, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.659, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.659", + "raw_stderr": "1.826", + "latency": "6.094", + "cost_per_test": "0.010214", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.826, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.9, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.9", + "raw_stderr": "1.198", + "latency": "5.639", + "cost_per_test": "0.008379", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.198, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 66.77, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.77", + "raw_stderr": "1.513", + "latency": "8.402", + "cost_per_test": "0.012613", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.513, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.362, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.362", + "raw_stderr": "1.44", + "latency": "6.152", + "cost_per_test": "0.008436", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.44, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.165, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.165", + "raw_stderr": "2.151", + "latency": "6.155", + "cost_per_test": "0.011353", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.151, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 66.848, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.848", + "raw_stderr": "1.419", + "latency": "6.963", + "cost_per_test": "0.010008", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.419, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.158, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.158", + "raw_stderr": "0.94", + "latency": "6.24", + "cost_per_test": "0.012018", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.94, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.169, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.169", + "raw_stderr": "1.286", + "latency": "5.192", + "cost_per_test": "0.006876", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.286, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.663, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.663", + "raw_stderr": "0.383", + "latency": "6.302", + "cost_per_test": "0.009664", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.383, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.154, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.154", + "raw_stderr": "1.879", + "latency": "6.058", + "cost_per_test": "0.007848", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.879, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.296, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.296", + "raw_stderr": "1.01", + "latency": "6.592", + "cost_per_test": "0.009941", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.01, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.208, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.208", + "raw_stderr": "1.323", + "latency": "5.9", + "cost_per_test": "0.007899", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.323, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-3-7-sonnet-20250219/4ce6baee-b64c-4a55-84ee-0bb979600801.json b/data/vals_ai/anthropic/claude-3-7-sonnet-20250219/4ce6baee-b64c-4a55-84ee-0bb979600801.json new file mode 100644 index 0000000000000000000000000000000000000000..72a60cbd061450dfc0294b4abeb8947e5935f53c --- /dev/null +++ b/data/vals_ai/anthropic/claude-3-7-sonnet-20250219/4ce6baee-b64c-4a55-84ee-0bb979600801.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/anthropic_claude-3-7-sonnet-20250219/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-3-7-sonnet-20250219", + "id": "anthropic/claude-3-7-sonnet-20250219", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-3-7-sonnet-20250219", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 96.584, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "96.584", + "raw_stderr": "1.012", + "latency": "6.162", + "cost_per_test": "0.007004", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.012, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 18.571, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "18.571", + "raw_stderr": "2.079", + "latency": "27.525", + "cost_per_test": "0.03584", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.079, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 54.83, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "54.83", + "raw_stderr": "2.543", + "latency": "15.01", + "cost_per_test": "0.01881", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.543, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 56.662, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "56.662", + "raw_stderr": "1.146", + "latency": "16.232", + "cost_per_test": "0.020551", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.146, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-3-7-sonnet-20250219/9efe604c-224a-40b3-bd2a-37c045384635.json b/data/vals_ai/anthropic/claude-3-7-sonnet-20250219/9efe604c-224a-40b3-bd2a-37c045384635.json new file mode 100644 index 0000000000000000000000000000000000000000..7f9b8f122fad29ed468abc2785ddda38cad89fd3 --- /dev/null +++ b/data/vals_ai/anthropic/claude-3-7-sonnet-20250219/9efe604c-224a-40b3-bd2a-37c045384635.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/anthropic_claude-3-7-sonnet-20250219/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-3-7-sonnet-20250219", + "id": "anthropic/claude-3-7-sonnet-20250219", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-3-7-sonnet-20250219", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 77.742, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "77.742", + "raw_stderr": "2.297", + "latency": "5.183", + "cost_per_test": "0.015277", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.297, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 68.68, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.68", + "raw_stderr": "1.775", + "latency": "5.67", + "cost_per_test": "0.015796", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.775, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 59.618, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.618", + "raw_stderr": "2.707", + "latency": "6.156", + "cost_per_test": "0.016315", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.707, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-3-7-sonnet-20250219/b4ba4d44-8dea-422c-af12-4f0bbd2ce555.json b/data/vals_ai/anthropic/claude-3-7-sonnet-20250219/b4ba4d44-8dea-422c-af12-4f0bbd2ce555.json new file mode 100644 index 0000000000000000000000000000000000000000..b7fd13fd4ddddb3d60c4da7f7d086c2f90ec44a3 --- /dev/null +++ b/data/vals_ai/anthropic/claude-3-7-sonnet-20250219/b4ba4d44-8dea-422c-af12-4f0bbd2ce555.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/anthropic_claude-3-7-sonnet-20250219/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-3-7-sonnet-20250219", + "id": "anthropic/claude-3-7-sonnet-20250219", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-3-7-sonnet-20250219", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 61.897, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.897", + "raw_stderr": "1.389", + "latency": "6.28", + "cost_per_test": "0.006427", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.389, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 72.404, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "72.404", + "raw_stderr": "0.879", + "latency": "6.28", + "cost_per_test": "0.006427", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.879, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 82.911, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "82.911", + "raw_stderr": "1.076", + "latency": "6.28", + "cost_per_test": "0.006427", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.076, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-3-7-sonnet-20250219/f8396979-fc30-4946-9f5c-1732fa693243.json b/data/vals_ai/anthropic/claude-3-7-sonnet-20250219/f8396979-fc30-4946-9f5c-1732fa693243.json new file mode 100644 index 0000000000000000000000000000000000000000..b775315da5ae0cc854dfd4c185259cde91e72fd8 --- /dev/null +++ b/data/vals_ai/anthropic/claude-3-7-sonnet-20250219/f8396979-fc30-4946-9f5c-1732fa693243.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/anthropic_claude-3-7-sonnet-20250219/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-3-7-sonnet-20250219", + "id": "anthropic/claude-3-7-sonnet-20250219", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-3-7-sonnet-20250219", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 68.687, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "68.687", + "raw_stderr": "3.296", + "latency": "9.856", + "cost_per_test": "0.015357", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.296, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 66.162, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.162", + "raw_stderr": "3.363", + "latency": "8.579", + "cost_per_test": "0.008588", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.363, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 67.424, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "67.424", + "raw_stderr": "2.354", + "latency": "9.218", + "cost_per_test": "0.011972", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.354, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-3-7-sonnet-20250219/ff969f3c-8090-4652-84eb-5e91788a6b17.json b/data/vals_ai/anthropic/claude-3-7-sonnet-20250219/ff969f3c-8090-4652-84eb-5e91788a6b17.json new file mode 100644 index 0000000000000000000000000000000000000000..93e5886ef30951212af24b63b59db26f6e08a02b --- /dev/null +++ b/data/vals_ai/anthropic/claude-3-7-sonnet-20250219/ff969f3c-8090-4652-84eb-5e91788a6b17.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/anthropic_claude-3-7-sonnet-20250219/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-3-7-sonnet-20250219", + "id": "anthropic/claude-3-7-sonnet-20250219", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-3-7-sonnet-20250219", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.482, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.482", + "raw_stderr": "0.826", + "latency": "1.108", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.826, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.452, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.452", + "raw_stderr": "0.202", + "latency": "1.109", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.202, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.521, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.521", + "raw_stderr": "0.656", + "latency": "1.198", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.656, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.001, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.001", + "raw_stderr": "0.484", + "latency": "1.199", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.484, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.422, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.422", + "raw_stderr": "1.055", + "latency": "1.084", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.055, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:anthropic/claude-3-7-sonnet-20250219:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.128, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.128", + "raw_stderr": "1.896", + "latency": "1.495", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.896, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-fable-5/054ba24e-f39f-47c0-b953-53c2c545117f.json b/data/vals_ai/anthropic/claude-fable-5/054ba24e-f39f-47c0-b953-53c2c545117f.json new file mode 100644 index 0000000000000000000000000000000000000000..2af1431f87665280e46ef1dce77e315f30a938bf --- /dev/null +++ b/data/vals_ai/anthropic/claude-fable-5/054ba24e-f39f-47c0-b953-53c2c545117f.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/anthropic_claude-fable-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-fable-5", + "id": "anthropic/claude-fable-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-fable-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 44.444, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "44.444", + "raw_stderr": "6.24", + "latency": "121.425", + "cost_per_test": "0.6238", + "temperature": "1", + "max_output_tokens": "30000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.24, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "sage:automata:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 39.682, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "39.682", + "raw_stderr": "6.183", + "latency": "152.558", + "cost_per_test": "0.779331", + "temperature": "1", + "max_output_tokens": "30000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.183, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 71.544, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "71.544", + "raw_stderr": "5.182", + "latency": "76.765", + "cost_per_test": "0.55329", + "temperature": "1", + "max_output_tokens": "30000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.182, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "sage:overall:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 51.89, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "51.89", + "raw_stderr": "3.4", + "latency": "116.916", + "cost_per_test": "0.65214", + "temperature": "1", + "max_output_tokens": "30000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.4, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-fable-5/1cca33e4-0187-4ea2-a87f-ce93dc499fec.json b/data/vals_ai/anthropic/claude-fable-5/1cca33e4-0187-4ea2-a87f-ce93dc499fec.json new file mode 100644 index 0000000000000000000000000000000000000000..d827646b6481252698af430d0146a1ece70d7ca6 --- /dev/null +++ b/data/vals_ai/anthropic/claude-fable-5/1cca33e4-0187-4ea2-a87f-ce93dc499fec.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/anthropic_claude-fable-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-fable-5", + "id": "anthropic/claude-fable-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-fable-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 76.709, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "76.709", + "raw_stderr": "1.192", + "latency": "17.2", + "cost_per_test": "0.183368", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.192, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 68.919, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.919", + "raw_stderr": "0.909", + "latency": "16.096", + "cost_per_test": "0.175013", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.909, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 61.129, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.129", + "raw_stderr": "1.374", + "latency": "14.993", + "cost_per_test": "0.166658", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.374, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-fable-5/2c094d3e-89e1-4404-9aa7-3a21a196a453.json b/data/vals_ai/anthropic/claude-fable-5/2c094d3e-89e1-4404-9aa7-3a21a196a453.json new file mode 100644 index 0000000000000000000000000000000000000000..143b9d2bff0d7ec448679eee13ef6f1cfcb4154f --- /dev/null +++ b/data/vals_ai/anthropic/claude-fable-5/2c094d3e-89e1-4404-9aa7-3a21a196a453.json @@ -0,0 +1,3486 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/hlab/anthropic_claude-fable-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Harvey's Legal Agent Benchmark", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-fable-5", + "id": "anthropic/claude-fable-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-fable-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "hlab:antitrust_competition:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1501.226", + "cost_per_test": "16.832828", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:arbitration_international_dispute_resolution:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1715.527", + "cost_per_test": "22.492624", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:banking_finance:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1390.381", + "cost_per_test": "15.803642", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:bankruptcy_restructuring:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1683.265", + "cost_per_test": "23.012162", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:capital_markets:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1359.531", + "cost_per_test": "17.509413", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:corporate_governance:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "1644.857", + "cost_per_test": "18.661957", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:corporate_ma:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "1471.023", + "cost_per_test": "18.410087", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Criteria Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Criteria Pass Rate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 90.48, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate", + "task_name": "Criteria Pass Rate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.48", + "raw_stderr": "0.354", + "latency": "1613.04", + "cost_per_test": "19.225253", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.354, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__antitrust_competition:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 92.035, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "92.035", + "raw_stderr": "1.805", + "latency": "1501.226", + "cost_per_test": "16.832828", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.805, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__arbitration_international_dispute_resolution:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 91.398, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "91.398", + "raw_stderr": "1.682", + "latency": "1715.527", + "cost_per_test": "22.492624", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.682, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__banking_finance:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 92.565, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "92.565", + "raw_stderr": "1.602", + "latency": "1390.381", + "cost_per_test": "15.803642", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.602, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__bankruptcy_restructuring:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 84.429, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.429", + "raw_stderr": "2.137", + "latency": "1683.265", + "cost_per_test": "23.012162", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.137, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__capital_markets:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 91.372, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "91.372", + "raw_stderr": "1.762", + "latency": "1359.531", + "cost_per_test": "17.509413", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.762, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_governance:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 92.123, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "92.123", + "raw_stderr": "1.579", + "latency": "1644.857", + "cost_per_test": "18.661957", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.579, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_ma:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 91.987, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "91.987", + "raw_stderr": "1.539", + "latency": "1471.023", + "cost_per_test": "18.410087", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.539, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__data_privacy_cybersecurity:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 94.346, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "94.346", + "raw_stderr": "1.375", + "latency": "1595.649", + "cost_per_test": "20.041934", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.375, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__emerging_companies_venture_capital:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 93.857, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "93.857", + "raw_stderr": "1.405", + "latency": "1722.228", + "cost_per_test": "17.394862", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.405, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__employment_labor:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 87.64, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.64", + "raw_stderr": "2.018", + "latency": "1459.55", + "cost_per_test": "14.366748", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.018, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__energy_natural_resources:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 96.04, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "96.04", + "raw_stderr": "1.122", + "latency": "1808.583", + "cost_per_test": "32.1783", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.122, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__environmental_esg:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 87.814, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.814", + "raw_stderr": "1.962", + "latency": "1679.817", + "cost_per_test": "17.039801", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.962, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__funds_asset_management:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 84.84, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.84", + "raw_stderr": "1.939", + "latency": "2223.169", + "cost_per_test": "32.834156", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.939, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__healthcare_life_sciences:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 91.419, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "91.419", + "raw_stderr": "1.612", + "latency": "1645.04", + "cost_per_test": "23.076834", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.612, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__immigration:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 90.182, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.182", + "raw_stderr": "1.798", + "latency": "1489.501", + "cost_per_test": "14.198276", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.798, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__insurance:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 93.407, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "93.407", + "raw_stderr": "1.505", + "latency": "1417.023", + "cost_per_test": "15.611186", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.505, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__intellectual_property:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 92.58, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "92.58", + "raw_stderr": "1.561", + "latency": "1536.356", + "cost_per_test": "19.843555", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.561, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__international_trade_sanctions:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 92.46, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "92.46", + "raw_stderr": "1.667", + "latency": "1640.229", + "cost_per_test": "15.317586", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.667, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__litigation_dispute_resolution:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 89.139, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.139", + "raw_stderr": "1.908", + "latency": "1467.627", + "cost_per_test": "16.209915", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.908, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__real_estate:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 93.294, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "93.294", + "raw_stderr": "1.352", + "latency": "1678.201", + "cost_per_test": "21.067576", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.352, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__structured_finance_securitization:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 87.302, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.302", + "raw_stderr": "1.879", + "latency": "1580.749", + "cost_per_test": "17.134562", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.879, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__tax:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 83.612, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.612", + "raw_stderr": "2.144", + "latency": "1763.374", + "cost_per_test": "17.075749", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.144, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__trusts_estates_private_client:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 91.985, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "91.985", + "raw_stderr": "1.681", + "latency": "1662.67", + "cost_per_test": "18.186942", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.681, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__white_collar_defense_investigations:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 87.205, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.205", + "raw_stderr": "1.942", + "latency": "1555.025", + "cost_per_test": "16.626886", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.942, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:data_privacy_cybersecurity:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1595.649", + "cost_per_test": "20.041934", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:emerging_companies_venture_capital:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "1722.228", + "cost_per_test": "17.394862", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:employment_labor:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1459.55", + "cost_per_test": "14.366748", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:energy_natural_resources:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "1808.583", + "cost_per_test": "32.1783", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:environmental_esg:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1679.817", + "cost_per_test": "17.039801", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:funds_asset_management:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "2223.169", + "cost_per_test": "32.834156", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:healthcare_life_sciences:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "1645.04", + "cost_per_test": "23.076834", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:immigration:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1489.501", + "cost_per_test": "14.198276", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:insurance:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1417.023", + "cost_per_test": "15.611186", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:intellectual_property:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1536.356", + "cost_per_test": "19.843555", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:international_trade_sanctions:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1640.229", + "cost_per_test": "15.317586", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:litigation_dispute_resolution:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "1467.627", + "cost_per_test": "16.209915", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:overall:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.overall", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Overall).", + "metric_id": "vals_ai.hlab.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 11.25, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "11.25", + "raw_stderr": "2.166", + "latency": "1613.04", + "cost_per_test": "19.225253", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.166, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:real_estate:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1678.201", + "cost_per_test": "21.067576", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:structured_finance_securitization:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1580.749", + "cost_per_test": "17.134562", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:tax:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1763.374", + "cost_per_test": "17.075749", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:trusts_estates_private_client:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "1662.67", + "cost_per_test": "18.186942", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:white_collar_defense_investigations:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.hlab.white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1555.025", + "cost_per_test": "16.626886", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-fable-5/401914f7-a623-459e-abfe-3cf700f5497e.json b/data/vals_ai/anthropic/claude-fable-5/401914f7-a623-459e-abfe-3cf700f5497e.json new file mode 100644 index 0000000000000000000000000000000000000000..1729a6bfee39d51285d1ccd8b1b26bb2d5a30a7d --- /dev/null +++ b/data/vals_ai/anthropic/claude-fable-5/401914f7-a623-459e-abfe-3cf700f5497e.json @@ -0,0 +1,260 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/anthropic_claude-fable-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-fable-5", + "id": "anthropic/claude-fable-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-fable-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 100.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "100", + "raw_stderr": "0", + "latency": "11225.953", + "cost_per_test": "9.538742", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 70.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "70", + "raw_stderr": "3.333", + "latency": "1496.794", + "cost_per_test": "1.271832", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.333, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 84.848, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "84.848", + "raw_stderr": "0.606", + "latency": "816.433", + "cost_per_test": "0.693727", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.606, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 80.524, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "80.524", + "raw_stderr": "1.35", + "latency": "504.537", + "cost_per_test": "1.429025", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.35, + "method": "vals_reported" + } + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-fable-5/45ef2f07-02b8-4045-b587-55fb751b4d09.json b/data/vals_ai/anthropic/claude-fable-5/45ef2f07-02b8-4045-b587-55fb751b4d09.json new file mode 100644 index 0000000000000000000000000000000000000000..c2fd066c9dad15d0ef5b5ea926abb28470a3f70a --- /dev/null +++ b/data/vals_ai/anthropic/claude-fable-5/45ef2f07-02b8-4045-b587-55fb751b4d09.json @@ -0,0 +1,1071 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/anthropic_claude-fable-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-fable-5", + "id": "anthropic/claude-fable-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-fable-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.584, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.584", + "raw_stderr": "0.915", + "latency": "18.862", + "cost_per_test": "0.086148", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.915, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.156, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.156", + "raw_stderr": "0.899", + "latency": "15.86", + "cost_per_test": "0.054887", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.899, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.905, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.905", + "raw_stderr": "0.711", + "latency": "39.622", + "cost_per_test": "0.094911", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.711, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.146, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.146", + "raw_stderr": "1.159", + "latency": "17.469", + "cost_per_test": "0.08492", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.159, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.839, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.839", + "raw_stderr": "0.828", + "latency": "12.804", + "cost_per_test": "0.052862", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.828, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.99, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.99", + "raw_stderr": "0.964", + "latency": "80.54", + "cost_per_test": "0.189415", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.964, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.597, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.597", + "raw_stderr": "1.262", + "latency": "20.082", + "cost_per_test": "0.084721", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.262, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.302, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.302", + "raw_stderr": "1.814", + "latency": "32.513", + "cost_per_test": "0.079154", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.814, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.467, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.467", + "raw_stderr": "1.031", + "latency": "23.395", + "cost_per_test": "0.104874", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.031, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 97.557, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "97.557", + "raw_stderr": "0.42", + "latency": "21.371", + "cost_per_test": "0.07686", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.42, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.827, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.827", + "raw_stderr": "0.994", + "latency": "14.291", + "cost_per_test": "0.052645", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.994, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.502, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.502", + "raw_stderr": "0.278", + "latency": "24.998", + "cost_per_test": "0.082276", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.278, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.581, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.581", + "raw_stderr": "1.308", + "latency": "24.035", + "cost_per_test": "0.069071", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.308, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 95.843, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "95.843", + "raw_stderr": "0.554", + "latency": "15.9", + "cost_per_test": "0.071684", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.554, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.231, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.231", + "raw_stderr": "0.948", + "latency": "13.228", + "cost_per_test": "0.04971", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.948, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-fable-5/480ea8f5-4e36-4085-9611-d5f137cd7152.json b/data/vals_ai/anthropic/claude-fable-5/480ea8f5-4e36-4085-9611-d5f137cd7152.json new file mode 100644 index 0000000000000000000000000000000000000000..d20dbecd8858bd986022e78fca2561eb534875f9 --- /dev/null +++ b/data/vals_ai/anthropic/claude-fable-5/480ea8f5-4e36-4085-9611-d5f137cd7152.json @@ -0,0 +1,450 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/anthropic_claude-fable-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-fable-5", + "id": "anthropic/claude-fable-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-fable-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 91.775, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "91.775", + "raw_stderr": "0.667", + "latency": "7.255", + "cost_per_test": "0.019734", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.667, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 86.863, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.863", + "raw_stderr": "0.163", + "latency": "8.119", + "cost_per_test": "0.028526", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.163, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 92.064, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "92.064", + "raw_stderr": "0.505", + "latency": "6.922", + "cost_per_test": "0.030005", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.505, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 88.561, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.561", + "raw_stderr": "0.331", + "latency": "8.964", + "cost_per_test": "0.030639", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.331, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.772, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.772", + "raw_stderr": "0.839", + "latency": "6.836", + "cost_per_test": "0.020967", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.839, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 88.33, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.33", + "raw_stderr": "1.145", + "latency": "15.686", + "cost_per_test": "0.053962", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.145, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-fable-5/4b936b2f-4e40-428b-8982-3b8f4e81a63a.json b/data/vals_ai/anthropic/claude-fable-5/4b936b2f-4e40-428b-8982-3b8f4e81a63a.json new file mode 100644 index 0000000000000000000000000000000000000000..cfa222a8252ff18c0dbeb473a35183357deaeb64 --- /dev/null +++ b/data/vals_ai/anthropic/claude-fable-5/4b936b2f-4e40-428b-8982-3b8f4e81a63a.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/anthropic_claude-fable-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-fable-5", + "id": "anthropic/claude-fable-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-fable-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 66.9, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.9", + "raw_stderr": "1.607", + "latency": "18.601", + "cost_per_test": "0.090437", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.607, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 77.04, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "77.04", + "raw_stderr": "1.436", + "latency": "81.472", + "cost_per_test": "3.734685", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.436, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 71.834, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.834", + "raw_stderr": "0.883", + "latency": "52.99", + "cost_per_test": "1.737573", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.883, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 71.562, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.562", + "raw_stderr": "1.54", + "latency": "58.898", + "cost_per_test": "1.387598", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.54, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-fable-5/52dcdbc1-dd7e-48a7-9ffa-91f1fe9b4bac.json b/data/vals_ai/anthropic/claude-fable-5/52dcdbc1-dd7e-48a7-9ffa-91f1fe9b4bac.json new file mode 100644 index 0000000000000000000000000000000000000000..9ce073e17240525fbfbae686a641d6d546f9976a --- /dev/null +++ b/data/vals_ai/anthropic/claude-fable-5/52dcdbc1-dd7e-48a7-9ffa-91f1fe9b4bac.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/anthropic_claude-fable-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-fable-5", + "id": "anthropic/claude-fable-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-fable-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 92.929, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.929", + "raw_stderr": "2.043", + "latency": "132.309", + "cost_per_test": "0.610919", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.043, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 93.434, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.434", + "raw_stderr": "3.296", + "latency": "67.496", + "cost_per_test": "0.290028", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.296, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 93.182, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.182", + "raw_stderr": "1.939", + "latency": "99.902", + "cost_per_test": "0.450473", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.939, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-fable-5/598f25f2-fac7-44aa-9851-6265adcbc3ee.json b/data/vals_ai/anthropic/claude-fable-5/598f25f2-fac7-44aa-9851-6265adcbc3ee.json new file mode 100644 index 0000000000000000000000000000000000000000..3d4200998567397143d4229abb6ef56c00bf24f0 --- /dev/null +++ b/data/vals_ai/anthropic/claude-fable-5/598f25f2-fac7-44aa-9851-6265adcbc3ee.json @@ -0,0 +1,381 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/anthropic_claude-fable-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-fable-5", + "id": "anthropic/claude-fable-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-fable-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 92.857, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "92.857", + "raw_stderr": "4.022", + "latency": "617.517", + "cost_per_test": "1.253749", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 4.022, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 94.636, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "94.636", + "raw_stderr": "1.397", + "latency": "377.223", + "cost_per_test": "0.643848", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.397, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 95.876, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "95.876", + "raw_stderr": "1.431", + "latency": "260.529", + "cost_per_test": "0.410248", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.431, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 100.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "100", + "raw_stderr": "0", + "latency": "1052.442", + "cost_per_test": "2.240184", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:overall:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 95.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "95", + "raw_stderr": "0.976", + "latency": "356.182", + "cost_per_test": "2.046736", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.976, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-fable-5/6ef4c50c-aec8-4a80-b742-ca8949023d03.json b/data/vals_ai/anthropic/claude-fable-5/6ef4c50c-aec8-4a80-b742-ca8949023d03.json new file mode 100644 index 0000000000000000000000000000000000000000..b114102f14802de0020e830cc14923045710e481 --- /dev/null +++ b/data/vals_ai/anthropic/claude-fable-5/6ef4c50c-aec8-4a80-b742-ca8949023d03.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/anthropic_claude-fable-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-fable-5", + "id": "anthropic/claude-fable-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-fable-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 65.822, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.822", + "raw_stderr": "1.356", + "latency": "56.865", + "cost_per_test": "0.220133", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.356, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 76.942, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "76.942", + "raw_stderr": "0.821", + "latency": "56.865", + "cost_per_test": "0.220133", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.821, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 88.062, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "88.062", + "raw_stderr": "0.927", + "latency": "56.865", + "cost_per_test": "0.220133", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.927, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-fable-5/755fd2f3-b277-41e0-ac3b-a53a6866aa65.json b/data/vals_ai/anthropic/claude-fable-5/755fd2f3-b277-41e0-ac3b-a53a6866aa65.json new file mode 100644 index 0000000000000000000000000000000000000000..907d429d9085c303e4a5d4b440e5ae2f264c4e66 --- /dev/null +++ b/data/vals_ai/anthropic/claude-fable-5/755fd2f3-b277-41e0-ac3b-a53a6866aa65.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/code-migration/anthropic_claude-fable-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Code Migration", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-fable-5", + "id": "anthropic/claude-fable-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-fable-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "code-migration:cli:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.code-migration.cli", + "source_data": { + "dataset_name": "Code Migration - CLI", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cli", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (CLI).", + "metric_id": "vals_ai.code-migration.cli.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 60.087, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cli", + "task_name": "CLI", + "dataset_type": "private", + "industry": "coding", + "raw_score": "60.087", + "raw_stderr": "2.841", + "latency": "7275.185", + "cost_per_test": "139.102342", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.841, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "code-migration:cobol:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.code-migration.cobol", + "source_data": { + "dataset_name": "Code Migration - COBOL", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cobol", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (COBOL).", + "metric_id": "vals_ai.code-migration.cobol.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 40.0, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cobol", + "task_name": "COBOL", + "dataset_type": "private", + "industry": "coding", + "raw_score": "40", + "raw_stderr": "16.33", + "latency": "4411.994", + "cost_per_test": "31.08541", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 16.33, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "code-migration:code_quality:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.code-migration.code_quality", + "source_data": { + "dataset_name": "Code Migration - Code Quality", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "code_quality", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Code Quality).", + "metric_id": "vals_ai.code-migration.code_quality.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 4.442, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "code_quality", + "task_name": "Code Quality", + "dataset_type": "private", + "industry": "coding", + "raw_score": "4.442", + "raw_stderr": "0.029", + "latency": "6559.388", + "cost_per_test": "112.098109", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.029, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "code-migration:overall:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.code-migration.overall", + "source_data": { + "dataset_name": "Code Migration - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Overall).", + "metric_id": "vals_ai.code-migration.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 55.065, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "55.065", + "raw_stderr": "4.605", + "latency": "6559.388", + "cost_per_test": "112.098109", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 4.605, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-fable-5/7cda0d11-2a87-4536-b869-86cbb0cf065a.json b/data/vals_ai/anthropic/claude-fable-5/7cda0d11-2a87-4536-b869-86cbb0cf065a.json new file mode 100644 index 0000000000000000000000000000000000000000..5402042ee8c03b8257055c8ade9612374620eabc --- /dev/null +++ b/data/vals_ai/anthropic/claude-fable-5/7cda0d11-2a87-4536-b869-86cbb0cf065a.json @@ -0,0 +1,405 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/anthropic_claude-fable-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-fable-5", + "id": "anthropic/claude-fable-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-fable-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 71.562, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "71.562", + "raw_stderr": "1.54", + "latency": "58.898", + "cost_per_test": "1.737573", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.54, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 56.349, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "56.349", + "raw_stderr": "0.831", + "latency": "537.752", + "cost_per_test": "8.061563", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.831, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_index:overall:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 75.145, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "75.145", + "raw_stderr": "0.637", + "latency": "1050.682", + "cost_per_test": "5.157861", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.637, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 97.059, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "97.059", + "raw_stderr": "1.673", + "latency": "441.198", + "cost_per_test": "2.046736" + }, + "uncertainty": { + "standard_error": { + "value": 1.673, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 80.524, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "80.524", + "raw_stderr": "1.35", + "latency": "504.537", + "cost_per_test": "1.429025", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.35, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 93.467, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "93.467", + "raw_stderr": "1.473", + "latency": "3711.026", + "cost_per_test": "12.514408", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.473, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-fable-5/7fe99d04-b55d-4ad6-b8fe-f41404c26be5.json b/data/vals_ai/anthropic/claude-fable-5/7fe99d04-b55d-4ad6-b8fe-f41404c26be5.json new file mode 100644 index 0000000000000000000000000000000000000000..a3a05a35eeb291a41e8a18ce71765c61e3cff97c --- /dev/null +++ b/data/vals_ai/anthropic/claude-fable-5/7fe99d04-b55d-4ad6-b8fe-f41404c26be5.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/anthropic_claude-fable-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-fable-5", + "id": "anthropic/claude-fable-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-fable-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 77.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "77", + "raw_stderr": "4.23", + "latency": "918.917", + "cost_per_test": "1.635455", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 4.23, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-fable-5/9a3910a0-f644-4096-b97a-ca7cbe4e385a.json b/data/vals_ai/anthropic/claude-fable-5/9a3910a0-f644-4096-b97a-ca7cbe4e385a.json new file mode 100644 index 0000000000000000000000000000000000000000..c8fcaa4cbecdf52f7f47cea6d2c8e2a9cefb06a2 --- /dev/null +++ b/data/vals_ai/anthropic/claude-fable-5/9a3910a0-f644-4096-b97a-ca7cbe4e385a.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/anthropic_claude-fable-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-fable-5", + "id": "anthropic/claude-fable-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-fable-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.758, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.758", + "raw_stderr": "0.617", + "latency": "12.692", + "cost_per_test": "0.033455", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.617, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "lcb:hard:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 79.714, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "79.714", + "raw_stderr": "2.149", + "latency": "281.848", + "cost_per_test": "1.056984", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.149, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "lcb:medium:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 90.862, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "90.862", + "raw_stderr": "1.472", + "latency": "61.049", + "cost_per_test": "0.194043", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.472, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "lcb:overall:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 89.778, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "89.778", + "raw_stderr": "0.892", + "latency": "118.53", + "cost_per_test": "0.42816", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.892, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-fable-5/9e0f216b-a714-40f4-b323-f8582b6076ce.json b/data/vals_ai/anthropic/claude-fable-5/9e0f216b-a714-40f4-b323-f8582b6076ce.json new file mode 100644 index 0000000000000000000000000000000000000000..6b6bb449823630f0a08d37b5098a36e100594162 --- /dev/null +++ b/data/vals_ai/anthropic/claude-fable-5/9e0f216b-a714-40f4-b323-f8582b6076ce.json @@ -0,0 +1,762 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/anthropic_claude-fable-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-fable-5", + "id": "anthropic/claude-fable-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-fable-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 51.092, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "51.092", + "raw_stderr": "0.338", + "latency": "644.817", + "cost_per_test": "7.417657", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.338, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 45.694, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "45.694", + "raw_stderr": "0.871", + "latency": "610.952", + "cost_per_test": "8.061563", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.871, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 43.986, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "43.986", + "raw_stderr": "0.722", + "latency": "812.322", + "cost_per_test": "10.480979", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.722, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 68.82, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.82", + "raw_stderr": "2.01", + "latency": "887.51", + "cost_per_test": "11.594564", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.01, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 77.263, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "77.263", + "raw_stderr": "1.276", + "latency": "369.254", + "cost_per_test": "2.990129", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.276, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 17.609, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "17.609", + "raw_stderr": "1.578", + "latency": "599.956", + "cost_per_test": "7.073379", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.578, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 70.324, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "70.324", + "raw_stderr": "0.379", + "latency": "1140.824", + "cost_per_test": "11.155704", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.379, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 80.751, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "80.751", + "raw_stderr": "1.207", + "latency": "709.258", + "cost_per_test": "8.098129", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.207, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 68.882, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.882", + "raw_stderr": "2.144", + "latency": "474.571", + "cost_per_test": "4.244023", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.144, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:overall:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 56.314, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.314", + "raw_stderr": "0.84", + "latency": "610.952", + "cost_per_test": "8.061563", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.84, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 28.097, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "28.097", + "raw_stderr": "2.764", + "latency": "890.071", + "cost_per_test": "9.631473", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.764, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-fable-5/a97bcef7-24c8-4262-a167-0ccbb0d013bc.json b/data/vals_ai/anthropic/claude-fable-5/a97bcef7-24c8-4262-a167-0ccbb0d013bc.json new file mode 100644 index 0000000000000000000000000000000000000000..ad7a0f48370943ee774f610397299ec6d2f9f93d --- /dev/null +++ b/data/vals_ai/anthropic/claude-fable-5/a97bcef7-24c8-4262-a167-0ccbb0d013bc.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/anthropic_claude-fable-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-fable-5", + "id": "anthropic/claude-fable-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-fable-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 90.352, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "90.352", + "raw_stderr": "2.097", + "latency": "3711.026", + "cost_per_test": "12.514408", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.097, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-fable-5/af40c132-c7d0-4ac7-bdf8-72f0acda82e8.json b/data/vals_ai/anthropic/claude-fable-5/af40c132-c7d0-4ac7-bdf8-72f0acda82e8.json new file mode 100644 index 0000000000000000000000000000000000000000..daaf3e9de395d1f3f73096557bfcd0654e9f8fbe --- /dev/null +++ b/data/vals_ai/anthropic/claude-fable-5/af40c132-c7d0-4ac7-bdf8-72f0acda82e8.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/programbench/anthropic_claude-fable-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProgramBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-fable-5", + "id": "anthropic/claude-fable-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-fable-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "programbench:almost:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.programbench.almost", + "source_data": { + "dataset_name": "ProgramBench - Almost Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "almost", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Almost Resolved).", + "metric_id": "vals_ai.programbench.almost.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 33.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "almost", + "task_name": "Almost Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33", + "raw_stderr": "3.333", + "latency": "9447.46", + "cost_per_test": "75.675797", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "programbench:overall:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.programbench.overall", + "source_data": { + "dataset_name": "ProgramBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Overall).", + "metric_id": "vals_ai.programbench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 2.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "2", + "raw_stderr": "0.992", + "latency": "9447.46", + "cost_per_test": "75.675797", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.992, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "programbench:partial:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.programbench.partial", + "source_data": { + "dataset_name": "ProgramBench - Raw Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "partial", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Raw Pass Rate).", + "metric_id": "vals_ai.programbench.partial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 76.803, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "partial", + "task_name": "Raw Pass Rate", + "dataset_type": "public", + "industry": "coding", + "raw_score": "76.803", + "raw_stderr": "1.914", + "latency": "9447.46", + "cost_per_test": "75.675797", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.914, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "programbench:strict:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.programbench.strict", + "source_data": { + "dataset_name": "ProgramBench - Fully Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "strict", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Fully Resolved).", + "metric_id": "vals_ai.programbench.strict.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 2.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "strict", + "task_name": "Fully Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "2", + "raw_stderr": "0.992", + "latency": "9447.46", + "cost_per_test": "75.675797", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.992, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-fable-5/db158c70-4d46-4e93-b333-22b53c5688cb.json b/data/vals_ai/anthropic/claude-fable-5/db158c70-4d46-4e93-b333-22b53c5688cb.json new file mode 100644 index 0000000000000000000000000000000000000000..d5005834b9f936db562dd75b23be9461753135d6 --- /dev/null +++ b/data/vals_ai/anthropic/claude-fable-5/db158c70-4d46-4e93-b333-22b53c5688cb.json @@ -0,0 +1,105 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/anthropic_claude-fable-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-fable-5", + "id": "anthropic/claude-fable-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-fable-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 89.306, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.306", + "raw_stderr": "0.743", + "latency": "61.44", + "cost_per_test": "0.16064", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.743, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-fable-5/db78c6b9-89b9-4e63-9e60-492517537e65.json b/data/vals_ai/anthropic/claude-fable-5/db78c6b9-89b9-4e63-9e60-492517537e65.json new file mode 100644 index 0000000000000000000000000000000000000000..86581f185c9b9286d71a6a981a3b75222ea3a254 --- /dev/null +++ b/data/vals_ai/anthropic/claude-fable-5/db78c6b9-89b9-4e63-9e60-492517537e65.json @@ -0,0 +1,537 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_multimodal_index/anthropic_claude-fable-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Multimodal Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-fable-5", + "id": "anthropic/claude-fable-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-fable-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_multimodal_index:corp_fin_v2:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Multimodal Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (CorpFin v2).", + "metric_id": "vals_ai.vals_multimodal_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 71.562, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "71.562", + "raw_stderr": "1.54", + "latency": "58.898", + "cost_per_test": "1.737573", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.54, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:finance_agent:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.finance_agent", + "source_data": { + "dataset_name": "Vals Multimodal Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_multimodal_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 56.349, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "56.349", + "raw_stderr": "0.831", + "latency": "537.752", + "cost_per_test": "8.061563", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.831, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:mortgage_tax:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.mortgage_tax", + "source_data": { + "dataset_name": "Vals Multimodal Index - Mortgage Tax", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "mortgage_tax", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Mortgage Tax).", + "metric_id": "vals_ai.vals_multimodal_index.mortgage_tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 68.919, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "mortgage_tax", + "task_name": "Mortgage Tax", + "dataset_type": "private", + "industry": "index", + "raw_score": "68.919", + "raw_stderr": "0.909", + "latency": "16.096", + "cost_per_test": "0.175013", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.909, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:overall:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.overall", + "source_data": { + "dataset_name": "Vals Multimodal Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Overall).", + "metric_id": "vals_ai.vals_multimodal_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 74.154, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "74.154", + "raw_stderr": "0.567", + "latency": "769.489", + "cost_per_test": "3.816553", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.567, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:sage:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.sage", + "source_data": { + "dataset_name": "Vals Multimodal Index - SAGE", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "sage", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SAGE).", + "metric_id": "vals_ai.vals_multimodal_index.sage.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 51.89, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "sage", + "task_name": "SAGE", + "dataset_type": "private", + "industry": "index", + "raw_score": "51.89", + "raw_stderr": "3.4", + "latency": "116.916", + "cost_per_test": "0.618162", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.4, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:swebench:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.swebench", + "source_data": { + "dataset_name": "Vals Multimodal Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SWE-bench).", + "metric_id": "vals_ai.vals_multimodal_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 97.059, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "97.059", + "raw_stderr": "1.673", + "latency": "441.198", + "cost_per_test": "2.046736" + }, + "uncertainty": { + "standard_error": { + "value": 1.673, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:terminal_bench_2_1:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Multimodal Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_multimodal_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 80.524, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "80.524", + "raw_stderr": "1.35", + "latency": "504.537", + "cost_per_test": "1.429025", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.35, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:vibe_code_bench:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Multimodal Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_multimodal_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 93.467, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "93.467", + "raw_stderr": "1.473", + "latency": "3711.026", + "cost_per_test": "12.514408", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.473, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-fable-5/f78b403f-e233-4f22-8d7e-fe67381de012.json b/data/vals_ai/anthropic/claude-fable-5/f78b403f-e233-4f22-8d7e-fe67381de012.json new file mode 100644 index 0000000000000000000000000000000000000000..da828243cbde17ea4cbef9767879921a1034d18f --- /dev/null +++ b/data/vals_ai/anthropic/claude-fable-5/f78b403f-e233-4f22-8d7e-fe67381de012.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/anthropic_claude-fable-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-fable-5", + "id": "anthropic/claude-fable-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-fable-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 88.522, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "88.522", + "raw_stderr": "1.945", + "latency": "119.475", + "cost_per_test": "0.583239", + "temperature": "1", + "max_output_tokens": "30000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.945, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-fable-5/f90f3ac0-4bb5-43d0-be5d-747f22ae57c5.json b/data/vals_ai/anthropic/claude-fable-5/f90f3ac0-4bb5-43d0-be5d-747f22ae57c5.json new file mode 100644 index 0000000000000000000000000000000000000000..d38b72ee9a8da75702daa363e5cb83c178e05a16 --- /dev/null +++ b/data/vals_ai/anthropic/claude-fable-5/f90f3ac0-4bb5-43d0-be5d-747f22ae57c5.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/anthropic_claude-fable-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-fable-5", + "id": "anthropic/claude-fable-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-fable-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 100.0, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "100", + "raw_stderr": "0", + "latency": "2540.44", + "cost_per_test": "15.68634", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 44.5, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "44.5", + "raw_stderr": "18.714", + "latency": "4097.498", + "cost_per_test": "14.517016", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 18.714, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "ioi:overall:anthropic/claude-fable-5:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 72.25, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "72.25", + "raw_stderr": "9.357", + "latency": "3318.969", + "cost_per_test": "15.101678", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 9.357, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/2324d26a-636c-4f7c-943f-ed34e984fa4b.json b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/2324d26a-636c-4f7c-943f-ed34e984fa4b.json new file mode 100644 index 0000000000000000000000000000000000000000..285da6477fe55e9cd91ee7e1e1ef4ec10b5fc887 --- /dev/null +++ b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/2324d26a-636c-4f7c-943f-ed34e984fa4b.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/anthropic_claude-haiku-4-5-20251001-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-haiku-4-5-20251001-thinking", + "id": "anthropic/claude-haiku-4-5-20251001-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-haiku-4-5-20251001-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 85.23, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "85.23", + "raw_stderr": "1.899", + "latency": "66.197", + "cost_per_test": "0.042375", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.899, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/32fb05ef-ace1-4669-9539-caf736c37e40.json b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/32fb05ef-ace1-4669-9539-caf736c37e40.json new file mode 100644 index 0000000000000000000000000000000000000000..fbbe83eba1590e90123175efbd2e7fb86b1ef6f5 --- /dev/null +++ b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/32fb05ef-ace1-4669-9539-caf736c37e40.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/anthropic_claude-haiku-4-5-20251001-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-haiku-4-5-20251001-thinking", + "id": "anthropic/claude-haiku-4-5-20251001-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-haiku-4-5-20251001-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 67.091, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.091", + "raw_stderr": "1.325", + "latency": "42.056", + "cost_per_test": "0.028418", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.325, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 62.162, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.162", + "raw_stderr": "0.962", + "latency": "29.792", + "cost_per_test": "0.020742", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.962, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 57.234, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.234", + "raw_stderr": "1.395", + "latency": "17.528", + "cost_per_test": "0.013065", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.395, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/34f1d376-102f-42f2-9c0a-393f01a09540.json b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/34f1d376-102f-42f2-9c0a-393f01a09540.json new file mode 100644 index 0000000000000000000000000000000000000000..a3c99513bcceb324e3ce9115060eab2f18eea598 --- /dev/null +++ b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/34f1d376-102f-42f2-9c0a-393f01a09540.json @@ -0,0 +1,346 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/anthropic_claude-haiku-4-5-20251001-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-haiku-4-5-20251001-thinking", + "id": "anthropic/claude-haiku-4-5-20251001-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-haiku-4-5-20251001-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 19.048, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "19.048", + "raw_stderr": "4.928", + "latency": "73.227", + "cost_per_test": "0.052613", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 4.928, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 27.778, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "27.778", + "raw_stderr": "5.711", + "latency": "61.3", + "cost_per_test": "0.045446", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.711, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 49.383, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "49.383", + "raw_stderr": "6.324", + "latency": "62.372", + "cost_per_test": "0.097577", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.324, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 48.641, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "48.641", + "raw_stderr": "5.614", + "latency": "64.809", + "cost_per_test": "0.06133", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.614, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 31.822, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "31.822", + "raw_stderr": "3.134", + "latency": "66.445", + "cost_per_test": "0.05313", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.134, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/39ef2c8d-06b8-4f97-8ca1-250c607a58f7.json b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/39ef2c8d-06b8-4f97-8ca1-250c607a58f7.json new file mode 100644 index 0000000000000000000000000000000000000000..d2765769655ea7bf658bdeecd4fba18b4bb8cdb6 --- /dev/null +++ b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/39ef2c8d-06b8-4f97-8ca1-250c607a58f7.json @@ -0,0 +1,718 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/anthropic_claude-haiku-4-5-20251001-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-haiku-4-5-20251001-thinking", + "id": "anthropic/claude-haiku-4-5-20251001-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-haiku-4-5-20251001-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 30.462, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "30.462", + "raw_stderr": "0.789", + "latency": "182.053", + "cost_per_test": "0.675348", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.789, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 19.83, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "19.83", + "raw_stderr": "1.293", + "latency": "176.769", + "cost_per_test": "0.602129", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.293, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 31.453, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "31.453", + "raw_stderr": "1.208", + "latency": "203.88", + "cost_per_test": "0.700359", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.208, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 42.517, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "42.517", + "raw_stderr": "1.158", + "latency": "192.263", + "cost_per_test": "0.782271", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.158, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 26.692, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "26.692", + "raw_stderr": "1.753", + "latency": "110.802", + "cost_per_test": "0.305043", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.753, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 5.141, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "5.141", + "raw_stderr": "1.119", + "latency": "214.099", + "cost_per_test": "0.652418", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.119, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 45.11, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "45.11", + "raw_stderr": "2.169", + "latency": "208.374", + "cost_per_test": "0.61919", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.169, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 46.028, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "46.028", + "raw_stderr": "2.926", + "latency": "146.919", + "cost_per_test": "0.69854", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.926, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 40.674, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "40.674", + "raw_stderr": "1.737", + "latency": "109.817", + "cost_per_test": "0.301453", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.737, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:overall:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 31.01, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "31.01", + "raw_stderr": "0.564", + "latency": "176.769", + "cost_per_test": "0.602129", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.564, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 11.012, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "11.012", + "raw_stderr": "2.202", + "latency": "229.949", + "cost_per_test": "0.667183", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.202, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/40611bd4-19ef-4c8f-9921-ae018d3be230.json b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/40611bd4-19ef-4c8f-9921-ae018d3be230.json new file mode 100644 index 0000000000000000000000000000000000000000..def3d0fa702b9aa807fbce232a02b5b7b73685e7 --- /dev/null +++ b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/40611bd4-19ef-4c8f-9921-ae018d3be230.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/anthropic_claude-haiku-4-5-20251001-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-haiku-4-5-20251001-thinking", + "id": "anthropic/claude-haiku-4-5-20251001-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-haiku-4-5-20251001-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 73.232, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.232", + "raw_stderr": "3.542", + "latency": "75.421", + "cost_per_test": "0.050791", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.542, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 71.212, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.212", + "raw_stderr": "3.456", + "latency": "90.985", + "cost_per_test": "0.059337", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.456, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 72.222, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.222", + "raw_stderr": "2.474", + "latency": "83.203", + "cost_per_test": "0.055064", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.474, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/4289476d-969c-4b20-b121-962fa2161584.json b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/4289476d-969c-4b20-b121-962fa2161584.json new file mode 100644 index 0000000000000000000000000000000000000000..7f01f05c8a1d4e565f51e147291983d2254c9e54 --- /dev/null +++ b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/4289476d-969c-4b20-b121-962fa2161584.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/programbench/anthropic_claude-haiku-4-5-20251001-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProgramBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-haiku-4-5-20251001-thinking", + "id": "anthropic/claude-haiku-4-5-20251001-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-haiku-4-5-20251001-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "programbench:almost:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.programbench.almost", + "source_data": { + "dataset_name": "ProgramBench - Almost Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "almost", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Almost Resolved).", + "metric_id": "vals_ai.programbench.almost.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "almost", + "task_name": "Almost Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "457.113", + "cost_per_test": "0.726007", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "programbench:overall:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.programbench.overall", + "source_data": { + "dataset_name": "ProgramBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Overall).", + "metric_id": "vals_ai.programbench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "457.113", + "cost_per_test": "0.726007", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "programbench:partial:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.programbench.partial", + "source_data": { + "dataset_name": "ProgramBench - Raw Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "partial", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Raw Pass Rate).", + "metric_id": "vals_ai.programbench.partial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 25.422, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "partial", + "task_name": "Raw Pass Rate", + "dataset_type": "public", + "industry": "coding", + "raw_score": "25.422", + "raw_stderr": "1.636", + "latency": "457.113", + "cost_per_test": "0.726007", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.636, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "programbench:strict:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.programbench.strict", + "source_data": { + "dataset_name": "ProgramBench - Fully Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "strict", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Fully Resolved).", + "metric_id": "vals_ai.programbench.strict.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "strict", + "task_name": "Fully Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "457.113", + "cost_per_test": "0.726007", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/6dbb2250-68d5-4db3-acb3-7909c0f7cc60.json b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/6dbb2250-68d5-4db3-acb3-7909c0f7cc60.json new file mode 100644 index 0000000000000000000000000000000000000000..738e009e99b692039eac30d7414804a6cb8f7a14 --- /dev/null +++ b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/6dbb2250-68d5-4db3-acb3-7909c0f7cc60.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/anthropic_claude-haiku-4-5-20251001-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-haiku-4-5-20251001-thinking", + "id": "anthropic/claude-haiku-4-5-20251001-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-haiku-4-5-20251001-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 55.274, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "55.274", + "raw_stderr": "1.422", + "latency": "47.524", + "cost_per_test": "0.02743", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.422, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 67.539, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.539", + "raw_stderr": "0.914", + "latency": "47.524", + "cost_per_test": "0.02743", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.914, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 79.804, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "79.804", + "raw_stderr": "1.148", + "latency": "47.524", + "cost_per_test": "0.02743", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.148, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/791a24f2-2d9b-4e3e-9413-624ae2724a63.json b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/791a24f2-2d9b-4e3e-9413-624ae2724a63.json new file mode 100644 index 0000000000000000000000000000000000000000..602352c575c823e688ce6f6a464d89ee4959fa98 --- /dev/null +++ b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/791a24f2-2d9b-4e3e-9413-624ae2724a63.json @@ -0,0 +1,398 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/anthropic_claude-haiku-4-5-20251001-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-haiku-4-5-20251001-thinking", + "id": "anthropic/claude-haiku-4-5-20251001-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-haiku-4-5-20251001-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 63.17, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "63.17", + "raw_stderr": "1.647", + "latency": "15.049", + "cost_per_test": "0.096327", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.647, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 30.999, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "30.999", + "raw_stderr": "0.088", + "latency": "180.056", + "cost_per_test": "0.617074", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.088, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:overall:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 40.903, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "40.903", + "raw_stderr": "1.023", + "latency": "376.932", + "cost_per_test": "0.558143", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.023, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 64.706, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "64.706", + "raw_stderr": "4.732", + "latency": "264.186", + "cost_per_test": "0.407017" + }, + "uncertainty": { + "standard_error": { + "value": 4.732, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 43.82, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "43.82", + "raw_stderr": "1.946", + "latency": "649.489", + "cost_per_test": "0.363822", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.946, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 9.882, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "9.882", + "raw_stderr": "3.549", + "latency": "775.882", + "cost_per_test": "1.306474", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.549, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/8f456206-8250-4903-9b05-48b59e8797c2.json b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/8f456206-8250-4903-9b05-48b59e8797c2.json new file mode 100644 index 0000000000000000000000000000000000000000..6d0640732cfa449413e4b20f609e1c7a8d9bef65 --- /dev/null +++ b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/8f456206-8250-4903-9b05-48b59e8797c2.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/anthropic_claude-haiku-4-5-20251001-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-haiku-4-5-20251001-thinking", + "id": "anthropic/claude-haiku-4-5-20251001-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-haiku-4-5-20251001-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 58.625, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.625", + "raw_stderr": "1.681", + "latency": "9.317", + "cost_per_test": "0.004717", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.681, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.023, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.023", + "raw_stderr": "1.672", + "latency": "18.36", + "cost_per_test": "0.166751", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.672, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.606, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.606", + "raw_stderr": "0.962", + "latency": "14.242", + "cost_per_test": "0.089265", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.962, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 63.17, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.17", + "raw_stderr": "1.647", + "latency": "15.049", + "cost_per_test": "0.096327", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.647, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/a476d27a-f6ff-4ade-a54c-8617778c7517.json b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/a476d27a-f6ff-4ade-a54c-8617778c7517.json new file mode 100644 index 0000000000000000000000000000000000000000..3708cf5f778bb0e79d44ec2d720f0579ce7bfc82 --- /dev/null +++ b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/a476d27a-f6ff-4ade-a54c-8617778c7517.json @@ -0,0 +1,361 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/anthropic_claude-haiku-4-5-20251001-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-haiku-4-5-20251001-thinking", + "id": "anthropic/claude-haiku-4-5-20251001-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-haiku-4-5-20251001-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 23.81, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "23.81", + "raw_stderr": "6.652", + "latency": "313.361", + "cost_per_test": "0.524195", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.652, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 62.069, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "62.069", + "raw_stderr": "3.009", + "latency": "265.78", + "cost_per_test": "0.404706", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.009, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 82.474, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "82.474", + "raw_stderr": "2.737", + "latency": "194.792", + "cost_per_test": "0.279496", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.737, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "254.527", + "cost_per_test": "0.415981", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 66.6, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66.6", + "raw_stderr": "2.111", + "latency": "242.166", + "cost_per_test": "0.366229", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.111, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/b9b5439d-ef76-4799-9cda-c7ae5865cae8.json b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/b9b5439d-ef76-4799-9cda-c7ae5865cae8.json new file mode 100644 index 0000000000000000000000000000000000000000..1ba7f0d77a5311e9e8678c5957b0ca7e0f029e9c --- /dev/null +++ b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/b9b5439d-ef76-4799-9cda-c7ae5865cae8.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/code-migration/anthropic_claude-haiku-4-5-20251001-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Code Migration", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-haiku-4-5-20251001-thinking", + "id": "anthropic/claude-haiku-4-5-20251001-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-haiku-4-5-20251001-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "code-migration:cli:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.code-migration.cli", + "source_data": { + "dataset_name": "Code Migration - CLI", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cli", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (CLI).", + "metric_id": "vals_ai.code-migration.cli.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 10.068, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cli", + "task_name": "CLI", + "dataset_type": "private", + "industry": "coding", + "raw_score": "10.068", + "raw_stderr": "1.408", + "latency": "579", + "cost_per_test": "0.949776", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.408, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "code-migration:cobol:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.code-migration.cobol", + "source_data": { + "dataset_name": "Code Migration - COBOL", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cobol", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (COBOL).", + "metric_id": "vals_ai.code-migration.cobol.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cobol", + "task_name": "COBOL", + "dataset_type": "private", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "637.86", + "cost_per_test": "1.076978", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "code-migration:code_quality:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.code-migration.code_quality", + "source_data": { + "dataset_name": "Code Migration - Code Quality", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "code_quality", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Code Quality).", + "metric_id": "vals_ai.code-migration.code_quality.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 3.348, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "code_quality", + "task_name": "Code Quality", + "dataset_type": "private", + "industry": "coding", + "raw_score": "3.348", + "raw_stderr": "0.036", + "latency": "593.715", + "cost_per_test": "0.981576", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.036, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "code-migration:overall:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.code-migration.overall", + "source_data": { + "dataset_name": "Code Migration - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Overall).", + "metric_id": "vals_ai.code-migration.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 7.551, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "7.551", + "raw_stderr": "1.056", + "latency": "593.715", + "cost_per_test": "0.981576", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.056, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/bd4751a9-c5c7-4460-aa06-f933f4c4ec11.json b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/bd4751a9-c5c7-4460-aa06-f933f4c4ec11.json new file mode 100644 index 0000000000000000000000000000000000000000..4cc412c5ab95e3e7e2859f5867dd076084bca7dc --- /dev/null +++ b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/bd4751a9-c5c7-4460-aa06-f933f4c4ec11.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/anthropic_claude-haiku-4-5-20251001-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-haiku-4-5-20251001-thinking", + "id": "anthropic/claude-haiku-4-5-20251001-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-haiku-4-5-20251001-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 86.259, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.259", + "raw_stderr": "1.077", + "latency": "13.842", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.077, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.6, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.6", + "raw_stderr": "0.21", + "latency": "9.361", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.21, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 86.743, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.743", + "raw_stderr": "0.634", + "latency": "12.412", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.634, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.238, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.238", + "raw_stderr": "0.487", + "latency": "10.372", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.487, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.634, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.634", + "raw_stderr": "1.119", + "latency": "7.66", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.119, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 74.954, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "74.954", + "raw_stderr": "1.753", + "latency": "8.586", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.753, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/c34804d7-1db3-4fb2-aef1-8dbc2985668c.json b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/c34804d7-1db3-4fb2-aef1-8dbc2985668c.json new file mode 100644 index 0000000000000000000000000000000000000000..a57f21c05e138e5609f32bfe54e000c97f768596 --- /dev/null +++ b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/c34804d7-1db3-4fb2-aef1-8dbc2985668c.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/anthropic_claude-haiku-4-5-20251001-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-haiku-4-5-20251001-thinking", + "id": "anthropic/claude-haiku-4-5-20251001-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-haiku-4-5-20251001-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 87.578, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "87.578", + "raw_stderr": "1.838", + "latency": "27.485", + "cost_per_test": "0.022583", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.838, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 11.143, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "11.143", + "raw_stderr": "1.682", + "latency": "130.412", + "cost_per_test": "0.092739", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.682, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 24.804, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "24.804", + "raw_stderr": "2.207", + "latency": "86.787", + "cost_per_test": "0.06398", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.207, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 41.175, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "41.175", + "raw_stderr": "1.109", + "latency": "81.561", + "cost_per_test": "0.059767", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.109, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/c88fceaf-ebae-4e9e-8479-83b463fac752.json b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/c88fceaf-ebae-4e9e-8479-83b463fac752.json new file mode 100644 index 0000000000000000000000000000000000000000..d2b33524fc4e43908161f6e95b9623df920855d3 --- /dev/null +++ b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/c88fceaf-ebae-4e9e-8479-83b463fac752.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/anthropic_claude-haiku-4-5-20251001-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-haiku-4-5-20251001-thinking", + "id": "anthropic/claude-haiku-4-5-20251001-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-haiku-4-5-20251001-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.563, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.563", + "raw_stderr": "1.604", + "latency": "18.955", + "cost_per_test": "0.013073", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.604, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.003, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.003", + "raw_stderr": "1.702", + "latency": "19.528", + "cost_per_test": "0.01488", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.702, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.247, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.247", + "raw_stderr": "1.482", + "latency": "31.291", + "cost_per_test": "0.026179", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.482, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.707, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.707", + "raw_stderr": "2.398", + "latency": "28.835", + "cost_per_test": "0.020126", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.398, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.308, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.308", + "raw_stderr": "1.715", + "latency": "16.328", + "cost_per_test": "0.010968", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.715, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.723, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.723", + "raw_stderr": "1.57", + "latency": "59.966", + "cost_per_test": "0.044915", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.57, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.572, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.572", + "raw_stderr": "1.743", + "latency": "19.554", + "cost_per_test": "0.011481", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.743, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 70.341, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.341", + "raw_stderr": "2.561", + "latency": "18.203", + "cost_per_test": "0.012087", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.561, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 62.216, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "62.216", + "raw_stderr": "1.507", + "latency": "43.716", + "cost_per_test": "0.025813", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.507, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.752, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.752", + "raw_stderr": "1.217", + "latency": "23.459", + "cost_per_test": "0.020505", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.217, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.758, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.758", + "raw_stderr": "1.64", + "latency": "14.09", + "cost_per_test": "0.008857", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.64, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.715, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.715", + "raw_stderr": "0.477", + "latency": "25.784", + "cost_per_test": "0.018312", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.477, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.549, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.549", + "raw_stderr": "2.231", + "latency": "18.668", + "cost_per_test": "0.012033", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.231, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.45, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.45", + "raw_stderr": "1.299", + "latency": "34.158", + "cost_per_test": "0.026522", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.299, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.825, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.825", + "raw_stderr": "1.76", + "latency": "14.232", + "cost_per_test": "0.008931", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.76, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/c8ee83a5-e7b5-467d-a70a-1fd14640d0de.json b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/c8ee83a5-e7b5-467d-a70a-1fd14640d0de.json new file mode 100644 index 0000000000000000000000000000000000000000..5680db7f3b92569bdbb0b45492a8e31a4451d0a4 --- /dev/null +++ b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/c8ee83a5-e7b5-467d-a70a-1fd14640d0de.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/anthropic_claude-haiku-4-5-20251001-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-haiku-4-5-20251001-thinking", + "id": "anthropic/claude-haiku-4-5-20251001-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-haiku-4-5-20251001-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 6.5, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "6.5", + "raw_stderr": "4.383", + "latency": "4391.967", + "cost_per_test": "2.10517", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 4.383, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 5.833, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "5.833", + "raw_stderr": "2.466", + "latency": "10438.736", + "cost_per_test": "2.161642", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.466, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 6.166, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "6.166", + "raw_stderr": "2.515", + "latency": "7415.352", + "cost_per_test": "2.133406", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.515, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/cdac0192-3e4b-47c5-8da6-137c962e3bd1.json b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/cdac0192-3e4b-47c5-8da6-137c962e3bd1.json new file mode 100644 index 0000000000000000000000000000000000000000..12d9f96e996f925e7b70508e0411ee719d1b0bac --- /dev/null +++ b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/cdac0192-3e4b-47c5-8da6-137c962e3bd1.json @@ -0,0 +1,522 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_multimodal_index/anthropic_claude-haiku-4-5-20251001-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Multimodal Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-haiku-4-5-20251001-thinking", + "id": "anthropic/claude-haiku-4-5-20251001-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-haiku-4-5-20251001-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_multimodal_index:corp_fin_v2:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.vals_multimodal_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Multimodal Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (CorpFin v2).", + "metric_id": "vals_ai.vals_multimodal_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 63.17, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "63.17", + "raw_stderr": "1.647", + "latency": "15.049", + "cost_per_test": "0.096327", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.647, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:finance_agent:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.vals_multimodal_index.finance_agent", + "source_data": { + "dataset_name": "Vals Multimodal Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_multimodal_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 30.999, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "30.999", + "raw_stderr": "0.088", + "latency": "180.056", + "cost_per_test": "0.617074", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.088, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:mortgage_tax:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.vals_multimodal_index.mortgage_tax", + "source_data": { + "dataset_name": "Vals Multimodal Index - Mortgage Tax", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "mortgage_tax", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Mortgage Tax).", + "metric_id": "vals_ai.vals_multimodal_index.mortgage_tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 62.162, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "mortgage_tax", + "task_name": "Mortgage Tax", + "dataset_type": "private", + "industry": "index", + "raw_score": "62.162", + "raw_stderr": "0.962", + "latency": "29.792", + "cost_per_test": "0.020742", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.962, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:overall:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.vals_multimodal_index.overall", + "source_data": { + "dataset_name": "Vals Multimodal Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Overall).", + "metric_id": "vals_ai.vals_multimodal_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 42.883, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "42.883", + "raw_stderr": "0.932", + "latency": "282.986", + "cost_per_test": "0.409227", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.932, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:sage:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.vals_multimodal_index.sage", + "source_data": { + "dataset_name": "Vals Multimodal Index - SAGE", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "sage", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SAGE).", + "metric_id": "vals_ai.vals_multimodal_index.sage.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 31.822, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "sage", + "task_name": "SAGE", + "dataset_type": "private", + "industry": "index", + "raw_score": "31.822", + "raw_stderr": "3.134", + "latency": "66.445", + "cost_per_test": "0.05313", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.134, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:swebench:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.vals_multimodal_index.swebench", + "source_data": { + "dataset_name": "Vals Multimodal Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SWE-bench).", + "metric_id": "vals_ai.vals_multimodal_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 64.706, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "64.706", + "raw_stderr": "4.732", + "latency": "264.186", + "cost_per_test": "0.407017" + }, + "uncertainty": { + "standard_error": { + "value": 4.732, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:terminal_bench_2_1:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.vals_multimodal_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Multimodal Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_multimodal_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 43.82, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "43.82", + "raw_stderr": "1.946", + "latency": "649.489", + "cost_per_test": "0.363822", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.946, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:vibe_code_bench:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.vals_multimodal_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Multimodal Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_multimodal_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 9.882, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "9.882", + "raw_stderr": "3.549", + "latency": "775.882", + "cost_per_test": "1.306474", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.549, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/db0bbe3c-c672-4fe5-904e-6c5efac15c23.json b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/db0bbe3c-c672-4fe5-904e-6c5efac15c23.json new file mode 100644 index 0000000000000000000000000000000000000000..3baa23c5749bf47ba2926cae34ea874b570c056b --- /dev/null +++ b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/db0bbe3c-c672-4fe5-904e-6c5efac15c23.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/anthropic_claude-haiku-4-5-20251001-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-haiku-4-5-20251001-thinking", + "id": "anthropic/claude-haiku-4-5-20251001-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-haiku-4-5-20251001-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 83.333, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "83.333", + "raw_stderr": "8.333", + "latency": "14451.125", + "cost_per_test": "8.095034", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 8.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 22.222, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "22.222", + "raw_stderr": "1.111", + "latency": "1926.817", + "cost_per_test": "1.079338", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.111, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 52.727, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "52.727", + "raw_stderr": "3.785", + "latency": "1050.991", + "cost_per_test": "0.58873", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.785, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 43.82, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "43.82", + "raw_stderr": "1.946", + "latency": "649.489", + "cost_per_test": "0.363822", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.946, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/e45e2f26-98ab-4b8b-a9c1-f5f429a83338.json b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/e45e2f26-98ab-4b8b-a9c1-f5f429a83338.json new file mode 100644 index 0000000000000000000000000000000000000000..5ec69bd488906219a21b00e7d09d4652ea81be53 --- /dev/null +++ b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/e45e2f26-98ab-4b8b-a9c1-f5f429a83338.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/anthropic_claude-haiku-4-5-20251001-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-haiku-4-5-20251001-thinking", + "id": "anthropic/claude-haiku-4-5-20251001-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-haiku-4-5-20251001-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 11.393, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "11.393", + "raw_stderr": "3.129", + "latency": "775.882", + "cost_per_test": "1.306474", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.129, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/ec6db844-b251-4d78-959c-e1b92126629b.json b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/ec6db844-b251-4d78-959c-e1b92126629b.json new file mode 100644 index 0000000000000000000000000000000000000000..f648aa6defe481fe8ff0569e4d9c72bc67aed925 --- /dev/null +++ b/data/vals_ai/anthropic/claude-haiku-4-5-20251001-thinking/ec6db844-b251-4d78-959c-e1b92126629b.json @@ -0,0 +1,101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/anthropic_claude-haiku-4-5-20251001-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-haiku-4-5-20251001-thinking", + "id": "anthropic/claude-haiku-4-5-20251001-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-haiku-4-5-20251001-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:anthropic/claude-haiku-4-5-20251001-thinking:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 46.069, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "46.069", + "raw_stderr": "1.198", + "latency": "40.481", + "cost_per_test": "0.027476", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.198, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-haiku-4-5-20251001/555b4edc-e177-468e-a8fb-40ec4bfc5ea0.json b/data/vals_ai/anthropic/claude-haiku-4-5-20251001/555b4edc-e177-468e-a8fb-40ec4bfc5ea0.json new file mode 100644 index 0000000000000000000000000000000000000000..9a50f0132e3306659983db5b3f30a3dfa03ca207 --- /dev/null +++ b/data/vals_ai/anthropic/claude-haiku-4-5-20251001/555b4edc-e177-468e-a8fb-40ec4bfc5ea0.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/public-benefits-bench/anthropic_claude-haiku-4-5-20251001/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Public Benefits Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "social mobility", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-haiku-4-5-20251001", + "id": "anthropic/claude-haiku-4-5-20251001", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-haiku-4-5-20251001", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "public-benefits-bench:overall:anthropic/claude-haiku-4-5-20251001:score", + "evaluation_name": "vals_ai.public-benefits-bench.overall", + "source_data": { + "dataset_name": "Public Benefits Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Public Benefits Bench v1.1 (Overall).", + "metric_id": "vals_ai.public-benefits-bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "score_details": { + "score": 54.263, + "details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "social mobility", + "raw_score": "54.263", + "raw_stderr": "1.296", + "latency": "337.201", + "cost_per_test": "0.224512", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.296, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-haiku-4-5-20251001/b924b82e-494e-4b52-a8ef-6caae7dbbc25.json b/data/vals_ai/anthropic/claude-haiku-4-5-20251001/b924b82e-494e-4b52-a8ef-6caae7dbbc25.json new file mode 100644 index 0000000000000000000000000000000000000000..9b1f766bea8fa16588bcb5cb014f060e26070a59 --- /dev/null +++ b/data/vals_ai/anthropic/claude-haiku-4-5-20251001/b924b82e-494e-4b52-a8ef-6caae7dbbc25.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/anthropic_claude-haiku-4-5-20251001/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-haiku-4-5-20251001", + "id": "anthropic/claude-haiku-4-5-20251001", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-haiku-4-5-20251001", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:anthropic/claude-haiku-4-5-20251001:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.256, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.256", + "raw_stderr": "1.671", + "latency": "5.204", + "cost_per_test": "0.002564", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.671, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:anthropic/claude-haiku-4-5-20251001:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 58.858, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.858", + "raw_stderr": "1.68", + "latency": "10.123", + "cost_per_test": "0.150828", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.68, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:anthropic/claude-haiku-4-5-20251001:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.295, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.295", + "raw_stderr": "0.964", + "latency": "7.951", + "cost_per_test": "0.075779", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.964, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:anthropic/claude-haiku-4-5-20251001:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.772, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.772", + "raw_stderr": "1.659", + "latency": "8.526", + "cost_per_test": "0.073944", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.659, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-1-20250805-thinking/11aaf721-6c9c-4b51-8ed7-2b4683dcb481.json b/data/vals_ai/anthropic/claude-opus-4-1-20250805-thinking/11aaf721-6c9c-4b51-8ed7-2b4683dcb481.json new file mode 100644 index 0000000000000000000000000000000000000000..6d8935193b2a83df50b2f4173cc55ee28b1894d7 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-1-20250805-thinking/11aaf721-6c9c-4b51-8ed7-2b4683dcb481.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/anthropic_claude-opus-4-1-20250805-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-1-20250805-thinking", + "id": "anthropic/claude-opus-4-1-20250805-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-1-20250805-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.469, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.469", + "raw_stderr": "0.986", + "latency": "23.454", + "cost_per_test": "0.104728", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.986, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.128, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.128", + "raw_stderr": "1.012", + "latency": "22.186", + "cost_per_test": "0.099964", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.012, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.869, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.869", + "raw_stderr": "0.868", + "latency": "38.648", + "cost_per_test": "0.18329", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.868, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.024, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.024", + "raw_stderr": "1.643", + "latency": "27.701", + "cost_per_test": "0.129006", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.643, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.071, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.071", + "raw_stderr": "1.028", + "latency": "19.787", + "cost_per_test": "0.08772", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.028, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.043, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.043", + "raw_stderr": "1.268", + "latency": "60.787", + "cost_per_test": "0.265746", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.268, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.963, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.963", + "raw_stderr": "1.254", + "latency": "20.011", + "cost_per_test": "0.085317", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.254, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.89, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.89", + "raw_stderr": "1.973", + "latency": "21.033", + "cost_per_test": "0.102264", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.973, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.927, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.927", + "raw_stderr": "1.209", + "latency": "26.755", + "cost_per_test": "0.111326", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.209, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.226, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.226", + "raw_stderr": "0.653", + "latency": "36.869", + "cost_per_test": "0.180541", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.653, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.771, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.771", + "raw_stderr": "1.078", + "latency": "16.333", + "cost_per_test": "0.068106", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.078, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.924, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.924", + "raw_stderr": "0.325", + "latency": "27.668", + "cost_per_test": "0.124083", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.325, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.176, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.176", + "raw_stderr": "1.445", + "latency": "20.153", + "cost_per_test": "0.084337", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.445, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.915, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.915", + "raw_stderr": "0.841", + "latency": "34.366", + "cost_per_test": "0.154028", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.841, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.471, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.471", + "raw_stderr": "1.131", + "latency": "19.271", + "cost_per_test": "0.080786", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.131, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-1-20250805-thinking/2c4b993a-b13f-42fa-93c1-6390f782e279.json b/data/vals_ai/anthropic/claude-opus-4-1-20250805-thinking/2c4b993a-b13f-42fa-93c1-6390f782e279.json new file mode 100644 index 0000000000000000000000000000000000000000..337a42181c761f82f63b1b7c64cb98a120208152 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-1-20250805-thinking/2c4b993a-b13f-42fa-93c1-6390f782e279.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/anthropic_claude-opus-4-1-20250805-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-1-20250805-thinking", + "id": "anthropic/claude-opus-4-1-20250805-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-1-20250805-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 62.96, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.96", + "raw_stderr": "1.381", + "latency": "26.637", + "cost_per_test": "0.093482", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.381, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 73.672, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "73.672", + "raw_stderr": "0.864", + "latency": "26.637", + "cost_per_test": "0.093482", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.864, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 84.383, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "84.383", + "raw_stderr": "1.038", + "latency": "26.637", + "cost_per_test": "0.093482", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.038, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-1-20250805-thinking/8ac91ffe-a69a-4ae4-b580-b4ca6e82451c.json b/data/vals_ai/anthropic/claude-opus-4-1-20250805-thinking/8ac91ffe-a69a-4ae4-b580-b4ca6e82451c.json new file mode 100644 index 0000000000000000000000000000000000000000..40b1a8af6c6783edf14238af9346f9d2e2a4e2dd --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-1-20250805-thinking/8ac91ffe-a69a-4ae4-b580-b4ca6e82451c.json @@ -0,0 +1,346 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/anthropic_claude-opus-4-1-20250805-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-1-20250805-thinking", + "id": "anthropic/claude-opus-4-1-20250805-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-1-20250805-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 31.746, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "31.746", + "raw_stderr": "5.841", + "latency": "52.287", + "cost_per_test": "0.284017", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.841, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 21.825, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "21.825", + "raw_stderr": "5.266", + "latency": "26.102", + "cost_per_test": "0.250705", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.266, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 46.296, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "46.296", + "raw_stderr": "6.516", + "latency": "85.079", + "cost_per_test": "1.424207", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.516, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 37.594, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "37.594", + "raw_stderr": "5.441", + "latency": "44.045", + "cost_per_test": "0.416739", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.441, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 30.388, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "30.388", + "raw_stderr": "3.188", + "latency": "40.811", + "cost_per_test": "0.317154", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.188, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-1-20250805-thinking/8fc9a34c-168e-4275-b540-f50c89395b50.json b/data/vals_ai/anthropic/claude-opus-4-1-20250805-thinking/8fc9a34c-168e-4275-b540-f50c89395b50.json new file mode 100644 index 0000000000000000000000000000000000000000..ed1847c7249ea75a88b4f15c7231a307221da2e5 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-1-20250805-thinking/8fc9a34c-168e-4275-b540-f50c89395b50.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/anthropic_claude-opus-4-1-20250805-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-1-20250805-thinking", + "id": "anthropic/claude-opus-4-1-20250805-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-1-20250805-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 76.768, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.768", + "raw_stderr": "3.001", + "latency": "36.604", + "cost_per_test": "0.16189", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.001, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 75.758, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.758", + "raw_stderr": "3.108", + "latency": "34.939", + "cost_per_test": "0.127465", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.108, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 76.263, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.263", + "raw_stderr": "2.16", + "latency": "35.772", + "cost_per_test": "0.144678", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.16, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-1-20250805-thinking/a6f66b0e-0667-4e05-8d10-b13994eee3fa.json b/data/vals_ai/anthropic/claude-opus-4-1-20250805-thinking/a6f66b0e-0667-4e05-8d10-b13994eee3fa.json new file mode 100644 index 0000000000000000000000000000000000000000..738f07c8877f59f6d48769eee615377baeb55bdf --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-1-20250805-thinking/a6f66b0e-0667-4e05-8d10-b13994eee3fa.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/anthropic_claude-opus-4-1-20250805-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-1-20250805-thinking", + "id": "anthropic/claude-opus-4-1-20250805-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-1-20250805-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.758, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.758", + "raw_stderr": "0.617", + "latency": "38.083", + "cost_per_test": "0.154263", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.617, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 28.286, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "28.286", + "raw_stderr": "2.407", + "latency": "143.594", + "cost_per_test": "0.550873", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.407, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 72.324, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "72.324", + "raw_stderr": "2.286", + "latency": "94.824", + "cost_per_test": "0.376829", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.286, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 66.456, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66.456", + "raw_stderr": "1.125", + "latency": "92.167", + "cost_per_test": "0.360655", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.125, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-1-20250805-thinking/a9d5cc00-555d-4148-931a-c70b1dcecd91.json b/data/vals_ai/anthropic/claude-opus-4-1-20250805-thinking/a9d5cc00-555d-4148-931a-c70b1dcecd91.json new file mode 100644 index 0000000000000000000000000000000000000000..8350dcefee920e2ee10f7df3abb790af5fe1203a --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-1-20250805-thinking/a9d5cc00-555d-4148-931a-c70b1dcecd91.json @@ -0,0 +1,101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/anthropic_claude-opus-4-1-20250805-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-1-20250805-thinking", + "id": "anthropic/claude-opus-4-1-20250805-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-1-20250805-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 77.514, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.514", + "raw_stderr": "1.004", + "latency": "2.446", + "cost_per_test": "0.168667", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.004, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-1-20250805-thinking/c4a9cad7-52d0-4f16-befc-dfdd7034482e.json b/data/vals_ai/anthropic/claude-opus-4-1-20250805-thinking/c4a9cad7-52d0-4f16-befc-dfdd7034482e.json new file mode 100644 index 0000000000000000000000000000000000000000..0380ff0d62460416ad12e4f276715378c5df804a --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-1-20250805-thinking/c4a9cad7-52d0-4f16-befc-dfdd7034482e.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/anthropic_claude-opus-4-1-20250805-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-1-20250805-thinking", + "id": "anthropic/claude-opus-4-1-20250805-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-1-20250805-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 73.901, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "73.901", + "raw_stderr": "1.965", + "latency": "57.404", + "cost_per_test": "0.263427", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.965, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-1-20250805-thinking/def32f7f-c65b-4292-9d2f-1fee919f2434.json b/data/vals_ai/anthropic/claude-opus-4-1-20250805-thinking/def32f7f-c65b-4292-9d2f-1fee919f2434.json new file mode 100644 index 0000000000000000000000000000000000000000..dc4be5b2e3d66b0bbc5800b49c299f9ecd334399 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-1-20250805-thinking/def32f7f-c65b-4292-9d2f-1fee919f2434.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/anthropic_claude-opus-4-1-20250805-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-1-20250805-thinking", + "id": "anthropic/claude-opus-4-1-20250805-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-1-20250805-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 53.418, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "53.418", + "raw_stderr": "1.406", + "latency": "31.019", + "cost_per_test": "0.162339", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.406, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 56.121, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.121", + "raw_stderr": "0.988", + "latency": "23.52", + "cost_per_test": "0.135806", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.988, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:anthropic/claude-opus-4-1-20250805-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 58.824, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.824", + "raw_stderr": "1.388", + "latency": "16.022", + "cost_per_test": "0.109272", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.388, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-1-20250805/1acfc7de-0966-4e1c-86ef-f359cedd363a.json b/data/vals_ai/anthropic/claude-opus-4-1-20250805/1acfc7de-0966-4e1c-86ef-f359cedd363a.json new file mode 100644 index 0000000000000000000000000000000000000000..e6e8ba5086382033c74958e2d3dce9a61f8976f3 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-1-20250805/1acfc7de-0966-4e1c-86ef-f359cedd363a.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/anthropic_claude-opus-4-1-20250805/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-1-20250805", + "id": "anthropic/claude-opus-4-1-20250805", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-1-20250805", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 18.667, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "18.667", + "raw_stderr": "9.657", + "latency": "5926.555", + "cost_per_test": "15.554335", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 9.657, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 6.364, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "6.364", + "raw_stderr": "1.819", + "latency": "6323.238", + "cost_per_test": "19.617569", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.819, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 12.516, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "12.516", + "raw_stderr": "4.913", + "latency": "6124.896", + "cost_per_test": "17.585952", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 4.913, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-1-20250805/28936934-52d0-4cfd-a7ed-c70602e56c64.json b/data/vals_ai/anthropic/claude-opus-4-1-20250805/28936934-52d0-4cfd-a7ed-c70602e56c64.json new file mode 100644 index 0000000000000000000000000000000000000000..ca926621b859459d643ca593fcb7e2f126d0507b --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-1-20250805/28936934-52d0-4cfd-a7ed-c70602e56c64.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/anthropic_claude-opus-4-1-20250805/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-1-20250805", + "id": "anthropic/claude-opus-4-1-20250805", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-1-20250805", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.887, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.887", + "raw_stderr": "0.96", + "latency": "19.899", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.96, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.734, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.734", + "raw_stderr": "1.086", + "latency": "19.061", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.086, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.488, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.488", + "raw_stderr": "0.912", + "latency": "18.038", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.912, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.512, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.512", + "raw_stderr": "1.573", + "latency": "14.003", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.573, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.455, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.455", + "raw_stderr": "1.057", + "latency": "18.5", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.057, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.709, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.709", + "raw_stderr": "1.354", + "latency": "13.94", + "cost_per_test": "0.073918", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.354, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.108, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.108", + "raw_stderr": "1.282", + "latency": "15.984", + "cost_per_test": "0.015834", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.282, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.415, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.415", + "raw_stderr": "1.95", + "latency": "20.301", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.95, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.475, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.475", + "raw_stderr": "1.259", + "latency": "18.284", + "cost_per_test": "0.009339", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.259, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.597, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.597", + "raw_stderr": "0.619", + "latency": "19.462", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.619, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.771, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.771", + "raw_stderr": "1.078", + "latency": "14.998", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.078, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.214, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.214", + "raw_stderr": "0.328", + "latency": "17.757", + "cost_per_test": "0.007078", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.328, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.575, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.575", + "raw_stderr": "1.477", + "latency": "21.155", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.477, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.915, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.915", + "raw_stderr": "0.871", + "latency": "18.345", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.871, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.348, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.348", + "raw_stderr": "1.092", + "latency": "16.632", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.092, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-1-20250805/315e92c5-24f3-4f4b-9f07-daf4c63282ab.json b/data/vals_ai/anthropic/claude-opus-4-1-20250805/315e92c5-24f3-4f4b-9f07-daf4c63282ab.json new file mode 100644 index 0000000000000000000000000000000000000000..56dad06252ab64fe3cbc428d7cf7f279073a12a0 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-1-20250805/315e92c5-24f3-4f4b-9f07-daf4c63282ab.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/anthropic_claude-opus-4-1-20250805/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-1-20250805", + "id": "anthropic/claude-opus-4-1-20250805", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-1-20250805", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 65.183, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.183", + "raw_stderr": "1.343", + "latency": "12.004", + "cost_per_test": "0.093902", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.343, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 61.089, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.089", + "raw_stderr": "0.969", + "latency": "10.065", + "cost_per_test": "0.088232", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.969, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 56.995, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.995", + "raw_stderr": "1.396", + "latency": "8.126", + "cost_per_test": "0.082562", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.396, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-1-20250805/326ce173-9a7c-4152-93dd-ab64931a9aa4.json b/data/vals_ai/anthropic/claude-opus-4-1-20250805/326ce173-9a7c-4152-93dd-ab64931a9aa4.json new file mode 100644 index 0000000000000000000000000000000000000000..c01ad5d06628bc542d77f003e2ce4bb83159b801 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-1-20250805/326ce173-9a7c-4152-93dd-ab64931a9aa4.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/anthropic_claude-opus-4-1-20250805/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-1-20250805", + "id": "anthropic/claude-opus-4-1-20250805", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-1-20250805", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 88.951, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.951", + "raw_stderr": "0.817", + "latency": "2.637", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.817, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.408, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.408", + "raw_stderr": "0.202", + "latency": "2.442", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.202, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 88.534, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.534", + "raw_stderr": "0.579", + "latency": "2.692", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.579, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.458, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.458", + "raw_stderr": "0.449", + "latency": "2.79", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.449, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.038, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.038", + "raw_stderr": "1.168", + "latency": "2.416", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.168, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.36, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.36", + "raw_stderr": "1.619", + "latency": "3.763", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.619, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-1-20250805/3730ae04-567b-4e1d-a1a0-85dab6679c7e.json b/data/vals_ai/anthropic/claude-opus-4-1-20250805/3730ae04-567b-4e1d-a1a0-85dab6679c7e.json new file mode 100644 index 0000000000000000000000000000000000000000..5f10c32f8829c34723a73aef68f3984322036904 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-1-20250805/3730ae04-567b-4e1d-a1a0-85dab6679c7e.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/anthropic_claude-opus-4-1-20250805/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-1-20250805", + "id": "anthropic/claude-opus-4-1-20250805", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-1-20250805", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 97.516, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "97.516", + "raw_stderr": "0.613", + "latency": "22.583", + "cost_per_test": "0.037777", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.613, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 25.143, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "25.143", + "raw_stderr": "2.319", + "latency": "42.675", + "cost_per_test": "0.074597", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.319, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 71.018, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "71.018", + "raw_stderr": "2.318", + "latency": "32.269", + "cost_per_test": "0.055591", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.318, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 64.559, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "64.559", + "raw_stderr": "1.112", + "latency": "32.509", + "cost_per_test": "0.055988", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.112, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-1-20250805/3ec0525b-23d3-4acc-a681-dbf15f7bee30.json b/data/vals_ai/anthropic/claude-opus-4-1-20250805/3ec0525b-23d3-4acc-a681-dbf15f7bee30.json new file mode 100644 index 0000000000000000000000000000000000000000..f109e8dfa56499f364b899dbaff077b8d85477ec --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-1-20250805/3ec0525b-23d3-4acc-a681-dbf15f7bee30.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/anthropic_claude-opus-4-1-20250805/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-1-20250805", + "id": "anthropic/claude-opus-4-1-20250805", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-1-20250805", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 71.753, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "71.753", + "raw_stderr": "2.021", + "latency": "38.04", + "cost_per_test": "0.187162", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.021, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-1-20250805/61302201-25f2-4e8d-b136-ecef100eff5d.json b/data/vals_ai/anthropic/claude-opus-4-1-20250805/61302201-25f2-4e8d-b136-ecef100eff5d.json new file mode 100644 index 0000000000000000000000000000000000000000..e0d052e9c2244ee146fa8c70430a1a33f8891e56 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-1-20250805/61302201-25f2-4e8d-b136-ecef100eff5d.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/anthropic_claude-opus-4-1-20250805/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-1-20250805", + "id": "anthropic/claude-opus-4-1-20250805", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-1-20250805", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 71.212, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.212", + "raw_stderr": "3.218", + "latency": "38.525", + "cost_per_test": "0.083439", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.218, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 68.687, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "68.687", + "raw_stderr": "3.296", + "latency": "33.342", + "cost_per_test": "0.051403", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.296, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 69.95, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.95", + "raw_stderr": "2.303", + "latency": "35.934", + "cost_per_test": "0.067421", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.303, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-1-20250805/69e6bf69-ca95-474b-b504-144da32a572c.json b/data/vals_ai/anthropic/claude-opus-4-1-20250805/69e6bf69-ca95-474b-b504-144da32a572c.json new file mode 100644 index 0000000000000000000000000000000000000000..ecb41814f000bf177847ae2ced9da52bdfe12d10 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-1-20250805/69e6bf69-ca95-474b-b504-144da32a572c.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/anthropic_claude-opus-4-1-20250805/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-1-20250805", + "id": "anthropic/claude-opus-4-1-20250805", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-1-20250805", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 59.853, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.853", + "raw_stderr": "1.402", + "latency": "32.602", + "cost_per_test": "0.039842", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.402, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 71.464, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.464", + "raw_stderr": "0.882", + "latency": "32.602", + "cost_per_test": "0.039842", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.882, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 83.074, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "83.074", + "raw_stderr": "1.072", + "latency": "32.602", + "cost_per_test": "0.039842", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.072, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-1-20250805/a19de6c9-3034-489d-8067-49929eaafadf.json b/data/vals_ai/anthropic/claude-opus-4-1-20250805/a19de6c9-3034-489d-8067-49929eaafadf.json new file mode 100644 index 0000000000000000000000000000000000000000..b801f25aaa13a8e9819f585857ea000f866fc22c --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-1-20250805/a19de6c9-3034-489d-8067-49929eaafadf.json @@ -0,0 +1,101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/anthropic_claude-opus-4-1-20250805/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-1-20250805", + "id": "anthropic/claude-opus-4-1-20250805", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-1-20250805", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 73.715, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.715", + "raw_stderr": "1.058", + "latency": "46.277", + "cost_per_test": "0.044475", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.058, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-1-20250805/b3379029-001f-48f0-b22b-192193510f5c.json b/data/vals_ai/anthropic/claude-opus-4-1-20250805/b3379029-001f-48f0-b22b-192193510f5c.json new file mode 100644 index 0000000000000000000000000000000000000000..aff96c098b3cdbf0215144bef9383a1276d13cb0 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-1-20250805/b3379029-001f-48f0-b22b-192193510f5c.json @@ -0,0 +1,346 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/anthropic_claude-opus-4-1-20250805/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-1-20250805", + "id": "anthropic/claude-opus-4-1-20250805", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-1-20250805", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 25.397, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "25.397", + "raw_stderr": "5.462", + "latency": "19.359", + "cost_per_test": "0.147236", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.462, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 15.476, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "15.476", + "raw_stderr": "4.59", + "latency": "17.193", + "cost_per_test": "0.206172", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 4.59, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 26.852, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "26.852", + "raw_stderr": "5.818", + "latency": "25.246", + "cost_per_test": "0.984331", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.818, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 53.268, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "53.268", + "raw_stderr": "5.604", + "latency": "15.449", + "cost_per_test": "0.304378", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.604, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:anthropic/claude-opus-4-1-20250805:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 31.38, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "31.38", + "raw_stderr": "3.024", + "latency": "17.334", + "cost_per_test": "0.219262", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.024, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-20250514-thinking/4a79fae0-b38f-452c-aaee-d1a3fd9e124b.json b/data/vals_ai/anthropic/claude-opus-4-20250514-thinking/4a79fae0-b38f-452c-aaee-d1a3fd9e124b.json new file mode 100644 index 0000000000000000000000000000000000000000..0dff4cc975fbb1eea87ae2deb9fb1ddf961971d2 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-20250514-thinking/4a79fae0-b38f-452c-aaee-d1a3fd9e124b.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/anthropic_claude-opus-4-20250514-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-20250514-thinking", + "id": "anthropic/claude-opus-4-20250514-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-20250514-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:anthropic/claude-opus-4-20250514-thinking:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 99.068, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "99.068", + "raw_stderr": "0.535", + "latency": "39.894", + "cost_per_test": "0.173558", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:anthropic/claude-opus-4-20250514-thinking:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 33.429, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.429", + "raw_stderr": "2.522", + "latency": "145.059", + "cost_per_test": "0.609114", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.522, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:anthropic/claude-opus-4-20250514-thinking:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 78.068, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "78.068", + "raw_stderr": "2.114", + "latency": "95.662", + "cost_per_test": "0.409474", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.114, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:anthropic/claude-opus-4-20250514-thinking:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 70.188, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "70.188", + "raw_stderr": "1.111", + "latency": "93.538", + "cost_per_test": "0.397382", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.111, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-20250514/1593bb05-e8ac-445d-b7dd-828647a6035a.json b/data/vals_ai/anthropic/claude-opus-4-20250514/1593bb05-e8ac-445d-b7dd-828647a6035a.json new file mode 100644 index 0000000000000000000000000000000000000000..5387e541ce66b079bcaf4d21dbc7576e8e990634 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-20250514/1593bb05-e8ac-445d-b7dd-828647a6035a.json @@ -0,0 +1,101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/anthropic_claude-opus-4-20250514/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-20250514", + "id": "anthropic/claude-opus-4-20250514", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-20250514", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 73.31, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.31", + "raw_stderr": "1.063", + "latency": "14.023", + "cost_per_test": "0.061801", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.063, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-20250514/23e6d46d-6857-417a-9628-594081fac6fe.json b/data/vals_ai/anthropic/claude-opus-4-20250514/23e6d46d-6857-417a-9628-594081fac6fe.json new file mode 100644 index 0000000000000000000000000000000000000000..c088c2320014db18c59a8ceea730b4c769c4fde6 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-20250514/23e6d46d-6857-417a-9628-594081fac6fe.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/anthropic_claude-opus-4-20250514/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-20250514", + "id": "anthropic/claude-opus-4-20250514", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-20250514", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 73.737, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.737", + "raw_stderr": "3.127", + "latency": "20.274", + "cost_per_test": "0.083302", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.127, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 69.697, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.697", + "raw_stderr": "3.266", + "latency": "17.986", + "cost_per_test": "0.050314", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.266, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 71.717, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.717", + "raw_stderr": "2.261", + "latency": "19.13", + "cost_per_test": "0.066808", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.261, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-20250514/2667a9b1-0b47-445c-a911-ad5763358627.json b/data/vals_ai/anthropic/claude-opus-4-20250514/2667a9b1-0b47-445c-a911-ad5763358627.json new file mode 100644 index 0000000000000000000000000000000000000000..2a765a9e30ebf342782a7fc0e877c69a6d0566f9 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-20250514/2667a9b1-0b47-445c-a911-ad5763358627.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/anthropic_claude-opus-4-20250514/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-20250514", + "id": "anthropic/claude-opus-4-20250514", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-20250514", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 60.507, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.507", + "raw_stderr": "1.398", + "latency": "15.23", + "cost_per_test": "0.03694", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.398, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 71.914, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.914", + "raw_stderr": "0.879", + "latency": "15.23", + "cost_per_test": "0.03694", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.879, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 83.32, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "83.32", + "raw_stderr": "1.066", + "latency": "15.23", + "cost_per_test": "0.03694", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.066, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-20250514/3b4059cb-39f4-4ceb-aaaa-916504bc929c.json b/data/vals_ai/anthropic/claude-opus-4-20250514/3b4059cb-39f4-4ceb-aaaa-916504bc929c.json new file mode 100644 index 0000000000000000000000000000000000000000..fee90184fcbafb92d722d395ddc3f9eba00a84ca --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-20250514/3b4059cb-39f4-4ceb-aaaa-916504bc929c.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/anthropic_claude-opus-4-20250514/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-20250514", + "id": "anthropic/claude-opus-4-20250514", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-20250514", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.026, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.026", + "raw_stderr": "0.951", + "latency": "10.237", + "cost_per_test": "0.048629", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.951, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.354, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.354", + "raw_stderr": "1.098", + "latency": "8.321", + "cost_per_test": "0.039973", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.098, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.898, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.898", + "raw_stderr": "0.969", + "latency": "11.11", + "cost_per_test": "0.062945", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.969, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.293, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.293", + "raw_stderr": "1.588", + "latency": "10.875", + "cost_per_test": "0.054997", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.588, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.692, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.692", + "raw_stderr": "1.047", + "latency": "9.348", + "cost_per_test": "0.043404", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.047, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.367, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.367", + "raw_stderr": "1.365", + "latency": "14.014", + "cost_per_test": "0.067022", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.365, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.985, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.985", + "raw_stderr": "1.282", + "latency": "9.711", + "cost_per_test": "0.041684", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.282, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.315, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.315", + "raw_stderr": "2.047", + "latency": "11.317", + "cost_per_test": "0.059706", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.047, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.569, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.569", + "raw_stderr": "1.312", + "latency": "11.27", + "cost_per_test": "0.050234", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.312, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.708, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.708", + "raw_stderr": "0.661", + "latency": "10.815", + "cost_per_test": "0.064139", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.661, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.338, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.338", + "raw_stderr": "1.094", + "latency": "8.322", + "cost_per_test": "0.033476", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.094, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.166, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.166", + "raw_stderr": "0.337", + "latency": "10.405", + "cost_per_test": "0.049652", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.337, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.571, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.571", + "raw_stderr": "1.573", + "latency": "9.721", + "cost_per_test": "0.039418", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.573, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.991, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.991", + "raw_stderr": "0.902", + "latency": "10.681", + "cost_per_test": "0.050261", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.902, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.221, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.221", + "raw_stderr": "1.141", + "latency": "9.928", + "cost_per_test": "0.039244", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.141, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-20250514/b5a93e2d-7a55-4eef-ac00-80b423586a55.json b/data/vals_ai/anthropic/claude-opus-4-20250514/b5a93e2d-7a55-4eef-ac00-80b423586a55.json new file mode 100644 index 0000000000000000000000000000000000000000..cad8507377a4e3a4a1d3e919258e969ad31e80a3 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-20250514/b5a93e2d-7a55-4eef-ac00-80b423586a55.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/anthropic_claude-opus-4-20250514/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-20250514", + "id": "anthropic/claude-opus-4-20250514", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-20250514", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 96.273, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "96.273", + "raw_stderr": "1.056", + "latency": "9.857", + "cost_per_test": "0.035912", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.056, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 20.857, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "20.857", + "raw_stderr": "2.172", + "latency": "20.172", + "cost_per_test": "0.077502", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.172, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 70.757, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "70.757", + "raw_stderr": "2.324", + "latency": "14.302", + "cost_per_test": "0.053676", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.324, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 62.629, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "62.629", + "raw_stderr": "1.117", + "latency": "14.777", + "cost_per_test": "0.055697", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.117, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-20250514/d688d9cd-bb6a-4d57-89e0-11a01b732ba3.json b/data/vals_ai/anthropic/claude-opus-4-20250514/d688d9cd-bb6a-4d57-89e0-11a01b732ba3.json new file mode 100644 index 0000000000000000000000000000000000000000..3cd444da4b8c6134cdd7e536fe58397bcfe52289 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-20250514/d688d9cd-bb6a-4d57-89e0-11a01b732ba3.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/anthropic_claude-opus-4-20250514/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-20250514", + "id": "anthropic/claude-opus-4-20250514", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-20250514", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 58.983, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.983", + "raw_stderr": "1.387", + "latency": "16.297", + "cost_per_test": "0.093539", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.387, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 58.586, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.586", + "raw_stderr": "0.982", + "latency": "13.758", + "cost_per_test": "0.08824", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.982, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 58.188, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.188", + "raw_stderr": "1.391", + "latency": "11.218", + "cost_per_test": "0.082942", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.391, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-20250514/e22639eb-3c7d-4102-a89b-93c60fc9272f.json b/data/vals_ai/anthropic/claude-opus-4-20250514/e22639eb-3c7d-4102-a89b-93c60fc9272f.json new file mode 100644 index 0000000000000000000000000000000000000000..b018d201197a867389aa55fafc2e622c6b02e0dd --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-20250514/e22639eb-3c7d-4102-a89b-93c60fc9272f.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/anthropic_claude-opus-4-20250514/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-20250514", + "id": "anthropic/claude-opus-4-20250514", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-20250514", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.942, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.942", + "raw_stderr": "0.814", + "latency": "3.297", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.814, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.346, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.346", + "raw_stderr": "0.208", + "latency": "3.349", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.208, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.936, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.936", + "raw_stderr": "0.586", + "latency": "3.775", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.586, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.071, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.071", + "raw_stderr": "0.439", + "latency": "3.733", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.439, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.211, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.211", + "raw_stderr": "1.162", + "latency": "3.602", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.162, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:anthropic/claude-opus-4-20250514:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.92, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.92", + "raw_stderr": "1.552", + "latency": "4.643", + "temperature": "1", + "max_output_tokens": "32000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.552, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/029c0c88-e409-41d4-af79-2cc948e776db.json b/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/029c0c88-e409-41d4-af79-2cc948e776db.json new file mode 100644 index 0000000000000000000000000000000000000000..dfcde1642aa70cf56ef3637f0b88078893b24628 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/029c0c88-e409-41d4-af79-2cc948e776db.json @@ -0,0 +1,450 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/anthropic_claude-opus-4-5-20251101-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-5-20251101-thinking", + "id": "anthropic/claude-opus-4-5-20251101-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-5-20251101-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 89.436, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.436", + "raw_stderr": "0.722", + "latency": "7.635", + "cost_per_test": "0.033545", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.722, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.954, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.954", + "raw_stderr": "0.186", + "latency": "7.524", + "cost_per_test": "0.03889", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.186, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 88.866, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.866", + "raw_stderr": "0.592", + "latency": "8.434", + "cost_per_test": "0.04999", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.592, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 84.604, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.604", + "raw_stderr": "0.391", + "latency": "7.638", + "cost_per_test": "0.037097", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.391, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.184, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.184", + "raw_stderr": "0.914", + "latency": "6.208", + "cost_per_test": "0.032262", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.914, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.58, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.58", + "raw_stderr": "1.441", + "latency": "8.387", + "cost_per_test": "0.030796", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.441, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/0e6a3be8-cd74-4427-a1f9-3599391db279.json b/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/0e6a3be8-cd74-4427-a1f9-3599391db279.json new file mode 100644 index 0000000000000000000000000000000000000000..5fd3248268b95828e78b222cbfcc36502169f23d --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/0e6a3be8-cd74-4427-a1f9-3599391db279.json @@ -0,0 +1,1071 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/anthropic_claude-opus-4-5-20251101-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-5-20251101-thinking", + "id": "anthropic/claude-opus-4-5-20251101-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-5-20251101-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.724, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.724", + "raw_stderr": "1.122", + "latency": "24.916", + "cost_per_test": "0.139523", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.122, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.298, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.298", + "raw_stderr": "1.364", + "latency": "30.544", + "cost_per_test": "0.179429", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.364, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.813, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.813", + "raw_stderr": "1.205", + "latency": "47.605", + "cost_per_test": "0.304988", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.205, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.61, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.61", + "raw_stderr": "2.114", + "latency": "41.97", + "cost_per_test": "0.237579", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.114, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.981, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.981", + "raw_stderr": "1.312", + "latency": "24.478", + "cost_per_test": "0.130533", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.312, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.481, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.481", + "raw_stderr": "1.546", + "latency": "97.048", + "cost_per_test": "0.575076", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.546, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.597, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.597", + "raw_stderr": "1.446", + "latency": "29.677", + "cost_per_test": "0.15263", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.446, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.315, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.315", + "raw_stderr": "2.115", + "latency": "23.848", + "cost_per_test": "0.132189", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.115, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.656, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.656", + "raw_stderr": "1.342", + "latency": "66.38", + "cost_per_test": "0.312744", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.342, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.152, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.152", + "raw_stderr": "1.088", + "latency": "34.451", + "cost_per_test": "0.228935", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.088, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.498, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.498", + "raw_stderr": "1.176", + "latency": "19.814", + "cost_per_test": "0.099581", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.176, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.26, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.26", + "raw_stderr": "0.381", + "latency": "38.92", + "cost_per_test": "0.215138", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.381, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.379, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.379", + "raw_stderr": "1.391", + "latency": "24.329", + "cost_per_test": "0.124834", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.391, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.915, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.915", + "raw_stderr": "0.989", + "latency": "47.367", + "cost_per_test": "0.282054", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.989, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.221, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.221", + "raw_stderr": "1.187", + "latency": "32.451", + "cost_per_test": "0.111833", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.187, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/1925f8f6-ea04-4543-87bb-973c5c581e3b.json b/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/1925f8f6-ea04-4543-87bb-973c5c581e3b.json new file mode 100644 index 0000000000000000000000000000000000000000..869c9538f7199e06a54dca34dd2d764c70fc611a --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/1925f8f6-ea04-4543-87bb-973c5c581e3b.json @@ -0,0 +1,105 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/anthropic_claude-opus-4-5-20251101-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-5-20251101-thinking", + "id": "anthropic/claude-opus-4-5-20251101-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-5-20251101-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 82.948, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.948", + "raw_stderr": "0.904", + "latency": "75.28", + "cost_per_test": "0.126986", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.904, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/32b881ae-d628-48cf-9509-846cec823c03.json b/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/32b881ae-d628-48cf-9509-846cec823c03.json new file mode 100644 index 0000000000000000000000000000000000000000..652457cadcd4bb67cbd4e134f57b5f12b5ef70c9 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/32b881ae-d628-48cf-9509-846cec823c03.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/anthropic_claude-opus-4-5-20251101-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-5-20251101-thinking", + "id": "anthropic/claude-opus-4-5-20251101-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-5-20251101-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 76.232, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "76.232", + "raw_stderr": "1.2", + "latency": "37.862", + "cost_per_test": "0.221566", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.2, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 67.686, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.686", + "raw_stderr": "0.917", + "latency": "28.166", + "cost_per_test": "0.178388", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.917, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 59.141, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.141", + "raw_stderr": "1.386", + "latency": "18.469", + "cost_per_test": "0.135209", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.386, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/3e506f93-909e-47ab-9d29-562ae1aa0719.json b/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/3e506f93-909e-47ab-9d29-562ae1aa0719.json new file mode 100644 index 0000000000000000000000000000000000000000..9ff655d4afa8bd7fc689e0ee6647109d0905cb2d --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/3e506f93-909e-47ab-9d29-562ae1aa0719.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/anthropic_claude-opus-4-5-20251101-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-5-20251101-thinking", + "id": "anthropic/claude-opus-4-5-20251101-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-5-20251101-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 18.0, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "18", + "raw_stderr": "5.452", + "latency": "4779.544", + "cost_per_test": "3.923563", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.452, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 22.5, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "22.5", + "raw_stderr": "8.755", + "latency": "11643.441", + "cost_per_test": "3.923438", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 8.755, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "ioi:overall:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 20.25, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "20.25", + "raw_stderr": "5.157", + "latency": "8211.492", + "cost_per_test": "3.9235", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.157, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/65f091fb-f14b-4eeb-9c47-46e8dc517663.json b/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/65f091fb-f14b-4eeb-9c47-46e8dc517663.json new file mode 100644 index 0000000000000000000000000000000000000000..c7b09f96fecd57809dff5707261973b3e016c8b1 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/65f091fb-f14b-4eeb-9c47-46e8dc517663.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/anthropic_claude-opus-4-5-20251101-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-5-20251101-thinking", + "id": "anthropic/claude-opus-4-5-20251101-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-5-20251101-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 20.63, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "20.63", + "raw_stderr": "3.159", + "latency": "2283.142", + "cost_per_test": "32.873283", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.159, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/77624b58-5b2c-4d8c-bb99-e4a7c0b64ac4.json b/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/77624b58-5b2c-4d8c-bb99-e4a7c0b64ac4.json new file mode 100644 index 0000000000000000000000000000000000000000..6ef4e1eaaeedf1810ec9732857051bf80a3fc560 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/77624b58-5b2c-4d8c-bb99-e4a7c0b64ac4.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/anthropic_claude-opus-4-5-20251101-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-5-20251101-thinking", + "id": "anthropic/claude-opus-4-5-20251101-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-5-20251101-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 86.364, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.364", + "raw_stderr": "3.501", + "latency": "139.968", + "cost_per_test": "0.733867", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.501, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 85.354, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.354", + "raw_stderr": "3.501", + "latency": "137.082", + "cost_per_test": "0.691649", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.501, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 85.859, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.859", + "raw_stderr": "2.476", + "latency": "138.525", + "cost_per_test": "0.712758", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.476, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/a9e2139e-f41b-4233-98c9-6013976c3c55.json b/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/a9e2139e-f41b-4233-98c9-6013976c3c55.json new file mode 100644 index 0000000000000000000000000000000000000000..50b5b1559e2f98098ee770987ee16b6d0266948e --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/a9e2139e-f41b-4233-98c9-6013976c3c55.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/anthropic_claude-opus-4-5-20251101-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-5-20251101-thinking", + "id": "anthropic/claude-opus-4-5-20251101-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-5-20251101-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 63.859, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.859", + "raw_stderr": "1.374", + "latency": "47.448", + "cost_per_test": "0.218987", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.374, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 74.856, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "74.856", + "raw_stderr": "0.849", + "latency": "47.448", + "cost_per_test": "0.218987", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.849, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 85.854, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "85.854", + "raw_stderr": "0.997", + "latency": "47.448", + "cost_per_test": "0.218987", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.997, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/b0d13355-3ef7-4d14-a0df-2e5f454f7f94.json b/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/b0d13355-3ef7-4d14-a0df-2e5f454f7f94.json new file mode 100644 index 0000000000000000000000000000000000000000..98cffbc5be21506dfabd33c433c0c5a56b2d60ed --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/b0d13355-3ef7-4d14-a0df-2e5f454f7f94.json @@ -0,0 +1,381 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/anthropic_claude-opus-4-5-20251101-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-5-20251101-thinking", + "id": "anthropic/claude-opus-4-5-20251101-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-5-20251101-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 40.476, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "40.476", + "raw_stderr": "7.666", + "latency": "449.824", + "cost_per_test": "1.70738", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 7.666, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 74.713, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "74.713", + "raw_stderr": "2.696", + "latency": "348.779", + "cost_per_test": "1.146684", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.696, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 87.629, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "87.629", + "raw_stderr": "2.37", + "latency": "250.995", + "cost_per_test": "0.774675", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.37, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "551.372", + "cost_per_test": "2.045779", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:overall:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 76.4, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "76.4", + "raw_stderr": "1.901", + "latency": "320.542", + "cost_per_test": "1.054837", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.901, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/bad25b72-0219-4600-b040-ab94a7502242.json b/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/bad25b72-0219-4600-b040-ab94a7502242.json new file mode 100644 index 0000000000000000000000000000000000000000..0dacef4a54b9d2961c7f61c90f51ef168056b885 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/bad25b72-0219-4600-b040-ab94a7502242.json @@ -0,0 +1,362 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/anthropic_claude-opus-4-5-20251101-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-5-20251101-thinking", + "id": "anthropic/claude-opus-4-5-20251101-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-5-20251101-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 47.619, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "47.619", + "raw_stderr": "6.267", + "latency": "97.811", + "cost_per_test": "0.17995", + "temperature": "1", + "max_output_tokens": "30000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.267, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:automata:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 38.096, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "38.096", + "raw_stderr": "6.223", + "latency": "93.82", + "cost_per_test": "0.152698", + "temperature": "1", + "max_output_tokens": "30000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.223, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:calculus:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 73.457, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "73.457", + "raw_stderr": "5.53", + "latency": "103.527", + "cost_per_test": "0.145267", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.53, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 70.561, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "70.561", + "raw_stderr": "5.119", + "latency": "70.02", + "cost_per_test": "0.12869", + "temperature": "1", + "max_output_tokens": "30000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.119, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:overall:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 52.092, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "52.092", + "raw_stderr": "3.403", + "latency": "87.217", + "cost_per_test": "0.153779", + "temperature": "1", + "max_output_tokens": "30000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.403, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/bf73c96d-3c99-4730-971e-637bf12ea9a7.json b/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/bf73c96d-3c99-4730-971e-637bf12ea9a7.json new file mode 100644 index 0000000000000000000000000000000000000000..88e6ade64910ffdbdd892ff0d08bc9bb41c6ef14 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/bf73c96d-3c99-4730-971e-637bf12ea9a7.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/anthropic_claude-opus-4-5-20251101-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-5-20251101-thinking", + "id": "anthropic/claude-opus-4-5-20251101-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-5-20251101-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 36.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "36", + "raw_stderr": "4.8", + "latency": "3757.671", + "cost_per_test": "1.267043", + "temperature": "1", + "max_output_tokens": "32000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 4.8, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/c3a88a27-b24f-41ea-80bb-e3603ede7608.json b/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/c3a88a27-b24f-41ea-80bb-e3603ede7608.json new file mode 100644 index 0000000000000000000000000000000000000000..c97f4a5cd5428e66ddafccb020bc73f8b79a851e --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/c3a88a27-b24f-41ea-80bb-e3603ede7608.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/anthropic_claude-opus-4-5-20251101-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-5-20251101-thinking", + "id": "anthropic/claude-opus-4-5-20251101-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-5-20251101-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 85.321, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "85.321", + "raw_stderr": "1.896", + "latency": "72.112", + "cost_per_test": "0.410224", + "temperature": "1", + "max_output_tokens": "30000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.896, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/d25ae090-9aec-436b-bddc-34d1206ff30d.json b/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/d25ae090-9aec-436b-bddc-34d1206ff30d.json new file mode 100644 index 0000000000000000000000000000000000000000..e2d8315e76c816621ed88c0080e99550cea10d3b --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/d25ae090-9aec-436b-bddc-34d1206ff30d.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/anthropic_claude-opus-4-5-20251101-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-5-20251101-thinking", + "id": "anthropic/claude-opus-4-5-20251101-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-5-20251101-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 63.403, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.403", + "raw_stderr": "1.644", + "latency": "9.631", + "cost_per_test": "0.020493", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.644, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 65.618, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.618", + "raw_stderr": "1.622", + "latency": "23.184", + "cost_per_test": "0.024271", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.622, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 65.074, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.074", + "raw_stderr": "0.939", + "latency": "17.725", + "cost_per_test": "0.173423", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.939, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 66.2, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.2", + "raw_stderr": "1.615", + "latency": "20.361", + "cost_per_test": "0.475505", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.615, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/ffea60e2-b79a-4156-8e2f-3fc243021307.json b/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/ffea60e2-b79a-4156-8e2f-3fc243021307.json new file mode 100644 index 0000000000000000000000000000000000000000..d99fbcc147e0c08fb6e540772b85dda095a28d50 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-5-20251101-thinking/ffea60e2-b79a-4156-8e2f-3fc243021307.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/anthropic_claude-opus-4-5-20251101-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-5-20251101-thinking", + "id": "anthropic/claude-opus-4-5-20251101-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-5-20251101-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.758, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.758", + "raw_stderr": "0.617", + "latency": "40.099", + "cost_per_test": "0.261576", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.617, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:hard:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 64.0, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "64", + "raw_stderr": "2.566", + "latency": "302.373", + "cost_per_test": "1.713524", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.566, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:medium:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 88.251, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "88.251", + "raw_stderr": "1.645", + "latency": "150.809", + "cost_per_test": "0.937899", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.645, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:overall:anthropic/claude-opus-4-5-20251101-thinking:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 83.67, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "83.67", + "raw_stderr": "1.037", + "latency": "164.427", + "cost_per_test": "0.971", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.037, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-5-20251101/0f4be8ac-1539-4173-8e65-300fe3048484.json b/data/vals_ai/anthropic/claude-opus-4-5-20251101/0f4be8ac-1539-4173-8e65-300fe3048484.json new file mode 100644 index 0000000000000000000000000000000000000000..bee239fa4512bdd9ad2531538a0b0569bcae98ed --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-5-20251101/0f4be8ac-1539-4173-8e65-300fe3048484.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/anthropic_claude-opus-4-5-20251101/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-5-20251101", + "id": "anthropic/claude-opus-4-5-20251101", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-5-20251101", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 77.424, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "77.424", + "raw_stderr": "1.179", + "latency": "11.076", + "cost_per_test": "0.093589", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.179, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 68.68, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.68", + "raw_stderr": "0.908", + "latency": "10.534", + "cost_per_test": "0.092448", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.908, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 59.936, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.936", + "raw_stderr": "1.382", + "latency": "9.991", + "cost_per_test": "0.091307", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.382, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-5-20251101/58e08413-224e-4594-8603-fd847a21d655.json b/data/vals_ai/anthropic/claude-opus-4-5-20251101/58e08413-224e-4594-8603-fd847a21d655.json new file mode 100644 index 0000000000000000000000000000000000000000..845394a4687187ad0ac4c5b6dc5356b5cf2e8f73 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-5-20251101/58e08413-224e-4594-8603-fd847a21d655.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/anthropic_claude-opus-4-5-20251101/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-5-20251101", + "id": "anthropic/claude-opus-4-5-20251101", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-5-20251101", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 63.123, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.123", + "raw_stderr": "1.38", + "latency": "14.876", + "cost_per_test": "0.058665", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.38, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 74.325, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "74.325", + "raw_stderr": "0.854", + "latency": "14.876", + "cost_per_test": "0.058665", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.854, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 85.527, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "85.527", + "raw_stderr": "1.006", + "latency": "14.876", + "cost_per_test": "0.058665", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.006, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-5-20251101/7688c0ac-1b6f-452e-8c40-d0f80f8c5d43.json b/data/vals_ai/anthropic/claude-opus-4-5-20251101/7688c0ac-1b6f-452e-8c40-d0f80f8c5d43.json new file mode 100644 index 0000000000000000000000000000000000000000..b2c0901f3526d10c7e34e1c7771f32f8971a37c0 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-5-20251101/7688c0ac-1b6f-452e-8c40-d0f80f8c5d43.json @@ -0,0 +1,1071 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/anthropic_claude-opus-4-5-20251101/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-5-20251101", + "id": "anthropic/claude-opus-4-5-20251101", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-5-20251101", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.003, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.003", + "raw_stderr": "0.994", + "latency": "8.647", + "cost_per_test": "0.05251", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.994, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.862, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.862", + "raw_stderr": "1.448", + "latency": "6.531", + "cost_per_test": "0.04245", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.448, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.488, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.488", + "raw_stderr": "1.108", + "latency": "8.301", + "cost_per_test": "0.064481", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.108, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.146, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.146", + "raw_stderr": "1.975", + "latency": "8.62", + "cost_per_test": "0.057543", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.975, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.573, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.573", + "raw_stderr": "1.068", + "latency": "7.509", + "cost_per_test": "0.044699", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.068, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.161, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.161", + "raw_stderr": "1.543", + "latency": "11.153", + "cost_per_test": "0.072202", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.543, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.985, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.985", + "raw_stderr": "1.321", + "latency": "8.045", + "cost_per_test": "0.04467", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.321, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.478, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.478", + "raw_stderr": "2.124", + "latency": "8.614", + "cost_per_test": "0.061382", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.124, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.292, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.292", + "raw_stderr": "1.254", + "latency": "10.208", + "cost_per_test": "0.054909", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.254, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.302, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.302", + "raw_stderr": "1.063", + "latency": "7.848", + "cost_per_test": "0.065281", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.063, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.632, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.632", + "raw_stderr": "1.186", + "latency": "6.863", + "cost_per_test": "0.035681", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.186, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.59, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.59", + "raw_stderr": "0.367", + "latency": "8.248", + "cost_per_test": "0.052065", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.367, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.776, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.776", + "raw_stderr": "1.477", + "latency": "7.596", + "cost_per_test": "0.040147", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.477, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.838, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.838", + "raw_stderr": "0.902", + "latency": "7.714", + "cost_per_test": "0.050949", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.902, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.722, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.722", + "raw_stderr": "1.125", + "latency": "7.824", + "cost_per_test": "0.042011", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.125, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-5-20251101/a5a541b4-cfdd-4862-8218-8bdac88a1d45.json b/data/vals_ai/anthropic/claude-opus-4-5-20251101/a5a541b4-cfdd-4862-8218-8bdac88a1d45.json new file mode 100644 index 0000000000000000000000000000000000000000..424fcd1773ed5ec96f6c7afbaca4a9cb53443ebe --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-5-20251101/a5a541b4-cfdd-4862-8218-8bdac88a1d45.json @@ -0,0 +1,105 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/anthropic_claude-opus-4-5-20251101/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-5-20251101", + "id": "anthropic/claude-opus-4-5-20251101", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-5-20251101", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 81.098, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.098", + "raw_stderr": "0.941", + "latency": "19.557", + "cost_per_test": "0.021549", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.941, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-5-20251101/a9e16721-2f05-4712-a098-2867b51cc933.json b/data/vals_ai/anthropic/claude-opus-4-5-20251101/a9e16721-2f05-4712-a098-2867b51cc933.json new file mode 100644 index 0000000000000000000000000000000000000000..c4859378e107fe4c14461c6c25bade1763125594 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-5-20251101/a9e16721-2f05-4712-a098-2867b51cc933.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/anthropic_claude-opus-4-5-20251101/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-5-20251101", + "id": "anthropic/claude-opus-4-5-20251101", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-5-20251101", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 83.246, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "83.246", + "raw_stderr": "1.926", + "latency": "43.305", + "cost_per_test": "0.281674", + "temperature": "1", + "max_output_tokens": "30000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.926, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-5-20251101/af5ba849-89b0-46bf-9a60-5be3f963dcae.json b/data/vals_ai/anthropic/claude-opus-4-5-20251101/af5ba849-89b0-46bf-9a60-5be3f963dcae.json new file mode 100644 index 0000000000000000000000000000000000000000..b269a349b4234a28f06972a491ae88c581c9d091 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-5-20251101/af5ba849-89b0-46bf-9a60-5be3f963dcae.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/anthropic_claude-opus-4-5-20251101/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-5-20251101", + "id": "anthropic/claude-opus-4-5-20251101", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-5-20251101", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 79.293, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.293", + "raw_stderr": "3.494", + "latency": "14.455", + "cost_per_test": "0.092902", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.494, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 79.798, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.798", + "raw_stderr": "3.472", + "latency": "12.625", + "cost_per_test": "0.056963", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.472, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 79.546, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.546", + "raw_stderr": "2.463", + "latency": "13.54", + "cost_per_test": "0.074932", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.463, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-5-20251101/b643b25c-f340-493e-a3f7-9778d328d6a8.json b/data/vals_ai/anthropic/claude-opus-4-5-20251101/b643b25c-f340-493e-a3f7-9778d328d6a8.json new file mode 100644 index 0000000000000000000000000000000000000000..c444ba9e3ee50b3ec8a75d91def0951f9791f588 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-5-20251101/b643b25c-f340-493e-a3f7-9778d328d6a8.json @@ -0,0 +1,362 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/anthropic_claude-opus-4-5-20251101/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-5-20251101", + "id": "anthropic/claude-opus-4-5-20251101", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-5-20251101", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 41.27, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "41.27", + "raw_stderr": "6.178", + "latency": "23.709", + "cost_per_test": "0.020485", + "temperature": "1", + "max_output_tokens": "30000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.178, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:automata:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 29.365, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "29.365", + "raw_stderr": "5.832", + "latency": "23.07", + "cost_per_test": "0.016674", + "temperature": "1", + "max_output_tokens": "30000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.832, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:calculus:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 58.333, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "58.333", + "raw_stderr": "6.139", + "latency": "39.018", + "cost_per_test": "0.024556", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.139, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 64.372, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "64.372", + "raw_stderr": "5.379", + "latency": "21.402", + "cost_per_test": "0.019286", + "temperature": "1", + "max_output_tokens": "30000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.379, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:overall:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 45.002, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "45.002", + "raw_stderr": "3.352", + "latency": "22.727", + "cost_per_test": "0.018815", + "temperature": "1", + "max_output_tokens": "30000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.352, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-5-20251101/b99301c1-71c9-40da-a9ef-983fecf55d93.json b/data/vals_ai/anthropic/claude-opus-4-5-20251101/b99301c1-71c9-40da-a9ef-983fecf55d93.json new file mode 100644 index 0000000000000000000000000000000000000000..817049b717dd3f187b65341f49a80bc8530fff47 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-5-20251101/b99301c1-71c9-40da-a9ef-983fecf55d93.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/anthropic_claude-opus-4-5-20251101/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-5-20251101", + "id": "anthropic/claude-opus-4-5-20251101", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-5-20251101", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 20.667, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "20.667", + "raw_stderr": "11.43", + "latency": "11576.195", + "cost_per_test": "2.959818", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 11.43, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 26.5, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "26.5", + "raw_stderr": "9.857", + "latency": "3788.571", + "cost_per_test": "3.137652", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 9.857, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "ioi:overall:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 23.584, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "23.584", + "raw_stderr": "7.547", + "latency": "7682.383", + "cost_per_test": "3.048735", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 7.547, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-5-20251101/caa4eab3-980b-4255-88d8-a3aabd4854b9.json b/data/vals_ai/anthropic/claude-opus-4-5-20251101/caa4eab3-980b-4255-88d8-a3aabd4854b9.json new file mode 100644 index 0000000000000000000000000000000000000000..abde847c4f462063b3b10e68baad18f466666955 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-5-20251101/caa4eab3-980b-4255-88d8-a3aabd4854b9.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/anthropic_claude-opus-4-5-20251101/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-5-20251101", + "id": "anthropic/claude-opus-4-5-20251101", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-5-20251101", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 99.068, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "99.068", + "raw_stderr": "0.535", + "latency": "4.841", + "cost_per_test": "0.026073", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:hard:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 44.571, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "44.571", + "raw_stderr": "2.657", + "latency": "22.051", + "cost_per_test": "0.124578", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.657, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:medium:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 81.462, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "81.462", + "raw_stderr": "1.986", + "latency": "13.904", + "cost_per_test": "0.081319", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.986, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:overall:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 75.034, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "75.034", + "raw_stderr": "1.12", + "latency": "13.599", + "cost_per_test": "0.077323", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.12, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-5-20251101/ce369907-731e-4c38-8c2c-77b456ddc52b.json b/data/vals_ai/anthropic/claude-opus-4-5-20251101/ce369907-731e-4c38-8c2c-77b456ddc52b.json new file mode 100644 index 0000000000000000000000000000000000000000..f21770e97bbbe6d1e18057a1a253b6b76f6b4464 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-5-20251101/ce369907-731e-4c38-8c2c-77b456ddc52b.json @@ -0,0 +1,450 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/anthropic_claude-opus-4-5-20251101/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-5-20251101", + "id": "anthropic/claude-opus-4-5-20251101", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-5-20251101", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.7, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.7", + "raw_stderr": "0.875", + "latency": "2.138", + "cost_per_test": "0.007212", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.875, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.235, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.235", + "raw_stderr": "0.188", + "latency": "2.149", + "cost_per_test": "0.012917", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.188, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 84.152, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.152", + "raw_stderr": "0.638", + "latency": "2.225", + "cost_per_test": "0.024892", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.638, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.837, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.837", + "raw_stderr": "0.434", + "latency": "2.376", + "cost_per_test": "0.01319", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.434, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.777, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.777", + "raw_stderr": "1.202", + "latency": "2.187", + "cost_per_test": "0.013139", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.202, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.32, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.32", + "raw_stderr": "1.433", + "latency": "3.18", + "cost_per_test": "0.007788", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.433, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-5-20251101/ed2c5c94-e71a-4c0d-b334-ddf04d69d072.json b/data/vals_ai/anthropic/claude-opus-4-5-20251101/ed2c5c94-e71a-4c0d-b334-ddf04d69d072.json new file mode 100644 index 0000000000000000000000000000000000000000..43fbef177583cd98e136f7ea2d6bfd7e2a8e865f --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-5-20251101/ed2c5c94-e71a-4c0d-b334-ddf04d69d072.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/anthropic_claude-opus-4-5-20251101/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-5-20251101", + "id": "anthropic/claude-opus-4-5-20251101", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-5-20251101", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.839, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.839", + "raw_stderr": "1.666", + "latency": "11.462", + "cost_per_test": "0.007209", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.666, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.723, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.723", + "raw_stderr": "1.667", + "latency": "5.786", + "cost_per_test": "0.011333", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.667, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.305, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.305", + "raw_stderr": "0.96", + "latency": "9.401", + "cost_per_test": "0.150159", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.96, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:anthropic/claude-opus-4-5-20251101:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 62.354, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.354", + "raw_stderr": "1.654", + "latency": "10.956", + "cost_per_test": "0.431934", + "temperature": "1", + "max_output_tokens": "64000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.654, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-6-thinking/07dfbee9-0f1b-40ac-a9cf-c05f44a43c25.json b/data/vals_ai/anthropic/claude-opus-4-6-thinking/07dfbee9-0f1b-40ac-a9cf-c05f44a43c25.json new file mode 100644 index 0000000000000000000000000000000000000000..d5181716bd3d9eb376f79f18aaf8738c662c4967 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-6-thinking/07dfbee9-0f1b-40ac-a9cf-c05f44a43c25.json @@ -0,0 +1,104 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/anthropic_claude-opus-4-6-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-6-thinking", + "id": "anthropic/claude-opus-4-6-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-6-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 53.498, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "53.498", + "raw_stderr": "4.678", + "latency": "1386.056", + "cost_per_test": "8.279062", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 4.678, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-6-thinking/29184bcf-f11a-49a7-9724-9e155bee49a8.json b/data/vals_ai/anthropic/claude-opus-4-6-thinking/29184bcf-f11a-49a7-9724-9e155bee49a8.json new file mode 100644 index 0000000000000000000000000000000000000000..eb39556d2a7a8c166b7790b91f5c459d613d005f --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-6-thinking/29184bcf-f11a-49a7-9724-9e155bee49a8.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/anthropic_claude-opus-4-6-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-6-thinking", + "id": "anthropic/claude-opus-4-6-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-6-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 88.384, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.384", + "raw_stderr": "1.115", + "latency": "337.27", + "cost_per_test": "0.310954", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.115, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 90.909, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.909", + "raw_stderr": "2.093", + "latency": "300.77", + "cost_per_test": "0.278848", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.093, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 89.646, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.646", + "raw_stderr": "1.186", + "latency": "319.02", + "cost_per_test": "0.294901", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.186, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-6-thinking/408b7efa-3f76-4c27-891d-b3f9f080cc66.json b/data/vals_ai/anthropic/claude-opus-4-6-thinking/408b7efa-3f76-4c27-891d-b3f9f080cc66.json new file mode 100644 index 0000000000000000000000000000000000000000..0d494abe53c0bb3fa65506f15907e6af53659b9e --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-6-thinking/408b7efa-3f76-4c27-891d-b3f9f080cc66.json @@ -0,0 +1,105 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/anthropic_claude-opus-4-6-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-6-thinking", + "id": "anthropic/claude-opus-4-6-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-6-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 83.873, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.873", + "raw_stderr": "0.884", + "latency": "119.005", + "cost_per_test": "0.116094", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.884, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-6-thinking/4239a422-a805-43f1-b1ff-83a1703ee2c6.json b/data/vals_ai/anthropic/claude-opus-4-6-thinking/4239a422-a805-43f1-b1ff-83a1703ee2c6.json new file mode 100644 index 0000000000000000000000000000000000000000..f048957ad5e3dccc100822a6494bec6a66b7bbff --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-6-thinking/4239a422-a805-43f1-b1ff-83a1703ee2c6.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/anthropic_claude-opus-4-6-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-6-thinking", + "id": "anthropic/claude-opus-4-6-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-6-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 50.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "50", + "raw_stderr": "5", + "latency": "3973.892", + "cost_per_test": "2.703583", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-6-thinking/4ca370de-75a6-4d85-8121-8cdf2b922a86.json b/data/vals_ai/anthropic/claude-opus-4-6-thinking/4ca370de-75a6-4d85-8121-8cdf2b922a86.json new file mode 100644 index 0000000000000000000000000000000000000000..087442a853ed1fd37ac41b6c2bd64a85d9cd381f --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-6-thinking/4ca370de-75a6-4d85-8121-8cdf2b922a86.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/anthropic_claude-opus-4-6-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-6-thinking", + "id": "anthropic/claude-opus-4-6-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-6-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 86.13, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "86.13", + "raw_stderr": "1.944", + "latency": "127.564", + "cost_per_test": "0.224735", + "temperature": "1", + "max_output_tokens": "30000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.944, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-6-thinking/50b6dd42-5bb5-4220-a0d9-5e2397ede081.json b/data/vals_ai/anthropic/claude-opus-4-6-thinking/50b6dd42-5bb5-4220-a0d9-5e2397ede081.json new file mode 100644 index 0000000000000000000000000000000000000000..369d6b0193b00265443781923f0c49147d1317f0 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-6-thinking/50b6dd42-5bb5-4220-a0d9-5e2397ede081.json @@ -0,0 +1,450 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/anthropic_claude-opus-4-6-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-6-thinking", + "id": "anthropic/claude-opus-4-6-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-6-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 88.795, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.795", + "raw_stderr": "0.703", + "latency": "3.904", + "cost_per_test": "0.003218", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.703, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.164, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.164", + "raw_stderr": "0.177", + "latency": "2.584", + "cost_per_test": "0.003341", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.177, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 90.685, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.685", + "raw_stderr": "0.542", + "latency": "3.921", + "cost_per_test": "0.006478", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.542, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 85.301, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.301", + "raw_stderr": "0.368", + "latency": "3.755", + "cost_per_test": "0.004043", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.368, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.482, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.482", + "raw_stderr": "0.9", + "latency": "2.953", + "cost_per_test": "0.003578", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.9, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.377, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.377", + "raw_stderr": "1.322", + "latency": "5.414", + "cost_per_test": "0.0036", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.322, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-6-thinking/85d43500-07fd-48e6-811f-d368470e5a5b.json b/data/vals_ai/anthropic/claude-opus-4-6-thinking/85d43500-07fd-48e6-811f-d368470e5a5b.json new file mode 100644 index 0000000000000000000000000000000000000000..ec4f80efa183b143de211ae89a7faef0f388c492 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-6-thinking/85d43500-07fd-48e6-811f-d368470e5a5b.json @@ -0,0 +1,362 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/anthropic_claude-opus-4-6-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-6-thinking", + "id": "anthropic/claude-opus-4-6-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-6-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 41.27, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "41.27", + "raw_stderr": "6.178", + "latency": "200.147", + "cost_per_test": "0.429431", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.178, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "sage:automata:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 38.095, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "38.095", + "raw_stderr": "6.227", + "latency": "152.08", + "cost_per_test": "0.274423", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.227, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "sage:calculus:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 72.531, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "72.531", + "raw_stderr": "5.081", + "latency": "263.947", + "cost_per_test": "0.822692", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.081, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 75.361, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "75.361", + "raw_stderr": "4.84", + "latency": "130.677", + "cost_per_test": "0.35649", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 4.84, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "sage:overall:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 51.575, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "51.575", + "raw_stderr": "3.339", + "latency": "160.968", + "cost_per_test": "0.353448", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.339, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-6-thinking/a85b7213-0af9-4b35-9d0d-f3fe93a135e8.json b/data/vals_ai/anthropic/claude-opus-4-6-thinking/a85b7213-0af9-4b35-9d0d-f3fe93a135e8.json new file mode 100644 index 0000000000000000000000000000000000000000..e357715224ab16b7250e388c94e62278c778ea47 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-6-thinking/a85b7213-0af9-4b35-9d0d-f3fe93a135e8.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/anthropic_claude-opus-4-6-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-6-thinking", + "id": "anthropic/claude-opus-4-6-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-6-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.447, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.447", + "raw_stderr": "0.689", + "latency": "15.289", + "cost_per_test": "0.018572", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.689, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "lcb:hard:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 66.286, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66.286", + "raw_stderr": "2.527", + "latency": "798.957", + "cost_per_test": "0.829373", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.527, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "lcb:medium:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 89.295, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "89.295", + "raw_stderr": "1.58", + "latency": "159.213", + "cost_per_test": "0.182375", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.58, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "lcb:overall:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 84.676, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "84.676", + "raw_stderr": "1.02", + "latency": "324.486", + "cost_per_test": "0.34344", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.02, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-6-thinking/b6d8212c-f547-4c51-b52a-7e6b46e8e852.json b/data/vals_ai/anthropic/claude-opus-4-6-thinking/b6d8212c-f547-4c51-b52a-7e6b46e8e852.json new file mode 100644 index 0000000000000000000000000000000000000000..0d8b706030e3bc5eafc40e98f85a5a402f199b4e --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-6-thinking/b6d8212c-f547-4c51-b52a-7e6b46e8e852.json @@ -0,0 +1,381 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/anthropic_claude-opus-4-6-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-6-thinking", + "id": "anthropic/claude-opus-4-6-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-6-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 42.857, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "42.857", + "raw_stderr": "7.729", + "latency": "606.929", + "cost_per_test": "2.467066", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 7.729, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 75.479, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "75.479", + "raw_stderr": "2.668", + "latency": "384.184", + "cost_per_test": "1.345821", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.668, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 90.206, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "90.206", + "raw_stderr": "2.14", + "latency": "248.78", + "cost_per_test": "0.765978", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.14, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "451.137", + "cost_per_test": "1.586181", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:overall:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 78.2, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "78.2", + "raw_stderr": "1.848", + "latency": "350.76", + "cost_per_test": "1.216469", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.848, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-6-thinking/bf50bbce-45a3-4443-8066-c66e2d2c7b69.json b/data/vals_ai/anthropic/claude-opus-4-6-thinking/bf50bbce-45a3-4443-8066-c66e2d2c7b69.json new file mode 100644 index 0000000000000000000000000000000000000000..fd13d75eaf0ceb52171175b5703566a2dac81117 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-6-thinking/bf50bbce-45a3-4443-8066-c66e2d2c7b69.json @@ -0,0 +1,1071 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/anthropic_claude-opus-4-6-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-6-thinking", + "id": "anthropic/claude-opus-4-6-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-6-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.003, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.003", + "raw_stderr": "1.843", + "latency": "20.744", + "cost_per_test": "0.022805", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.843, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.861, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.861", + "raw_stderr": "1.528", + "latency": "42.002", + "cost_per_test": "0.045443", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.528, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.286, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.286", + "raw_stderr": "1.074", + "latency": "67.055", + "cost_per_test": "0.077784", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.074, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.732, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.732", + "raw_stderr": "2.303", + "latency": "48.764", + "cost_per_test": "0.052137", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.303, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.284, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.284", + "raw_stderr": "1.684", + "latency": "21.494", + "cost_per_test": "0.022582", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.684, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.1, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.1", + "raw_stderr": "1.159", + "latency": "225.755", + "cost_per_test": "0.236159", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.159, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.597, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.597", + "raw_stderr": "1.603", + "latency": "25.958", + "cost_per_test": "0.024715", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.603, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.677, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.677", + "raw_stderr": "2.553", + "latency": "20.649", + "cost_per_test": "0.022339", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.553, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.746, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.746", + "raw_stderr": "1.506", + "latency": "60.609", + "cost_per_test": "0.055171", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.506, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 95.929, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "95.929", + "raw_stderr": "1.024", + "latency": "41.961", + "cost_per_test": "0.051479", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.024, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.905, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.905", + "raw_stderr": "1.635", + "latency": "17.063", + "cost_per_test": "0.016102", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.635, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.107, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.107", + "raw_stderr": "0.455", + "latency": "49.451", + "cost_per_test": "0.052129", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.455, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.377, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.377", + "raw_stderr": "2.136", + "latency": "19.566", + "cost_per_test": "0.01872", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.136, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.149, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.149", + "raw_stderr": "1.175", + "latency": "61.982", + "cost_per_test": "0.066938", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.175, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.85, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.85", + "raw_stderr": "1.767", + "latency": "18.714", + "cost_per_test": "0.017434", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.767, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-6-thinking/c13e5bf3-22dd-433e-a2ca-adb730a6fe55.json b/data/vals_ai/anthropic/claude-opus-4-6-thinking/c13e5bf3-22dd-433e-a2ca-adb730a6fe55.json new file mode 100644 index 0000000000000000000000000000000000000000..81b069814cfb8c180d3eb7fee0fe418e6e11d570 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-6-thinking/c13e5bf3-22dd-433e-a2ca-adb730a6fe55.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/anthropic_claude-opus-4-6-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-6-thinking", + "id": "anthropic/claude-opus-4-6-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-6-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 76.948, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "76.948", + "raw_stderr": "1.187", + "latency": "39.421", + "cost_per_test": "0.079983", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.187, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 68.522, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.522", + "raw_stderr": "0.911", + "latency": "27.161", + "cost_per_test": "0.060749", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.911, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 60.095, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.095", + "raw_stderr": "1.381", + "latency": "14.901", + "cost_per_test": "0.041515", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.381, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-6-thinking/db68a098-8921-47f8-a592-d5cd075762c9.json b/data/vals_ai/anthropic/claude-opus-4-6-thinking/db68a098-8921-47f8-a592-d5cd075762c9.json new file mode 100644 index 0000000000000000000000000000000000000000..0fa1af6591c29e25cdc105e8ebe23803ec7ef803 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-6-thinking/db68a098-8921-47f8-a592-d5cd075762c9.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/anthropic_claude-opus-4-6-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-6-thinking", + "id": "anthropic/claude-opus-4-6-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-6-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 64.759, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.759", + "raw_stderr": "1.366", + "latency": "79.254", + "cost_per_test": "0.071368", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.366, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 75.961, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "75.961", + "raw_stderr": "0.834", + "latency": "79.254", + "cost_per_test": "0.071368", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.834, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 87.163, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "87.163", + "raw_stderr": "0.957", + "latency": "79.254", + "cost_per_test": "0.071368", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.957, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-6-thinking/ecd2e441-4506-4c07-ad01-ea77fa6d2281.json b/data/vals_ai/anthropic/claude-opus-4-6-thinking/ecd2e441-4506-4c07-ad01-ea77fa6d2281.json new file mode 100644 index 0000000000000000000000000000000000000000..9f43516a015c24e68011b12377f59256a1c61a42 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-6-thinking/ecd2e441-4506-4c07-ad01-ea77fa6d2281.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/anthropic_claude-opus-4-6-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-6-thinking", + "id": "anthropic/claude-opus-4-6-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-6-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 64.918, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.918", + "raw_stderr": "1.629", + "latency": "10.405", + "cost_per_test": "0.022285", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.629, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 67.133, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.133", + "raw_stderr": "1.604", + "latency": "28.963", + "cost_per_test": "0.9616", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.604, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 67.016, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.016", + "raw_stderr": "0.926", + "latency": "20.581", + "cost_per_test": "0.487461", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.926, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:anthropic/claude-opus-4-6-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 68.998, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.998", + "raw_stderr": "1.579", + "latency": "22.376", + "cost_per_test": "0.478497", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.579, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-6/d12210e1-de8c-41c6-a540-9a2883ab29ae.json b/data/vals_ai/anthropic/claude-opus-4-6/d12210e1-de8c-41c6-a540-9a2883ab29ae.json new file mode 100644 index 0000000000000000000000000000000000000000..c4626c82c44f6f10631bbf644e6fae505f9caf96 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-6/d12210e1-de8c-41c6-a540-9a2883ab29ae.json @@ -0,0 +1,104 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/anthropic_claude-opus-4-6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-6", + "id": "anthropic/claude-opus-4-6", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-6", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:anthropic/claude-opus-4-6:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 57.573, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "57.573", + "raw_stderr": "4.371", + "latency": "1278.658", + "cost_per_test": "8.687436", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "false", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 4.371, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "false", + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-6/f55c408e-6fa5-4b5c-b9ef-8c0c256600d6.json b/data/vals_ai/anthropic/claude-opus-4-6/f55c408e-6fa5-4b5c-b9ef-8c0c256600d6.json new file mode 100644 index 0000000000000000000000000000000000000000..a04b2afcd79fb5b488396cc0abdd9b474b0b39e4 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-6/f55c408e-6fa5-4b5c-b9ef-8c0c256600d6.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/anthropic_claude-opus-4-6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-6", + "id": "anthropic/claude-opus-4-6", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-6", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:anthropic/claude-opus-4-6:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 86.738, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "86.738", + "raw_stderr": "1.942", + "latency": "54.318", + "cost_per_test": "0.115121", + "temperature": "1", + "max_output_tokens": "30000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.942, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-7/0a744eb1-76b8-43f3-9d2d-e237572e3c1c.json b/data/vals_ai/anthropic/claude-opus-4-7/0a744eb1-76b8-43f3-9d2d-e237572e3c1c.json new file mode 100644 index 0000000000000000000000000000000000000000..c7699768d633121be51ca972b83a9722fe9aeb99 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-7/0a744eb1-76b8-43f3-9d2d-e237572e3c1c.json @@ -0,0 +1,762 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/anthropic_claude-opus-4-7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-7", + "id": "anthropic/claude-opus-4-7", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-7", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 45.366, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "45.366", + "raw_stderr": "0.975", + "latency": "294.357", + "cost_per_test": "3.219143", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.975, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 38.647, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "38.647", + "raw_stderr": "0.617", + "latency": "360.437", + "cost_per_test": "4.032269", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.617, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 45.556, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "45.556", + "raw_stderr": "0.209", + "latency": "435.039", + "cost_per_test": "6.088803", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.209, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 61.34, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.34", + "raw_stderr": "0.27", + "latency": "364.799", + "cost_per_test": "4.322777", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.27, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 71.562, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.562", + "raw_stderr": "1.312", + "latency": "264.962", + "cost_per_test": "1.342081", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.312, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 20.096, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "20.096", + "raw_stderr": "1.765", + "latency": "505.87", + "cost_per_test": "4.408222", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.765, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 65.432, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.432", + "raw_stderr": "0.781", + "latency": "351.773", + "cost_per_test": "4.402011", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.781, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 70.054, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "70.054", + "raw_stderr": "1.051", + "latency": "238.532", + "cost_per_test": "3.744437", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.051, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 62.673, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.673", + "raw_stderr": "1.978", + "latency": "253.644", + "cost_per_test": "1.96134", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.978, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:overall:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 51.509, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "51.509", + "raw_stderr": "0.49", + "latency": "360.437", + "cost_per_test": "4.032269", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.49, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 21.497, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "21.497", + "raw_stderr": "1.086", + "latency": "570.859", + "cost_per_test": "7.128385", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.086, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-7/15dcfea0-03b2-4731-9e26-66deca30b1c1.json b/data/vals_ai/anthropic/claude-opus-4-7/15dcfea0-03b2-4731-9e26-66deca30b1c1.json new file mode 100644 index 0000000000000000000000000000000000000000..13fb1f98176ccf002059d13e38fc70656cbe647d --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-7/15dcfea0-03b2-4731-9e26-66deca30b1c1.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/anthropic_claude-opus-4-7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-7", + "id": "anthropic/claude-opus-4-7", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-7", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 90.909, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.909", + "raw_stderr": "2.043", + "latency": "203.874", + "cost_per_test": "0.352728", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.043, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 89.394, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.394", + "raw_stderr": "2.319", + "latency": "201.167", + "cost_per_test": "0.377529", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.319, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 90.152, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.152", + "raw_stderr": "1.545", + "latency": "202.52", + "cost_per_test": "0.365129", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.545, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-7/1fa8f67f-f93d-4819-a9b6-445f150dda7e.json b/data/vals_ai/anthropic/claude-opus-4-7/1fa8f67f-f93d-4819-a9b6-445f150dda7e.json new file mode 100644 index 0000000000000000000000000000000000000000..3b96c127f062a9b45910d6a801503010a9aa8281 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-7/1fa8f67f-f93d-4819-a9b6-445f150dda7e.json @@ -0,0 +1,1071 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/anthropic_claude-opus-4-7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-7", + "id": "anthropic/claude-opus-4-7", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-7", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.561, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.561", + "raw_stderr": "0.847", + "latency": "64.371", + "cost_per_test": "0.03085", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.847, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.283, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.283", + "raw_stderr": "0.944", + "latency": "79.437", + "cost_per_test": "0.050755", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.944, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.286, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.286", + "raw_stderr": "0.744", + "latency": "549.138", + "cost_per_test": "0.104175", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.744, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.683, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.683", + "raw_stderr": "1.286", + "latency": "97.967", + "cost_per_test": "0.042102", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.286, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.299, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.299", + "raw_stderr": "0.955", + "latency": "609.105", + "cost_per_test": "0.036308", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.955, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.029, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.029", + "raw_stderr": "1.091", + "latency": "319.67", + "cost_per_test": "0.248968", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.091, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.396, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.396", + "raw_stderr": "1.332", + "latency": "78.119", + "cost_per_test": "0.022124", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.332, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.202, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.202", + "raw_stderr": "1.962", + "latency": "134.626", + "cost_per_test": "0.04407", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.962, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.107, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.107", + "raw_stderr": "1.184", + "latency": "259.071", + "cost_per_test": "0.098415", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.184, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 95.855, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "95.855", + "raw_stderr": "0.542", + "latency": "79.094", + "cost_per_test": "0.042614", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.542, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.905, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.905", + "raw_stderr": "1.162", + "latency": "35.837", + "cost_per_test": "0.031505", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.162, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.871, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.871", + "raw_stderr": "0.305", + "latency": "187.83", + "cost_per_test": "0.061711", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.305, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.978, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.978", + "raw_stderr": "1.456", + "latency": "195.014", + "cost_per_test": "0.030079", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.456, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.38, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.38", + "raw_stderr": "0.69", + "latency": "90.591", + "cost_per_test": "0.053061", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.69, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.228, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.228", + "raw_stderr": "1.045", + "latency": "37.575", + "cost_per_test": "0.028932", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.045, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-7/28876c7d-37fd-4ead-a318-816a344b8eb2.json b/data/vals_ai/anthropic/claude-opus-4-7/28876c7d-37fd-4ead-a318-816a344b8eb2.json new file mode 100644 index 0000000000000000000000000000000000000000..22413a9250ecaa41d94b2bae129eaeb37c3671a5 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-7/28876c7d-37fd-4ead-a318-816a344b8eb2.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/anthropic_claude-opus-4-7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-7", + "id": "anthropic/claude-opus-4-7", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-7", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 63.778, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.778", + "raw_stderr": "1.374", + "latency": "62.071", + "cost_per_test": "0.124592", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.374, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 75.266, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "75.266", + "raw_stderr": "0.841", + "latency": "62.071", + "cost_per_test": "0.124592", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.841, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 86.754, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "86.754", + "raw_stderr": "0.969", + "latency": "62.071", + "cost_per_test": "0.124592", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.969, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-7/3e8d3086-38ba-40f0-954c-5092702b52bb.json b/data/vals_ai/anthropic/claude-opus-4-7/3e8d3086-38ba-40f0-954c-5092702b52bb.json new file mode 100644 index 0000000000000000000000000000000000000000..56e4b9309909b47e648a9a0ff6a6f7e1222b7a27 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-7/3e8d3086-38ba-40f0-954c-5092702b52bb.json @@ -0,0 +1,450 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/anthropic_claude-opus-4-7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-7", + "id": "anthropic/claude-opus-4-7", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-7", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 89.983, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.983", + "raw_stderr": "0.934", + "latency": "16.717", + "cost_per_test": "0.007066", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.934, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.609, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.609", + "raw_stderr": "0.204", + "latency": "24.393", + "cost_per_test": "0.007695", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.204, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 88.947, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.947", + "raw_stderr": "0.573", + "latency": "27.488", + "cost_per_test": "0.013182", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.573, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 85.251, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.251", + "raw_stderr": "0.451", + "latency": "26.622", + "cost_per_test": "0.008332", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.451, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.722, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.722", + "raw_stderr": "1.206", + "latency": "17.819", + "cost_per_test": "0.006968", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.206, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 84.996, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.996", + "raw_stderr": "1.544", + "latency": "46.694", + "cost_per_test": "0.006751", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.544, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-7/45b666af-f510-4a3e-a254-d1d3d071595a.json b/data/vals_ai/anthropic/claude-opus-4-7/45b666af-f510-4a3e-a254-d1d3d071595a.json new file mode 100644 index 0000000000000000000000000000000000000000..82b9d24fa33fe59f33c28abd77524f9b55461d96 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-7/45b666af-f510-4a3e-a254-d1d3d071595a.json @@ -0,0 +1,362 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/anthropic_claude-opus-4-7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-7", + "id": "anthropic/claude-opus-4-7", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-7", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 47.619, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "47.619", + "raw_stderr": "6.27", + "latency": "131.155", + "cost_per_test": "0.353409", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.27, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "sage:automata:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 43.65, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "43.65", + "raw_stderr": "6.318", + "latency": "145.362", + "cost_per_test": "0.329478", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.318, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "sage:calculus:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 75.926, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "75.926", + "raw_stderr": "5.396", + "latency": "121.638", + "cost_per_test": "0.367952", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.396, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 77.039, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "77.039", + "raw_stderr": "4.779", + "latency": "97.423", + "cost_per_test": "0.309413", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 4.779, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "sage:overall:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 56.103, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "56.103", + "raw_stderr": "3.368", + "latency": "124.647", + "cost_per_test": "0.330767", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.368, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-7/49bc2a55-f3c6-4ceb-8a45-053cb54c3ed1.json b/data/vals_ai/anthropic/claude-opus-4-7/49bc2a55-f3c6-4ceb-8a45-053cb54c3ed1.json new file mode 100644 index 0000000000000000000000000000000000000000..0ed72f9826ba744b32fce6a6332dc8ca4b384385 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-7/49bc2a55-f3c6-4ceb-8a45-053cb54c3ed1.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/anthropic_claude-opus-4-7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-7", + "id": "anthropic/claude-opus-4-7", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-7", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 71.003, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "71.003", + "raw_stderr": "4.506", + "latency": "2141.902", + "cost_per_test": "21.407199", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 4.506, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-7/53558929-7b1c-4167-9ded-006531b6a278.json b/data/vals_ai/anthropic/claude-opus-4-7/53558929-7b1c-4167-9ded-006531b6a278.json new file mode 100644 index 0000000000000000000000000000000000000000..5c2f9e662399faa37cfbb7a279a66595be08f16c --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-7/53558929-7b1c-4167-9ded-006531b6a278.json @@ -0,0 +1,381 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/anthropic_claude-opus-4-7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-7", + "id": "anthropic/claude-opus-4-7", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-7", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 64.286, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "64.286", + "raw_stderr": "7.483", + "latency": "812.181", + "cost_per_test": "5.23707", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 7.483, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 78.927, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "78.927", + "raw_stderr": "2.529", + "latency": "494.019", + "cost_per_test": "2.715322", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.529, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 90.206, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "90.206", + "raw_stderr": "2.14", + "latency": "280.58", + "cost_per_test": "1.375301", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.14, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66.667", + "raw_stderr": "33.333", + "latency": "1170.395", + "cost_per_test": "5.416409", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:overall:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 82.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "82", + "raw_stderr": "1.72", + "latency": "441.988", + "cost_per_test": "2.423427", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.72, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-7/62409804-c454-4a9f-90f0-4bc4ac4aec3a.json b/data/vals_ai/anthropic/claude-opus-4-7/62409804-c454-4a9f-90f0-4bc4ac4aec3a.json new file mode 100644 index 0000000000000000000000000000000000000000..6bb29966c0348f092cc504eefd82adabe047a520 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-7/62409804-c454-4a9f-90f0-4bc4ac4aec3a.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/anthropic_claude-opus-4-7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-7", + "id": "anthropic/claude-opus-4-7", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-7", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 78.696, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "78.696", + "raw_stderr": "1.154", + "latency": "19.793", + "cost_per_test": "0.093695", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.154, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 70.27, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "70.27", + "raw_stderr": "0.896", + "latency": "16.101", + "cost_per_test": "0.084417", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.896, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 61.844, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.844", + "raw_stderr": "1.37", + "latency": "12.409", + "cost_per_test": "0.07514", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.37, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-7/6833c274-3b8e-4d32-81f9-23e3121aab08.json b/data/vals_ai/anthropic/claude-opus-4-7/6833c274-3b8e-4d32-81f9-23e3121aab08.json new file mode 100644 index 0000000000000000000000000000000000000000..9e374ea7b25162a369fdaca8085043e36713b78c --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-7/6833c274-3b8e-4d32-81f9-23e3121aab08.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/anthropic_claude-opus-4-7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-7", + "id": "anthropic/claude-opus-4-7", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-7", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 91.667, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "91.667", + "raw_stderr": "8.333", + "latency": "18169.067", + "cost_per_test": "43.990177", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 8.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 40.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "40", + "raw_stderr": "5.092", + "latency": "2422.542", + "cost_per_test": "5.865357", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.092, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 82.424, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "82.424", + "raw_stderr": "0.606", + "latency": "1321.387", + "cost_per_test": "3.199286", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.606, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 68.539, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "68.539", + "raw_stderr": "1.297", + "latency": "816.587", + "cost_per_test": "1.977087", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.297, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-7/98a6b389-e01b-4696-81db-9d537d09658a.json b/data/vals_ai/anthropic/claude-opus-4-7/98a6b389-e01b-4696-81db-9d537d09658a.json new file mode 100644 index 0000000000000000000000000000000000000000..4baf1e1d1958ed600b84bd5626163b5e264eeaeb --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-7/98a6b389-e01b-4696-81db-9d537d09658a.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/anthropic_claude-opus-4-7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-7", + "id": "anthropic/claude-opus-4-7", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-7", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 54.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "54", + "raw_stderr": "4.984", + "latency": "679.708", + "cost_per_test": "1.807786", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 4.984, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-7/a5011987-6075-4583-9b90-5c21bf2e0123.json b/data/vals_ai/anthropic/claude-opus-4-7/a5011987-6075-4583-9b90-5c21bf2e0123.json new file mode 100644 index 0000000000000000000000000000000000000000..df770675391f2b2be19ee0cffe770ef3cacc068c --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-7/a5011987-6075-4583-9b90-5c21bf2e0123.json @@ -0,0 +1,418 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/anthropic_claude-opus-4-7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-7", + "id": "anthropic/claude-opus-4-7", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-7", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 66.084, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "66.084", + "raw_stderr": "1.616", + "latency": "18.221", + "cost_per_test": "0.816017", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.616, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 51.45, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "51.45", + "raw_stderr": "0.251", + "latency": "322.275", + "cost_per_test": "3.638017", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.251, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:overall:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 66.099, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "66.099", + "raw_stderr": "1.268", + "latency": "761.527", + "cost_per_test": "6.167573", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.268, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 82.353, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "82.353", + "raw_stderr": "3.775", + "latency": "508.652", + "cost_per_test": "2.999546" + }, + "uncertainty": { + "standard_error": { + "value": 3.775, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 68.539, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "68.539", + "raw_stderr": "1.297", + "latency": "816.587", + "cost_per_test": "1.977087", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.297, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 77.701, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "77.701", + "raw_stderr": "5.336", + "latency": "2141.902", + "cost_per_test": "21.407199", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.336, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-7/ad4c8465-4301-4cbe-8b98-395f7f692b5a.json b/data/vals_ai/anthropic/claude-opus-4-7/ad4c8465-4301-4cbe-8b98-395f7f692b5a.json new file mode 100644 index 0000000000000000000000000000000000000000..64db9ee98bb31e56abe97588741c448b660b0c50 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-7/ad4c8465-4301-4cbe-8b98-395f7f692b5a.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/anthropic_claude-opus-4-7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-7", + "id": "anthropic/claude-opus-4-7", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-7", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 64.336, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.336", + "raw_stderr": "1.635", + "latency": "10.524", + "cost_per_test": "0.03284", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.635, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 67.832, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.832", + "raw_stderr": "1.595", + "latency": "24.506", + "cost_per_test": "1.760496", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.595, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 66.084, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.084", + "raw_stderr": "0.933", + "latency": "17.75", + "cost_per_test": "0.816017", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.933, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 66.084, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.084", + "raw_stderr": "1.616", + "latency": "18.221", + "cost_per_test": "0.654716", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.616, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-7/b3cbf0b6-103c-46e0-ade8-d7cbe4d7165c.json b/data/vals_ai/anthropic/claude-opus-4-7/b3cbf0b6-103c-46e0-ade8-d7cbe4d7165c.json new file mode 100644 index 0000000000000000000000000000000000000000..80b28b0848248ccf4ef3d4efb49ebf5ffd03bbe4 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-7/b3cbf0b6-103c-46e0-ade8-d7cbe4d7165c.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/code-migration/anthropic_claude-opus-4-7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Code Migration", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-7", + "id": "anthropic/claude-opus-4-7", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-7", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "code-migration:cli:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.code-migration.cli", + "source_data": { + "dataset_name": "Code Migration - CLI", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cli", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (CLI).", + "metric_id": "vals_ai.code-migration.cli.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 35.171, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cli", + "task_name": "CLI", + "dataset_type": "private", + "industry": "coding", + "raw_score": "35.171", + "raw_stderr": "2.394", + "latency": "4613.696", + "cost_per_test": "40.880803", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.394, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "code-migration:cobol:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.code-migration.cobol", + "source_data": { + "dataset_name": "Code Migration - COBOL", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cobol", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (COBOL).", + "metric_id": "vals_ai.code-migration.cobol.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 70.0, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cobol", + "task_name": "COBOL", + "dataset_type": "private", + "industry": "coding", + "raw_score": "70", + "raw_stderr": "15.275", + "latency": "3466.953", + "cost_per_test": "22.908316", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 15.275, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "code-migration:code_quality:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.code-migration.code_quality", + "source_data": { + "dataset_name": "Code Migration - Code Quality", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "code_quality", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Code Quality).", + "metric_id": "vals_ai.code-migration.code_quality.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 4.588, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "code_quality", + "task_name": "Code Quality", + "dataset_type": "private", + "industry": "coding", + "raw_score": "4.588", + "raw_stderr": "0.034", + "latency": "4327.01", + "cost_per_test": "36.387681", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.034, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "code-migration:overall:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.code-migration.overall", + "source_data": { + "dataset_name": "Code Migration - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Overall).", + "metric_id": "vals_ai.code-migration.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 43.878, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "43.878", + "raw_stderr": "4.22", + "latency": "4327.01", + "cost_per_test": "36.387681", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 4.22, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-7/b86c1fa6-745a-4f9c-b4af-2ebf2d8864b2.json b/data/vals_ai/anthropic/claude-opus-4-7/b86c1fa6-745a-4f9c-b4af-2ebf2d8864b2.json new file mode 100644 index 0000000000000000000000000000000000000000..516c30b5e92515a6b5a86d3607e013693c04b7dc --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-7/b86c1fa6-745a-4f9c-b4af-2ebf2d8864b2.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/anthropic_claude-opus-4-7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-7", + "id": "anthropic/claude-opus-4-7", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-7", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 53.167, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "53.167", + "raw_stderr": "14.854", + "latency": "2569.514", + "cost_per_test": "9.010059", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 14.854, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 41.0, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "41", + "raw_stderr": "9.077", + "latency": "3594.296", + "cost_per_test": "11.963848", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 9.077, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "ioi:overall:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 47.084, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "47.084", + "raw_stderr": "8.704", + "latency": "3081.905", + "cost_per_test": "10.486954", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 8.704, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-7/bb3303ea-2f3d-4bef-b5ed-7baebf32c965.json b/data/vals_ai/anthropic/claude-opus-4-7/bb3303ea-2f3d-4bef-b5ed-7baebf32c965.json new file mode 100644 index 0000000000000000000000000000000000000000..552fc235f7dde7487b269c6b4d00d802b7aeae9e --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-7/bb3303ea-2f3d-4bef-b5ed-7baebf32c965.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/anthropic_claude-opus-4-7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-7", + "id": "anthropic/claude-opus-4-7", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-7", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 99.068, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "99.068", + "raw_stderr": "0.535", + "latency": "171.189", + "cost_per_test": "0.032636", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "lcb:hard:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 66.857, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66.857", + "raw_stderr": "2.516", + "latency": "787.584", + "cost_per_test": "1.012836", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.516, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "lcb:medium:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 89.295, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "89.295", + "raw_stderr": "1.58", + "latency": "426.957", + "cost_per_test": "0.00629", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.58, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "lcb:overall:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 85.073, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "85.073", + "raw_stderr": "1.006", + "latency": "461.91", + "cost_per_test": "0.350587", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.006, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-7/bbf14f6a-f080-41e7-8957-1c55d9175552.json b/data/vals_ai/anthropic/claude-opus-4-7/bbf14f6a-f080-41e7-8957-1c55d9175552.json new file mode 100644 index 0000000000000000000000000000000000000000..185c190504c5d320e6e85d2c848a23609186058f --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-7/bbf14f6a-f080-41e7-8957-1c55d9175552.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/anthropic_claude-opus-4-7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-7", + "id": "anthropic/claude-opus-4-7", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-7", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 82.953, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "82.953", + "raw_stderr": "1.977", + "latency": "67.496", + "cost_per_test": "0.177841", + "temperature": "1", + "max_output_tokens": "30000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.977, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-7/c272d30d-af6b-445e-b8d4-4a5f4fbafdaa.json b/data/vals_ai/anthropic/claude-opus-4-7/c272d30d-af6b-445e-b8d4-4a5f4fbafdaa.json new file mode 100644 index 0000000000000000000000000000000000000000..e12e71364c2486be832e46d2d012af832d837c3b --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-7/c272d30d-af6b-445e-b8d4-4a5f4fbafdaa.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/programbench/anthropic_claude-opus-4-7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProgramBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-7", + "id": "anthropic/claude-opus-4-7", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-7", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "programbench:almost:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.programbench.almost", + "source_data": { + "dataset_name": "ProgramBench - Almost Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "almost", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Almost Resolved).", + "metric_id": "vals_ai.programbench.almost.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 9.5, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "almost", + "task_name": "Almost Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "9.5", + "raw_stderr": "2.079", + "latency": "3246.698", + "cost_per_test": "30.939088", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.079, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "programbench:overall:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.programbench.overall", + "source_data": { + "dataset_name": "ProgramBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Overall).", + "metric_id": "vals_ai.programbench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "3246.698", + "cost_per_test": "30.939088", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "programbench:partial:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.programbench.partial", + "source_data": { + "dataset_name": "ProgramBench - Raw Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "partial", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Raw Pass Rate).", + "metric_id": "vals_ai.programbench.partial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 64.225, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "partial", + "task_name": "Raw Pass Rate", + "dataset_type": "public", + "industry": "coding", + "raw_score": "64.225", + "raw_stderr": "1.894", + "latency": "3246.698", + "cost_per_test": "30.939088", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.894, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "programbench:strict:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.programbench.strict", + "source_data": { + "dataset_name": "ProgramBench - Fully Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "strict", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Fully Resolved).", + "metric_id": "vals_ai.programbench.strict.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "strict", + "task_name": "Fully Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "3246.698", + "cost_per_test": "30.939088", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-7/c3383c9f-d357-4a57-8b9d-6023f9a95fa6.json b/data/vals_ai/anthropic/claude-opus-4-7/c3383c9f-d357-4a57-8b9d-6023f9a95fa6.json new file mode 100644 index 0000000000000000000000000000000000000000..8d8aef4b7e3a896d25a80a3500cd6b807120e700 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-7/c3383c9f-d357-4a57-8b9d-6023f9a95fa6.json @@ -0,0 +1,195 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/cyber/anthropic_claude-opus-4-7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CyberBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "beta", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-7", + "id": "anthropic/claude-opus-4-7", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-7", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "cyber:overall:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.cyber.overall", + "source_data": { + "dataset_name": "CyberBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (Overall).", + "metric_id": "vals_ai.cyber.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 52.141, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "beta", + "raw_score": "52.141", + "raw_stderr": "5.614", + "latency": "752.381", + "cost_per_test": "4.089642", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.614, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "cyber:patch:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.cyber.patch", + "source_data": { + "dataset_name": "CyberBench - Patch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "patch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (Patch).", + "metric_id": "vals_ai.cyber.patch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 77.966, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "patch", + "task_name": "Patch", + "dataset_type": "private", + "industry": "beta", + "raw_score": "77.966", + "raw_stderr": "5.396", + "latency": "1028.162", + "cost_per_test": "2.770533", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.396, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "cyber:poc:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.cyber.poc", + "source_data": { + "dataset_name": "CyberBench - PoC", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "poc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (PoC).", + "metric_id": "vals_ai.cyber.poc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 26.316, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "poc", + "task_name": "PoC", + "dataset_type": "private", + "industry": "beta", + "raw_score": "26.316", + "raw_stderr": "5.833", + "latency": "476.601", + "cost_per_test": "5.408751", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.833, + "method": "vals_reported" + } + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-7/e008115a-7aff-45a0-9d77-5f8a8a9e4315.json b/data/vals_ai/anthropic/claude-opus-4-7/e008115a-7aff-45a0-9d77-5f8a8a9e4315.json new file mode 100644 index 0000000000000000000000000000000000000000..481362c0d74e6b9528a1b7f5d3149869eb0304a2 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-7/e008115a-7aff-45a0-9d77-5f8a8a9e4315.json @@ -0,0 +1,105 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/anthropic_claude-opus-4-7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-7", + "id": "anthropic/claude-opus-4-7", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-7", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 85.549, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.549", + "raw_stderr": "0.845", + "latency": "70.801", + "cost_per_test": "0.170461", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.845, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-7/e7e68dd6-0f5a-4195-bd7e-6200c32c634d.json b/data/vals_ai/anthropic/claude-opus-4-7/e7e68dd6-0f5a-4195-bd7e-6200c32c634d.json new file mode 100644 index 0000000000000000000000000000000000000000..42707be1547221a3dbdc14e54b3758da46ccbe90 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-7/e7e68dd6-0f5a-4195-bd7e-6200c32c634d.json @@ -0,0 +1,550 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_multimodal_index/anthropic_claude-opus-4-7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Multimodal Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-7", + "id": "anthropic/claude-opus-4-7", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-7", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_multimodal_index:corp_fin_v2:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.vals_multimodal_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Multimodal Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (CorpFin v2).", + "metric_id": "vals_ai.vals_multimodal_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 66.084, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "66.084", + "raw_stderr": "1.616", + "latency": "18.221", + "cost_per_test": "0.816017", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.616, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:finance_agent:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.vals_multimodal_index.finance_agent", + "source_data": { + "dataset_name": "Vals Multimodal Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_multimodal_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 51.45, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "51.45", + "raw_stderr": "0.251", + "latency": "322.275", + "cost_per_test": "3.638017", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.251, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:mortgage_tax:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.vals_multimodal_index.mortgage_tax", + "source_data": { + "dataset_name": "Vals Multimodal Index - Mortgage Tax", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "mortgage_tax", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Mortgage Tax).", + "metric_id": "vals_ai.vals_multimodal_index.mortgage_tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 70.27, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "mortgage_tax", + "task_name": "Mortgage Tax", + "dataset_type": "private", + "industry": "index", + "raw_score": "70.27", + "raw_stderr": "0.896", + "latency": "16.101", + "cost_per_test": "0.084417", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.896, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:overall:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.vals_multimodal_index.overall", + "source_data": { + "dataset_name": "Vals Multimodal Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Overall).", + "metric_id": "vals_ai.vals_multimodal_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 67.361, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "67.361", + "raw_stderr": "1.162", + "latency": "564.055", + "cost_per_test": "4.475671", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.162, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:sage:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.vals_multimodal_index.sage", + "source_data": { + "dataset_name": "Vals Multimodal Index - SAGE", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "sage", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SAGE).", + "metric_id": "vals_ai.vals_multimodal_index.sage.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 56.103, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "sage", + "task_name": "SAGE", + "dataset_type": "private", + "industry": "index", + "raw_score": "56.103", + "raw_stderr": "3.368", + "latency": "124.647", + "cost_per_test": "0.327557", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.368, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:swebench:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.vals_multimodal_index.swebench", + "source_data": { + "dataset_name": "Vals Multimodal Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SWE-bench).", + "metric_id": "vals_ai.vals_multimodal_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 82.353, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "82.353", + "raw_stderr": "3.775", + "latency": "508.652", + "cost_per_test": "2.999546" + }, + "uncertainty": { + "standard_error": { + "value": 3.775, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:terminal_bench_2_1:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.vals_multimodal_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Multimodal Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_multimodal_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 68.539, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "68.539", + "raw_stderr": "1.297", + "latency": "816.587", + "cost_per_test": "1.977087", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "high", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.297, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:vibe_code_bench:anthropic/claude-opus-4-7:score", + "evaluation_name": "vals_ai.vals_multimodal_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Multimodal Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_multimodal_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 77.701, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "77.701", + "raw_stderr": "5.336", + "latency": "2141.902", + "cost_per_test": "21.407199", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.336, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-8-claude-code/0acd22d7-67c7-49a9-ace7-e739697e2455.json b/data/vals_ai/anthropic/claude-opus-4-8-claude-code/0acd22d7-67c7-49a9-ace7-e739697e2455.json new file mode 100644 index 0000000000000000000000000000000000000000..2170b0f68639cc1e75ce5bee4862b0645e379125 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-8-claude-code/0acd22d7-67c7-49a9-ace7-e739697e2455.json @@ -0,0 +1,260 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/anthropic_claude-opus-4-8-claude-code/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-8-claude-code", + "id": "anthropic/claude-opus-4-8-claude-code", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-8-claude-code", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:anthropic/claude-opus-4-8-claude-code:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 100.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "100", + "raw_stderr": "0", + "latency": "8009.262", + "cost_per_test": "21.540183", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:anthropic/claude-opus-4-8-claude-code:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 63.333, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "63.333", + "raw_stderr": "8.949", + "latency": "1067.902", + "cost_per_test": "2.872024", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 8.949, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:anthropic/claude-opus-4-8-claude-code:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 70.909, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "70.909", + "raw_stderr": "6.181", + "latency": "582.492", + "cost_per_test": "1.566559", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.181, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:anthropic/claude-opus-4-8-claude-code:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 69.663, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "69.663", + "raw_stderr": "4.901", + "latency": "359.967", + "cost_per_test": "0.968098", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 4.901, + "method": "vals_reported" + } + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-8-claude-code/505f6a64-93de-4ee0-ab1e-620bf109831e.json b/data/vals_ai/anthropic/claude-opus-4-8-claude-code/505f6a64-93de-4ee0-ab1e-620bf109831e.json new file mode 100644 index 0000000000000000000000000000000000000000..53de095fad93ec2a6f0128227ba561332d59d573 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-8-claude-code/505f6a64-93de-4ee0-ab1e-620bf109831e.json @@ -0,0 +1,316 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/anthropic_claude-opus-4-8-claude-code/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-8-claude-code", + "id": "anthropic/claude-opus-4-8-claude-code", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-8-claude-code", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:anthropic/claude-opus-4-8-claude-code:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 71.429, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "71.429", + "raw_stderr": "7.055", + "latency": "296.765", + "cost_per_test": "1.488649", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 7.055, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:anthropic/claude-opus-4-8-claude-code:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 84.291, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "84.291", + "raw_stderr": "2.257", + "latency": "150.856", + "cost_per_test": "0.741644", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.257, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:anthropic/claude-opus-4-8-claude-code:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 91.237, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "91.237", + "raw_stderr": "2.035", + "latency": "71.328", + "cost_per_test": "0.375902", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.035, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:anthropic/claude-opus-4-8-claude-code:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66.667", + "raw_stderr": "33.333", + "latency": "674.074", + "cost_per_test": "2.402732", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "swebench:overall:anthropic/claude-opus-4-8-claude-code:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 85.8, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "85.8", + "raw_stderr": "1.563", + "latency": "135.395", + "cost_per_test": "0.672451", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.563, + "method": "vals_reported" + } + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-8-claude-code/869ee621-04fc-4aff-a5b5-5508bc42d313.json b/data/vals_ai/anthropic/claude-opus-4-8-claude-code/869ee621-04fc-4aff-a5b5-5508bc42d313.json new file mode 100644 index 0000000000000000000000000000000000000000..c02ba226824387bf7fe3be5b9ad4579b7266f70a --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-8-claude-code/869ee621-04fc-4aff-a5b5-5508bc42d313.json @@ -0,0 +1,89 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/anthropic_claude-opus-4-8-claude-code/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-8-claude-code", + "id": "anthropic/claude-opus-4-8-claude-code", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-8-claude-code", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:anthropic/claude-opus-4-8-claude-code:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 77.485, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "77.485", + "raw_stderr": "3.737", + "latency": "1142.14", + "cost_per_test": "6.8557", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.737, + "method": "vals_reported" + } + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-8/0102a4bf-c13a-45fb-8d63-7a525b6d1093.json b/data/vals_ai/anthropic/claude-opus-4-8/0102a4bf-c13a-45fb-8d63-7a525b6d1093.json new file mode 100644 index 0000000000000000000000000000000000000000..a870f49b5a80f042eb2570083f1d939ca4a5cd0e --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-8/0102a4bf-c13a-45fb-8d63-7a525b6d1093.json @@ -0,0 +1,550 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_multimodal_index/anthropic_claude-opus-4-8/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Multimodal Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-8", + "id": "anthropic/claude-opus-4-8", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-8", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_multimodal_index:corp_fin_v2:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.vals_multimodal_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Multimodal Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (CorpFin v2).", + "metric_id": "vals_ai.vals_multimodal_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 65.967, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "65.967", + "raw_stderr": "1.618", + "latency": "20.38", + "cost_per_test": "0.855534", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.618, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:finance_agent:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.vals_multimodal_index.finance_agent", + "source_data": { + "dataset_name": "Vals Multimodal Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_multimodal_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 53.949, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "53.949", + "raw_stderr": "1.764", + "latency": "522.678", + "cost_per_test": "4.218538", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.764, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:mortgage_tax:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.vals_multimodal_index.mortgage_tax", + "source_data": { + "dataset_name": "Vals Multimodal Index - Mortgage Tax", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "mortgage_tax", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Mortgage Tax).", + "metric_id": "vals_ai.vals_multimodal_index.mortgage_tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 69.912, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "mortgage_tax", + "task_name": "Mortgage Tax", + "dataset_type": "private", + "industry": "index", + "raw_score": "69.912", + "raw_stderr": "0.893", + "latency": "20.814", + "cost_per_test": "0.105627", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.893, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:overall:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.vals_multimodal_index.overall", + "source_data": { + "dataset_name": "Vals Multimodal Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Overall).", + "metric_id": "vals_ai.vals_multimodal_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 70.889, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "70.889", + "raw_stderr": "0.744", + "latency": "979.847", + "cost_per_test": "2.165957", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.744, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:sage:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.vals_multimodal_index.sage", + "source_data": { + "dataset_name": "Vals Multimodal Index - SAGE", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "sage", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SAGE).", + "metric_id": "vals_ai.vals_multimodal_index.sage.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 54.788, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "sage", + "task_name": "SAGE", + "dataset_type": "private", + "industry": "index", + "raw_score": "54.788", + "raw_stderr": "3.347", + "latency": "185.594", + "cost_per_test": "0.47466", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.347, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:swebench:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.vals_multimodal_index.swebench", + "source_data": { + "dataset_name": "Vals Multimodal Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SWE-bench).", + "metric_id": "vals_ai.vals_multimodal_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 89.216, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "89.216", + "raw_stderr": "3.071", + "latency": "630.305", + "cost_per_test": "1.923224" + }, + "uncertainty": { + "standard_error": { + "value": 3.071, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:terminal_bench_2_1:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.vals_multimodal_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Multimodal Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_multimodal_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 71.91, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "71.91", + "raw_stderr": "0.649", + "latency": "929.902", + "cost_per_test": "2.409772", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.649, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:vibe_code_bench:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.vals_multimodal_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Multimodal Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_multimodal_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 89.869, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "89.869", + "raw_stderr": "2.244", + "latency": "4549.256", + "cost_per_test": "5.094422", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.244, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-8/0334d187-0795-462a-9e71-6a50800f072c.json b/data/vals_ai/anthropic/claude-opus-4-8/0334d187-0795-462a-9e71-6a50800f072c.json new file mode 100644 index 0000000000000000000000000000000000000000..b4f40d2bd1acccf7bd2ffb3116770165ea6d9689 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-8/0334d187-0795-462a-9e71-6a50800f072c.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/anthropic_claude-opus-4-8/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-8", + "id": "anthropic/claude-opus-4-8", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-8", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 99.068, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "99.068", + "raw_stderr": "0.535", + "latency": "12.108", + "cost_per_test": "0.02112", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "lcb:hard:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 74.571, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "74.571", + "raw_stderr": "2.328", + "latency": "379.558", + "cost_per_test": "0.798518", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.328, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "lcb:medium:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 89.817, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "89.817", + "raw_stderr": "1.545", + "latency": "87.119", + "cost_per_test": "0.162481", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.545, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "lcb:overall:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 87.819, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "87.819", + "raw_stderr": "0.948", + "latency": "159.595", + "cost_per_test": "0.327373", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.948, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-8/071c622f-559a-459f-9ac3-f87c3d5d489b.json b/data/vals_ai/anthropic/claude-opus-4-8/071c622f-559a-459f-9ac3-f87c3d5d489b.json new file mode 100644 index 0000000000000000000000000000000000000000..983b1a6475fccbf0037d5d6491dced1119f812a7 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-8/071c622f-559a-459f-9ac3-f87c3d5d489b.json @@ -0,0 +1,418 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/anthropic_claude-opus-4-8/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-8", + "id": "anthropic/claude-opus-4-8", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-8", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 65.967, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "65.967", + "raw_stderr": "1.618", + "latency": "20.38", + "cost_per_test": "0.855534", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.618, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 53.949, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "53.949", + "raw_stderr": "1.764", + "latency": "522.678", + "cost_per_test": "4.218538", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.764, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_index:overall:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 70.358, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "70.358", + "raw_stderr": "0.902", + "latency": "1330.504", + "cost_per_test": "2.900298", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.902, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 89.216, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "89.216", + "raw_stderr": "3.071", + "latency": "630.305", + "cost_per_test": "1.923224" + }, + "uncertainty": { + "standard_error": { + "value": 3.071, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 71.91, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "71.91", + "raw_stderr": "0.649", + "latency": "929.902", + "cost_per_test": "2.409772", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.649, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 89.869, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "89.869", + "raw_stderr": "2.244", + "latency": "4549.256", + "cost_per_test": "5.094422", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.244, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-8/10a702e9-149d-4dd3-aa83-cf1db165eefb.json b/data/vals_ai/anthropic/claude-opus-4-8/10a702e9-149d-4dd3-aa83-cf1db165eefb.json new file mode 100644 index 0000000000000000000000000000000000000000..5c23520047a2fe6e583cda934274e62762d17027 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-8/10a702e9-149d-4dd3-aa83-cf1db165eefb.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/anthropic_claude-opus-4-8/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-8", + "id": "anthropic/claude-opus-4-8", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-8", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 91.919, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.919", + "raw_stderr": "2.277", + "latency": "131.379", + "cost_per_test": "0.3381", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.277, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 92.929, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.929", + "raw_stderr": "3.25", + "latency": "136.373", + "cost_per_test": "0.340421", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.25, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 92.424, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.424", + "raw_stderr": "1.984", + "latency": "133.876", + "cost_per_test": "0.339261", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.984, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-8/12bd9eb2-bdff-43ce-b038-0b5ca31686b4.json b/data/vals_ai/anthropic/claude-opus-4-8/12bd9eb2-bdff-43ce-b038-0b5ca31686b4.json new file mode 100644 index 0000000000000000000000000000000000000000..2263c3709093c381e219c2359309c64da2fd42f4 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-8/12bd9eb2-bdff-43ce-b038-0b5ca31686b4.json @@ -0,0 +1,762 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/anthropic_claude-opus-4-8/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-8", + "id": "anthropic/claude-opus-4-8", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-8", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 48.387, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "48.387", + "raw_stderr": "2.53", + "latency": "705.323", + "cost_per_test": "3.790165", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.53, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 43.517, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "43.517", + "raw_stderr": "0.303", + "latency": "547.397", + "cost_per_test": "4.218538", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.303, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 44.853, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "44.853", + "raw_stderr": "1.122", + "latency": "1048.58", + "cost_per_test": "6.341675", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.122, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 65.706, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.706", + "raw_stderr": "1.124", + "latency": "798.453", + "cost_per_test": "4.989067", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.124, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 72.07, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "72.07", + "raw_stderr": "1.3", + "latency": "444.246", + "cost_per_test": "1.386113", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.3, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 23.347, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "23.347", + "raw_stderr": "1.029", + "latency": "669.744", + "cost_per_test": "4.205596", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.029, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 66.414, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.414", + "raw_stderr": "2.134", + "latency": "842.804", + "cost_per_test": "5.273958", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.134, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 71.636, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.636", + "raw_stderr": "0.975", + "latency": "666.365", + "cost_per_test": "4.002255", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.975, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 67.219, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.219", + "raw_stderr": "0.47", + "latency": "414.757", + "cost_per_test": "2.017384", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.47, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:overall:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 53.918, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "53.918", + "raw_stderr": "0.159", + "latency": "547.397", + "cost_per_test": "4.218538", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.159, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 25.627, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "25.627", + "raw_stderr": "3.199", + "latency": "1099.702", + "cost_per_test": "6.192772", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.199, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-8/1f681401-9cb4-4520-af54-ef0a5190d00e.json b/data/vals_ai/anthropic/claude-opus-4-8/1f681401-9cb4-4520-af54-ef0a5190d00e.json new file mode 100644 index 0000000000000000000000000000000000000000..cfaaeb403bd1db480f8bd921c405ac7e460ba442 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-8/1f681401-9cb4-4520-af54-ef0a5190d00e.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/anthropic_claude-opus-4-8/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-8", + "id": "anthropic/claude-opus-4-8", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-8", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 82.725, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "82.725", + "raw_stderr": "3.081", + "latency": "4549.256", + "cost_per_test": "5.094422", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.081, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-8/30f03342-021e-4402-bf12-e85a055df239.json b/data/vals_ai/anthropic/claude-opus-4-8/30f03342-021e-4402-bf12-e85a055df239.json new file mode 100644 index 0000000000000000000000000000000000000000..23a306bbcaf95352ada564f039a1caed12a1c2b4 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-8/30f03342-021e-4402-bf12-e85a055df239.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/anthropic_claude-opus-4-8/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-8", + "id": "anthropic/claude-opus-4-8", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-8", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 44.444, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "44.444", + "raw_stderr": "6.24", + "latency": "203.742", + "cost_per_test": "0.494441", + "temperature": "1", + "max_output_tokens": "30000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.24, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "sage:automata:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 41.666, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "41.666", + "raw_stderr": "6.311", + "latency": "236.054", + "cost_per_test": "0.61324", + "temperature": "1", + "max_output_tokens": "30000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.311, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 78.253, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "78.253", + "raw_stderr": "4.694", + "latency": "116.985", + "cost_per_test": "0.40882", + "temperature": "1", + "max_output_tokens": "30000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 4.694, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "sage:overall:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 54.788, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "54.788", + "raw_stderr": "3.347", + "latency": "185.594", + "cost_per_test": "0.5055", + "temperature": "1", + "max_output_tokens": "30000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.347, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-8/32a777ca-e26d-4ec8-aa74-4fb656d39b4d.json b/data/vals_ai/anthropic/claude-opus-4-8/32a777ca-e26d-4ec8-aa74-4fb656d39b4d.json new file mode 100644 index 0000000000000000000000000000000000000000..0c930d195306516b24d0dceed91d8a97f174e1d7 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-8/32a777ca-e26d-4ec8-aa74-4fb656d39b4d.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/anthropic_claude-opus-4-8/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-8", + "id": "anthropic/claude-opus-4-8", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-8", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 64.513, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.513", + "raw_stderr": "1.368", + "latency": "70.33", + "cost_per_test": "0.159139", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.368, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 75.634, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "75.634", + "raw_stderr": "0.838", + "latency": "70.33", + "cost_per_test": "0.159139", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.838, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 86.754, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "86.754", + "raw_stderr": "0.969", + "latency": "70.33", + "cost_per_test": "0.159139", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.969, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-8/3b61c666-4882-46d4-aa23-79dd0c660693.json b/data/vals_ai/anthropic/claude-opus-4-8/3b61c666-4882-46d4-aa23-79dd0c660693.json new file mode 100644 index 0000000000000000000000000000000000000000..3b8850e4cf22f21b221e9a9d549551c30d2289e3 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-8/3b61c666-4882-46d4-aa23-79dd0c660693.json @@ -0,0 +1,696 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/emb/anthropic_claude-opus-4-8/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - EMB", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-8", + "id": "anthropic/claude-opus-4-8", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-8", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "emb:comps:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.emb.comps", + "source_data": { + "dataset_name": "EMB - Comps", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "comps", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Comps).", + "metric_id": "vals_ai.emb.comps.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 72.797, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "comps", + "task_name": "Comps", + "dataset_type": "private", + "industry": "finance", + "raw_score": "72.797", + "raw_stderr": "4.198", + "latency": "2442.712", + "cost_per_test": "12.886468", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 4.198, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:dataroom:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.emb.dataroom", + "source_data": { + "dataset_name": "EMB - Dataroom", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dataroom", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Dataroom).", + "metric_id": "vals_ai.emb.dataroom.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 85.948, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dataroom", + "task_name": "Dataroom", + "dataset_type": "private", + "industry": "finance", + "raw_score": "85.948", + "raw_stderr": "3.719", + "latency": "2182.513", + "cost_per_test": "8.994203", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.719, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:dcf:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.emb.dcf", + "source_data": { + "dataset_name": "EMB - DCF", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dcf", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (DCF).", + "metric_id": "vals_ai.emb.dcf.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 66.334, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dcf", + "task_name": "DCF", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.334", + "raw_stderr": "7.339", + "latency": "2796.508", + "cost_per_test": "12.058156", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 7.339, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:lbo:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.emb.lbo", + "source_data": { + "dataset_name": "EMB - LBO", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "lbo", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (LBO).", + "metric_id": "vals_ai.emb.lbo.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 54.384, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "lbo", + "task_name": "LBO", + "dataset_type": "private", + "industry": "finance", + "raw_score": "54.384", + "raw_stderr": "5.352", + "latency": "3411.868", + "cost_per_test": "17.127734", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.352, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:ma:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.emb.ma", + "source_data": { + "dataset_name": "EMB - M&A", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "ma", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (M&A).", + "metric_id": "vals_ai.emb.ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 80.091, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "ma", + "task_name": "M&A", + "dataset_type": "private", + "industry": "finance", + "raw_score": "80.091", + "raw_stderr": "3.258", + "latency": "2301.312", + "cost_per_test": "9.449638", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.258, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:misc:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.emb.misc", + "source_data": { + "dataset_name": "EMB - Misc", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "misc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Misc).", + "metric_id": "vals_ai.emb.misc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 59.355, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "misc", + "task_name": "Misc", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.355", + "raw_stderr": "8.87", + "latency": "2930.208", + "cost_per_test": "13.123916", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 8.87, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:operating:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.emb.operating", + "source_data": { + "dataset_name": "EMB - Operating", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "operating", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Operating).", + "metric_id": "vals_ai.emb.operating.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 66.699, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "operating", + "task_name": "Operating", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.699", + "raw_stderr": "6.296", + "latency": "2376.906", + "cost_per_test": "10.392758", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.296, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:overall:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.emb.overall", + "source_data": { + "dataset_name": "EMB - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Overall).", + "metric_id": "vals_ai.emb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 69.373, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "69.373", + "raw_stderr": "2.606", + "latency": "2632.285", + "cost_per_test": "12.05844", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.606, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:scratch:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.emb.scratch", + "source_data": { + "dataset_name": "EMB - Scratch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "scratch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Scratch).", + "metric_id": "vals_ai.emb.scratch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 67.437, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "scratch", + "task_name": "Scratch", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.437", + "raw_stderr": "2.853", + "latency": "2538.573", + "cost_per_test": "11.138836", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.853, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:template:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.emb.template", + "source_data": { + "dataset_name": "EMB - Template", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "template", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Template).", + "metric_id": "vals_ai.emb.template.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 71.143, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "template", + "task_name": "Template", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.143", + "raw_stderr": "3.696", + "latency": "2724.158", + "cost_per_test": "12.960012", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.696, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-8/4cdf8984-4a08-4ccc-9417-27b179248d36.json b/data/vals_ai/anthropic/claude-opus-4-8/4cdf8984-4a08-4ccc-9417-27b179248d36.json new file mode 100644 index 0000000000000000000000000000000000000000..2070c41ea8e0f15b7d5ed92f9b300fec47917ea4 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-8/4cdf8984-4a08-4ccc-9417-27b179248d36.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/anthropic_claude-opus-4-8/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-8", + "id": "anthropic/claude-opus-4-8", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-8", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 64.336, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.336", + "raw_stderr": "1.635", + "latency": "14.109", + "cost_per_test": "0.039715", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.635, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 69.814, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "69.814", + "raw_stderr": "1.567", + "latency": "26.189", + "cost_per_test": "1.847531", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.567, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 66.706, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.706", + "raw_stderr": "0.928", + "latency": "20.226", + "cost_per_test": "0.855534", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.928, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 65.967, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.967", + "raw_stderr": "1.618", + "latency": "20.38", + "cost_per_test": "0.679358", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.618, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-8/5ddb0369-0bb8-4424-8ca5-42df02931a8d.json b/data/vals_ai/anthropic/claude-opus-4-8/5ddb0369-0bb8-4424-8ca5-42df02931a8d.json new file mode 100644 index 0000000000000000000000000000000000000000..6441af2fa1e4b61d7234132177bc6ab01aec2a29 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-8/5ddb0369-0bb8-4424-8ca5-42df02931a8d.json @@ -0,0 +1,105 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/anthropic_claude-opus-4-8/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-8", + "id": "anthropic/claude-opus-4-8", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-8", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 86.59, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.59", + "raw_stderr": "0.819", + "latency": "49.044", + "cost_per_test": "0.120663", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.819, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-8/656b027d-b2a2-42b6-9a58-9b3ff767cacd.json b/data/vals_ai/anthropic/claude-opus-4-8/656b027d-b2a2-42b6-9a58-9b3ff767cacd.json new file mode 100644 index 0000000000000000000000000000000000000000..3e6fa028cc349784379ae29cfa716c1955007a25 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-8/656b027d-b2a2-42b6-9a58-9b3ff767cacd.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/code-migration/anthropic_claude-opus-4-8/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Code Migration", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-8", + "id": "anthropic/claude-opus-4-8", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-8", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "code-migration:cli:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.code-migration.cli", + "source_data": { + "dataset_name": "Code Migration - CLI", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cli", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (CLI).", + "metric_id": "vals_ai.code-migration.cli.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 40.143, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cli", + "task_name": "CLI", + "dataset_type": "private", + "industry": "coding", + "raw_score": "40.143", + "raw_stderr": "2.431", + "latency": "4819.54", + "cost_per_test": "35.418678", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.431, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "code-migration:cobol:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.code-migration.cobol", + "source_data": { + "dataset_name": "Code Migration - COBOL", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cobol", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (COBOL).", + "metric_id": "vals_ai.code-migration.cobol.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 68.571, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cobol", + "task_name": "COBOL", + "dataset_type": "private", + "industry": "coding", + "raw_score": "68.571", + "raw_stderr": "15.028", + "latency": "3536.152", + "cost_per_test": "15.800701", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 15.028, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "code-migration:code_quality:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.code-migration.code_quality", + "source_data": { + "dataset_name": "Code Migration - Code Quality", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "code_quality", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Code Quality).", + "metric_id": "vals_ai.code-migration.code_quality.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 4.7, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "code_quality", + "task_name": "Code Quality", + "dataset_type": "private", + "industry": "coding", + "raw_score": "4.7", + "raw_stderr": "0.025", + "latency": "4498.693", + "cost_per_test": "30.514184", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.025, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "code-migration:overall:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.code-migration.overall", + "source_data": { + "dataset_name": "Code Migration - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Overall).", + "metric_id": "vals_ai.code-migration.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 47.25, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "47.25", + "raw_stderr": "4.176", + "latency": "4498.693", + "cost_per_test": "30.514184", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 4.176, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-8/661431f3-205d-478c-be92-27ec8debd0f7.json b/data/vals_ai/anthropic/claude-opus-4-8/661431f3-205d-478c-be92-27ec8debd0f7.json new file mode 100644 index 0000000000000000000000000000000000000000..9544ecb760d643d866d44ab716ba9fc1374283aa --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-8/661431f3-205d-478c-be92-27ec8debd0f7.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/public-benefits-bench/anthropic_claude-opus-4-8/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Public Benefits Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "social mobility", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-8", + "id": "anthropic/claude-opus-4-8", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-8", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "public-benefits-bench:overall:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.public-benefits-bench.overall", + "source_data": { + "dataset_name": "Public Benefits Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Public Benefits Bench v1.1 (Overall).", + "metric_id": "vals_ai.public-benefits-bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "score_details": { + "score": 68.133, + "details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "social mobility", + "raw_score": "68.133", + "raw_stderr": "1.212", + "latency": "897.428", + "cost_per_test": "1.889795", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.212, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-8/6d32d099-c0ab-4723-b503-3d2dc53fc16e.json b/data/vals_ai/anthropic/claude-opus-4-8/6d32d099-c0ab-4723-b503-3d2dc53fc16e.json new file mode 100644 index 0000000000000000000000000000000000000000..25fa784b80aade36e112d556ddfad6f41d91c682 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-8/6d32d099-c0ab-4723-b503-3d2dc53fc16e.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/anthropic_claude-opus-4-8/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-8", + "id": "anthropic/claude-opus-4-8", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-8", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 79.73, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "79.73", + "raw_stderr": "1.133", + "latency": "29.835", + "cost_per_test": "0.127846", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.133, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 69.912, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "69.912", + "raw_stderr": "0.893", + "latency": "20.814", + "cost_per_test": "0.105627", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.893, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 60.095, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.095", + "raw_stderr": "1.381", + "latency": "11.794", + "cost_per_test": "0.083407", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.381, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-8/78ddcd4a-f5f7-4281-8d7f-7f1be9f9482b.json b/data/vals_ai/anthropic/claude-opus-4-8/78ddcd4a-f5f7-4281-8d7f-7f1be9f9482b.json new file mode 100644 index 0000000000000000000000000000000000000000..77e043b6a4647ee820fafb559c5dab0f85f46067 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-8/78ddcd4a-f5f7-4281-8d7f-7f1be9f9482b.json @@ -0,0 +1,1224 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_research/anthropic_claude-opus-4-8/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Legal Research Bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-8", + "id": "anthropic/claude-opus-4-8", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-8", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_research:administrative_regulatory:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.legal_research.administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 43.939, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "43.939", + "raw_stderr": "6.156", + "latency": "636.543", + "cost_per_test": "2.731524", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.156, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:business_commercial:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.legal_research.business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 41.86, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "41.86", + "raw_stderr": "7.612", + "latency": "777.706", + "cost_per_test": "2.888428", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 7.612, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:civil_litigation:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.legal_research.civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 51.852, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "51.852", + "raw_stderr": "9.799", + "latency": "806.663", + "cost_per_test": "2.997652", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 9.799, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:constitutional_civil_rights:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.legal_research.constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 39.474, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "39.474", + "raw_stderr": "8.036", + "latency": "609.116", + "cost_per_test": "2.462184", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 8.036, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:criminal:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.legal_research.criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 45.614, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "45.614", + "raw_stderr": "6.656", + "latency": "755.757", + "cost_per_test": "2.852274", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.656, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:family:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.legal_research.family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 18.182, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "18.182", + "raw_stderr": "8.417", + "latency": "787.574", + "cost_per_test": "2.857814", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 8.417, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:health:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.legal_research.health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 53.333, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "53.333", + "raw_stderr": "9.264", + "latency": "663.425", + "cost_per_test": "3.159294", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 9.264, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:immigration:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.legal_research.immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 27.273, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "27.273", + "raw_stderr": "14.084", + "latency": "728.3", + "cost_per_test": "3.441413", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 14.084, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:overall:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.legal_research.overall", + "source_data": { + "dataset_name": "Legal Research Bench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Overall).", + "metric_id": "vals_ai.legal_research.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 43.75, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "legal", + "raw_score": "43.75", + "raw_stderr": "3.448", + "latency": "714.036", + "cost_per_test": "2.820942", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.448, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass", + "source_data": { + "dataset_name": "Legal Research Bench - Weighted Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Weighted Pass).", + "metric_id": "vals_ai.legal_research.weighted_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 86.067, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass", + "task_name": "Weighted Pass", + "dataset_type": "private", + "industry": "legal", + "raw_score": "86.067", + "raw_stderr": "1.441", + "latency": "714.036", + "cost_per_test": "2.820942", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.441, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__administrative_regulatory:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.weighted_pass__administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 87.365, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "87.365", + "raw_stderr": "2.257", + "latency": "636.543", + "cost_per_test": "2.731524", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.257, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__business_commercial:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.weighted_pass__business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 86.112, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "86.112", + "raw_stderr": "3.565", + "latency": "777.706", + "cost_per_test": "2.888428", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.565, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__civil_litigation:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.weighted_pass__civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 90.753, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "90.753", + "raw_stderr": "2.985", + "latency": "806.663", + "cost_per_test": "2.997652", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.985, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__constitutional_civil_rights:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 87.016, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "87.016", + "raw_stderr": "2.634", + "latency": "609.116", + "cost_per_test": "2.462184", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.634, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__criminal:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.weighted_pass__criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 87.747, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "87.747", + "raw_stderr": "2.443", + "latency": "755.757", + "cost_per_test": "2.852274", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.443, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__family:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.weighted_pass__family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 77.323, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "77.323", + "raw_stderr": "5.197", + "latency": "787.574", + "cost_per_test": "2.857814", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.197, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__health:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.weighted_pass__health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 90.816, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "90.816", + "raw_stderr": "2.648", + "latency": "663.425", + "cost_per_test": "3.159294", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.648, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__immigration:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.weighted_pass__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 78.35, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "78.35", + "raw_stderr": "8.403", + "latency": "728.3", + "cost_per_test": "3.441413", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 8.403, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-8/81cb71fc-2727-4a0b-b0d1-811c312cff0e.json b/data/vals_ai/anthropic/claude-opus-4-8/81cb71fc-2727-4a0b-b0d1-811c312cff0e.json new file mode 100644 index 0000000000000000000000000000000000000000..9d6974a41b287ce2f57061896335310b51014cb8 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-8/81cb71fc-2727-4a0b-b0d1-811c312cff0e.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/anthropic_claude-opus-4-8/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-8", + "id": "anthropic/claude-opus-4-8", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-8", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 100.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "100", + "raw_stderr": "0", + "latency": "20690.317", + "cost_per_test": "32.170449", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 60.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "60", + "raw_stderr": "1.925", + "latency": "2758.709", + "cost_per_test": "4.289393", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.925, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 76.364, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "76.364", + "raw_stderr": "1.05", + "latency": "1504.75", + "cost_per_test": "2.339669", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.05, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 71.91, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "71.91", + "raw_stderr": "0.649", + "latency": "929.902", + "cost_per_test": "2.409772", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.649, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-8/905c4f54-d9c1-4cae-b8f6-5917a9215bb1.json b/data/vals_ai/anthropic/claude-opus-4-8/905c4f54-d9c1-4cae-b8f6-5917a9215bb1.json new file mode 100644 index 0000000000000000000000000000000000000000..963fc3d46666d70d7f9c8781859ebfdd6cb6fba0 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-8/905c4f54-d9c1-4cae-b8f6-5917a9215bb1.json @@ -0,0 +1,1071 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/anthropic_claude-opus-4-8/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-8", + "id": "anthropic/claude-opus-4-8", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-8", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.7, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.7", + "raw_stderr": "0.837", + "latency": "16.602", + "cost_per_test": "0.042", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.837, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.649, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.649", + "raw_stderr": "0.929", + "latency": "16.722", + "cost_per_test": "0.041351", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.929, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.845, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.845", + "raw_stderr": "0.766", + "latency": "29.316", + "cost_per_test": "0.081956", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.766, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.683, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.683", + "raw_stderr": "1.286", + "latency": "19.219", + "cost_per_test": "0.056235", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.286, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.825, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.825", + "raw_stderr": "0.943", + "latency": "17.061", + "cost_per_test": "0.04071", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.943, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.719, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.719", + "raw_stderr": "1.054", + "latency": "72.117", + "cost_per_test": "0.177437", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.054, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.086, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.086", + "raw_stderr": "1.246", + "latency": "17.113", + "cost_per_test": "0.040556", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.246, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.677, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.677", + "raw_stderr": "1.939", + "latency": "12.47", + "cost_per_test": "0.04551", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.939, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.926, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.926", + "raw_stderr": "1.16", + "latency": "46.643", + "cost_per_test": "0.113953", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.16, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 95.855, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "95.855", + "raw_stderr": "0.542", + "latency": "18.449", + "cost_per_test": "0.055314", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.542, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.147, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.147", + "raw_stderr": "1.136", + "latency": "13.617", + "cost_per_test": "0.030976", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.136, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.585, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.585", + "raw_stderr": "0.303", + "latency": "23.44", + "cost_per_test": "0.060716", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.303, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.573, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.573", + "raw_stderr": "1.526", + "latency": "12.119", + "cost_per_test": "0.031561", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.526, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.533, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.533", + "raw_stderr": "0.682", + "latency": "23.621", + "cost_per_test": "0.061096", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.682, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.975, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.975", + "raw_stderr": "1.063", + "latency": "13.084", + "cost_per_test": "0.031378", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.063, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-8/9fdcf58c-41a1-466f-8529-ae7fd8e25aae.json b/data/vals_ai/anthropic/claude-opus-4-8/9fdcf58c-41a1-466f-8529-ae7fd8e25aae.json new file mode 100644 index 0000000000000000000000000000000000000000..99bff594f996befd0a4a4d279b2b166dc71e75a8 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-8/9fdcf58c-41a1-466f-8529-ae7fd8e25aae.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/anthropic_claude-opus-4-8/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-8", + "id": "anthropic/claude-opus-4-8", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-8", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 69.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "69", + "raw_stderr": "4.648", + "latency": "1166.322", + "cost_per_test": "2.293129", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 4.648, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-8/a26ae388-b6a0-4f81-81ed-70f87683f9bb.json b/data/vals_ai/anthropic/claude-opus-4-8/a26ae388-b6a0-4f81-81ed-70f87683f9bb.json new file mode 100644 index 0000000000000000000000000000000000000000..2d8912e58228d2d6717043e04b82d639c5dcd3f4 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-8/a26ae388-b6a0-4f81-81ed-70f87683f9bb.json @@ -0,0 +1,3486 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/hlab/anthropic_claude-opus-4-8/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Harvey's Legal Agent Benchmark", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-8", + "id": "anthropic/claude-opus-4-8", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-8", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "hlab:antitrust_competition:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1337.226", + "cost_per_test": "9.296892", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:arbitration_international_dispute_resolution:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1580.331", + "cost_per_test": "12.077575", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:banking_finance:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1245.965", + "cost_per_test": "9.097239", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:bankruptcy_restructuring:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1398.957", + "cost_per_test": "11.784326", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:capital_markets:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "1231.115", + "cost_per_test": "9.299876", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:corporate_governance:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1412.554", + "cost_per_test": "11.140007", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:corporate_ma:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1524.989", + "cost_per_test": "10.04374", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Criteria Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Criteria Pass Rate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 87.856, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate", + "task_name": "Criteria Pass Rate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.856", + "raw_stderr": "0.393", + "latency": "1447.796", + "cost_per_test": "10.222944", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.393, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__antitrust_competition:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 87.319, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.319", + "raw_stderr": "2.007", + "latency": "1337.226", + "cost_per_test": "9.296892", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.007, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__arbitration_international_dispute_resolution:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 88.172, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.172", + "raw_stderr": "1.937", + "latency": "1580.331", + "cost_per_test": "12.077575", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.937, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__banking_finance:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 87.732, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.732", + "raw_stderr": "2.004", + "latency": "1245.965", + "cost_per_test": "9.097239", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.004, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__bankruptcy_restructuring:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 83.045, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.045", + "raw_stderr": "2.211", + "latency": "1398.957", + "cost_per_test": "11.784326", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.211, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__capital_markets:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 92.157, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "92.157", + "raw_stderr": "1.687", + "latency": "1231.115", + "cost_per_test": "9.299876", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.687, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_governance:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 89.041, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.041", + "raw_stderr": "1.831", + "latency": "1412.554", + "cost_per_test": "11.140007", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.831, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_ma:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 92.949, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "92.949", + "raw_stderr": "1.452", + "latency": "1524.989", + "cost_per_test": "10.04374", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.452, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__data_privacy_cybersecurity:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 94.346, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "94.346", + "raw_stderr": "1.375", + "latency": "1370.19", + "cost_per_test": "11.031307", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.375, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__emerging_companies_venture_capital:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 89.078, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.078", + "raw_stderr": "1.825", + "latency": "1500.003", + "cost_per_test": "8.692762", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.825, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__employment_labor:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 82.772, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.772", + "raw_stderr": "2.315", + "latency": "1324.746", + "cost_per_test": "7.024491", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.315, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__energy_natural_resources:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 97.03, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "97.03", + "raw_stderr": "0.977", + "latency": "1738.257", + "cost_per_test": "17.557582", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.977, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__environmental_esg:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 86.738, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.738", + "raw_stderr": "2.034", + "latency": "1687.494", + "cost_per_test": "10.410854", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.034, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__funds_asset_management:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 84.548, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.548", + "raw_stderr": "1.954", + "latency": "1687.575", + "cost_per_test": "12.061978", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.954, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__healthcare_life_sciences:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 91.419, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "91.419", + "raw_stderr": "1.612", + "latency": "1343.378", + "cost_per_test": "10.479166", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.612, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__immigration:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 85.818, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.818", + "raw_stderr": "2.108", + "latency": "1347.383", + "cost_per_test": "7.846506", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.108, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__insurance:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 87.18, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.18", + "raw_stderr": "2.027", + "latency": "1460.758", + "cost_per_test": "9.118767", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.027, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__intellectual_property:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 90.459, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.459", + "raw_stderr": "1.749", + "latency": "1413.435", + "cost_per_test": "9.530409", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.749, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__international_trade_sanctions:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 85.714, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.714", + "raw_stderr": "2.209", + "latency": "1470.574", + "cost_per_test": "7.692416", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.209, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__litigation_dispute_resolution:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 89.888, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.888", + "raw_stderr": "1.849", + "latency": "1464.809", + "cost_per_test": "10.4087", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.849, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__real_estate:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 90.962, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.962", + "raw_stderr": "1.55", + "latency": "1485.698", + "cost_per_test": "10.731391", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.55, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__structured_finance_securitization:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 81.27, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.27", + "raw_stderr": "2.202", + "latency": "1487.616", + "cost_per_test": "11.288588", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.202, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__tax:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 80.602, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.602", + "raw_stderr": "2.291", + "latency": "1692.91", + "cost_per_test": "9.85827", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.291, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__trusts_estates_private_client:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 87.786, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.786", + "raw_stderr": "2.027", + "latency": "1340.901", + "cost_per_test": "9.820962", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.027, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__white_collar_defense_investigations:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 82.492, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.492", + "raw_stderr": "2.209", + "latency": "1200.234", + "cost_per_test": "9.056848", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.209, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:data_privacy_cybersecurity:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "1370.19", + "cost_per_test": "11.031307", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:emerging_companies_venture_capital:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "1500.003", + "cost_per_test": "8.692762", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:employment_labor:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1324.746", + "cost_per_test": "7.024491", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:energy_natural_resources:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "1738.257", + "cost_per_test": "17.557582", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:environmental_esg:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1687.494", + "cost_per_test": "10.410854", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:funds_asset_management:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1687.575", + "cost_per_test": "12.061978", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:healthcare_life_sciences:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 40.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "40", + "raw_stderr": "24.495", + "latency": "1343.378", + "cost_per_test": "10.479166", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 24.495, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:immigration:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1347.383", + "cost_per_test": "7.846506", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:insurance:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1460.758", + "cost_per_test": "9.118767", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:intellectual_property:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1413.435", + "cost_per_test": "9.530409", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:international_trade_sanctions:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1470.574", + "cost_per_test": "7.692416", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:litigation_dispute_resolution:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1464.809", + "cost_per_test": "10.4087", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:overall:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.overall", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Overall).", + "metric_id": "vals_ai.hlab.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 9.583, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "9.583", + "raw_stderr": "2.148", + "latency": "1447.796", + "cost_per_test": "10.222944", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.148, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:real_estate:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1485.698", + "cost_per_test": "10.731391", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:structured_finance_securitization:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "1487.616", + "cost_per_test": "11.288588", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:tax:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1692.91", + "cost_per_test": "9.85827", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:trusts_estates_private_client:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1340.901", + "cost_per_test": "9.820962", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:white_collar_defense_investigations:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.hlab.white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1200.234", + "cost_per_test": "9.056848", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-8/b38e5cbf-dd84-438a-b679-6c10ab52ab06.json b/data/vals_ai/anthropic/claude-opus-4-8/b38e5cbf-dd84-438a-b679-6c10ab52ab06.json new file mode 100644 index 0000000000000000000000000000000000000000..87bdea30768c8db57dc8931de24eace34a8be33f --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-8/b38e5cbf-dd84-438a-b679-6c10ab52ab06.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/programbench/anthropic_claude-opus-4-8/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProgramBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-8", + "id": "anthropic/claude-opus-4-8", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-8", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "programbench:almost:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.programbench.almost", + "source_data": { + "dataset_name": "ProgramBench - Almost Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "almost", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Almost Resolved).", + "metric_id": "vals_ai.programbench.almost.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 15.5, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "almost", + "task_name": "Almost Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "15.5", + "raw_stderr": "2.565", + "latency": "5509.223", + "cost_per_test": "36.104116", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.565, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "programbench:overall:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.programbench.overall", + "source_data": { + "dataset_name": "ProgramBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Overall).", + "metric_id": "vals_ai.programbench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 1.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "1", + "raw_stderr": "0.705", + "latency": "5509.223", + "cost_per_test": "36.104116", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.705, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "programbench:partial:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.programbench.partial", + "source_data": { + "dataset_name": "ProgramBench - Raw Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "partial", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Raw Pass Rate).", + "metric_id": "vals_ai.programbench.partial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 71.882, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "partial", + "task_name": "Raw Pass Rate", + "dataset_type": "public", + "industry": "coding", + "raw_score": "71.882", + "raw_stderr": "1.829", + "latency": "5509.223", + "cost_per_test": "36.104116", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.829, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "programbench:strict:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.programbench.strict", + "source_data": { + "dataset_name": "ProgramBench - Fully Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "strict", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Fully Resolved).", + "metric_id": "vals_ai.programbench.strict.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 1.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "strict", + "task_name": "Fully Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "1", + "raw_stderr": "0.705", + "latency": "5509.223", + "cost_per_test": "36.104116", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.705, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-8/b410d178-8862-4644-ae4f-a750b03af7c4.json b/data/vals_ai/anthropic/claude-opus-4-8/b410d178-8862-4644-ae4f-a750b03af7c4.json new file mode 100644 index 0000000000000000000000000000000000000000..c5f879b1c154f4e9c530ff79f1172f5dc246a584 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-8/b410d178-8862-4644-ae4f-a750b03af7c4.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/anthropic_claude-opus-4-8/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-8", + "id": "anthropic/claude-opus-4-8", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-8", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 85.755, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "85.755", + "raw_stderr": "1.928", + "latency": "82.364", + "cost_per_test": "0.259121", + "temperature": "1", + "max_output_tokens": "30000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.928, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-8/c6a4b226-122b-461f-bfd9-f88dc1de4a4a.json b/data/vals_ai/anthropic/claude-opus-4-8/c6a4b226-122b-461f-bfd9-f88dc1de4a4a.json new file mode 100644 index 0000000000000000000000000000000000000000..e249f9b070683a56fbc09747c94bf3d3df769b57 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-8/c6a4b226-122b-461f-bfd9-f88dc1de4a4a.json @@ -0,0 +1,381 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/anthropic_claude-opus-4-8/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-8", + "id": "anthropic/claude-opus-4-8", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-8", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 73.81, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "73.81", + "raw_stderr": "6.867", + "latency": "772.142", + "cost_per_test": "2.209393", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.867, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 88.123, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "88.123", + "raw_stderr": "2.006", + "latency": "636.405", + "cost_per_test": "1.286456", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.006, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 92.784, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "92.784", + "raw_stderr": "1.863", + "latency": "418.303", + "cost_per_test": "0.715925", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.863, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66.667", + "raw_stderr": "33.333", + "latency": "1263.337", + "cost_per_test": "3.172757", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:overall:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 88.6, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "88.6", + "raw_stderr": "1.423", + "latency": "566.945", + "cost_per_test": "1.923224", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.423, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-8/dd2f74fd-884d-4fe6-a23b-5f54579fb034.json b/data/vals_ai/anthropic/claude-opus-4-8/dd2f74fd-884d-4fe6-a23b-5f54579fb034.json new file mode 100644 index 0000000000000000000000000000000000000000..4287ab773fddced22892803381f55a839bced47a --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-8/dd2f74fd-884d-4fe6-a23b-5f54579fb034.json @@ -0,0 +1,450 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/anthropic_claude-opus-4-8/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-8", + "id": "anthropic/claude-opus-4-8", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-8", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 88.937, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.937", + "raw_stderr": "0.61", + "latency": "4.886", + "cost_per_test": "0.011109", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.61, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.703, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.703", + "raw_stderr": "0.202", + "latency": "2.245", + "cost_per_test": "0.007534", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.202, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.834, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.834", + "raw_stderr": "0.601", + "latency": "2.135", + "cost_per_test": "0.013321", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.601, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.568, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.568", + "raw_stderr": "0.474", + "latency": "3.106", + "cost_per_test": "0.009106", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.474, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.389, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.389", + "raw_stderr": "1.376", + "latency": "2.188", + "cost_per_test": "0.006984", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.376, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.976, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.976", + "raw_stderr": "1.72", + "latency": "4.075", + "cost_per_test": "0.006584", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.72, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-opus-4-8/e6c8d429-1124-4d52-8c48-885c122e30ed.json b/data/vals_ai/anthropic/claude-opus-4-8/e6c8d429-1124-4d52-8c48-885c122e30ed.json new file mode 100644 index 0000000000000000000000000000000000000000..bd559dfc4ec8b3b873cf8ad4aee1ed62d4d7a976 --- /dev/null +++ b/data/vals_ai/anthropic/claude-opus-4-8/e6c8d429-1124-4d52-8c48-885c122e30ed.json @@ -0,0 +1,195 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/cyber/anthropic_claude-opus-4-8/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CyberBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "beta", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-opus-4-8", + "id": "anthropic/claude-opus-4-8", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-opus-4-8", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "cyber:overall:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.cyber.overall", + "source_data": { + "dataset_name": "CyberBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (Overall).", + "metric_id": "vals_ai.cyber.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 50.847, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "beta", + "raw_score": "50.847", + "raw_stderr": "5.155", + "latency": "896.536", + "cost_per_test": "1.708026", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.155, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "cyber:patch:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.cyber.patch", + "source_data": { + "dataset_name": "CyberBench - Patch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "patch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (Patch).", + "metric_id": "vals_ai.cyber.patch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 81.356, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "patch", + "task_name": "Patch", + "dataset_type": "private", + "industry": "beta", + "raw_score": "81.356", + "raw_stderr": "5.07", + "latency": "1209.05", + "cost_per_test": "2.318165", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.07, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "cyber:poc:anthropic/claude-opus-4-8:score", + "evaluation_name": "vals_ai.cyber.poc", + "source_data": { + "dataset_name": "CyberBench - PoC", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "poc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (PoC).", + "metric_id": "vals_ai.cyber.poc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 20.339, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "poc", + "task_name": "PoC", + "dataset_type": "private", + "industry": "beta", + "raw_score": "20.339", + "raw_stderr": "5.24", + "latency": "584.022", + "cost_per_test": "1.097887", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.24, + "method": "vals_reported" + } + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/10e1d478-d401-4489-b0d6-ea7078eaab38.json b/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/10e1d478-d401-4489-b0d6-ea7078eaab38.json new file mode 100644 index 0000000000000000000000000000000000000000..3c6dfaadfdab5df8f5e2aaa4ba3a1e94be1e14d3 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/10e1d478-d401-4489-b0d6-ea7078eaab38.json @@ -0,0 +1,336 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/anthropic_claude-sonnet-4-20250514-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-20250514-thinking", + "id": "anthropic/claude-sonnet-4-20250514-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-20250514-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 20.635, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "20.635", + "raw_stderr": "5.078", + "latency": "127.03", + "cost_per_test": "0.115638", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.078, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 33.334, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "33.334", + "raw_stderr": "6.041", + "latency": "89.374", + "cost_per_test": "0.089769", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.041, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 46.605, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "46.605", + "raw_stderr": "6.475", + "latency": "75.8", + "cost_per_test": "0.235685", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.475, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 42.047, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "42.047", + "raw_stderr": "5.545", + "latency": "98.095", + "cost_per_test": "0.127554", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.545, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 32.005, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "32.005", + "raw_stderr": "3.215", + "latency": "104.833", + "cost_per_test": "0.110987", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.215, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/61dc529e-fb71-4966-8c84-4ebd714ad2de.json b/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/61dc529e-fb71-4966-8c84-4ebd714ad2de.json new file mode 100644 index 0000000000000000000000000000000000000000..e60da1f61f17b1da2893a369e587d19be3ab4b29 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/61dc529e-fb71-4966-8c84-4ebd714ad2de.json @@ -0,0 +1,216 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/anthropic_claude-sonnet-4-20250514-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-20250514-thinking", + "id": "anthropic/claude-sonnet-4-20250514-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-20250514-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 42.289, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "42.289", + "raw_stderr": "1.393", + "latency": "75.535", + "cost_per_test": "0.078978", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.393, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 49.88, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "49.88", + "raw_stderr": "0.985", + "latency": "46.316", + "cost_per_test": "0.052253", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.985, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 57.472, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.472", + "raw_stderr": "1.394", + "latency": "17.097", + "cost_per_test": "0.025528", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.394, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/95198574-0f00-400e-ad54-ed9f7e029d14.json b/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/95198574-0f00-400e-ad54-ed9f7e029d14.json new file mode 100644 index 0000000000000000000000000000000000000000..d71b61f73dda574f553fd98098063e0e616e6aa3 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/95198574-0f00-400e-ad54-ed9f7e029d14.json @@ -0,0 +1,225 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/anthropic_claude-sonnet-4-20250514-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-20250514-thinking", + "id": "anthropic/claude-sonnet-4-20250514-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-20250514-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 74.747, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.747", + "raw_stderr": "3.088", + "latency": "112.167", + "cost_per_test": "0.112814", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.088, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 75.253, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.253", + "raw_stderr": "3.108", + "latency": "124.748", + "cost_per_test": "0.120597", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.108, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 75.0, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75", + "raw_stderr": "2.191", + "latency": "118.458", + "cost_per_test": "0.116706", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.191, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/a0f7ac77-9385-4ddb-bd45-c20c6bc1dd24.json b/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/a0f7ac77-9385-4ddb-bd45-c20c6bc1dd24.json new file mode 100644 index 0000000000000000000000000000000000000000..c16594e681386ba91e0e41d652bb75a9615a57ba --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/a0f7ac77-9385-4ddb-bd45-c20c6bc1dd24.json @@ -0,0 +1,99 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/anthropic_claude-sonnet-4-20250514-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-20250514-thinking", + "id": "anthropic/claude-sonnet-4-20250514-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-20250514-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 74.928, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.928", + "raw_stderr": "1.042", + "latency": "48.854", + "cost_per_test": "0.050728", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.042, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/bd63dd9f-eb09-41be-8f13-965618b9113e.json b/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/bd63dd9f-eb09-41be-8f13-965618b9113e.json new file mode 100644 index 0000000000000000000000000000000000000000..9d22bf74649add914ac384951df9f193e9f81419 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/bd63dd9f-eb09-41be-8f13-965618b9113e.json @@ -0,0 +1,288 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/anthropic_claude-sonnet-4-20250514-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-20250514-thinking", + "id": "anthropic/claude-sonnet-4-20250514-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-20250514-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.758, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.758", + "raw_stderr": "0.617", + "latency": "26.601", + "cost_per_test": "0.027167", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.617, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 23.143, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "23.143", + "raw_stderr": "2.254", + "latency": "56.29", + "cost_per_test": "0.054828", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.254, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 65.274, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "65.274", + "raw_stderr": "2.433", + "latency": "44.943", + "cost_per_test": "0.044958", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.433, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 62.392, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "62.392", + "raw_stderr": "1.125", + "latency": "42.611", + "cost_per_test": "0.042318", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.125, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/c5dabd61-1fd4-4d08-a47d-d43b6c6292c5.json b/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/c5dabd61-1fd4-4d08-a47d-d43b6c6292c5.json new file mode 100644 index 0000000000000000000000000000000000000000..6c5b164d51b6374a3bcb0c5326993d07dd18945e --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/c5dabd61-1fd4-4d08-a47d-d43b6c6292c5.json @@ -0,0 +1,216 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/anthropic_claude-sonnet-4-20250514-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-20250514-thinking", + "id": "anthropic/claude-sonnet-4-20250514-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-20250514-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 60.507, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.507", + "raw_stderr": "1.398", + "latency": "32.385", + "cost_per_test": "0.029421", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.398, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 71.995, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.995", + "raw_stderr": "0.878", + "latency": "32.385", + "cost_per_test": "0.029421", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.878, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 83.483, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "83.483", + "raw_stderr": "1.062", + "latency": "32.385", + "cost_per_test": "0.029421", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.062, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/c77eb1d5-db50-4f75-8239-703645089d0f.json b/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/c77eb1d5-db50-4f75-8239-703645089d0f.json new file mode 100644 index 0000000000000000000000000000000000000000..0be0f1aed5c988bf1416fe67e3823a5b38e2c83d --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/c77eb1d5-db50-4f75-8239-703645089d0f.json @@ -0,0 +1,408 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/anthropic_claude-sonnet-4-20250514-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-20250514-thinking", + "id": "anthropic/claude-sonnet-4-20250514-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-20250514-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.084, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.084", + "raw_stderr": "0.881", + "latency": "12.413", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.881, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.902, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.902", + "raw_stderr": "0.197", + "latency": "10.91", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.197, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 88.693, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.693", + "raw_stderr": "0.555", + "latency": "9.827", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.555, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.065, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.065", + "raw_stderr": "0.455", + "latency": "10.605", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.455, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.558, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.558", + "raw_stderr": "1.148", + "latency": "8.476", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.148, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 75.086, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "75.086", + "raw_stderr": "1.651", + "latency": "11.4", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.651, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/d22dcd60-7e24-4e20-ad41-a426c868a5a2.json b/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/d22dcd60-7e24-4e20-ad41-a426c868a5a2.json new file mode 100644 index 0000000000000000000000000000000000000000..20f50174eeee92fc6d541af3b16f6e65e6dc9a32 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/d22dcd60-7e24-4e20-ad41-a426c868a5a2.json @@ -0,0 +1,981 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/anthropic_claude-sonnet-4-20250514-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-20250514-thinking", + "id": "anthropic/claude-sonnet-4-20250514-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-20250514-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.353, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.353", + "raw_stderr": "1.05", + "latency": "28.672", + "cost_per_test": "0.030852", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.05, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.819, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.819", + "raw_stderr": "1.204", + "latency": "38.29", + "cost_per_test": "0.035402", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.204, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.664, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.664", + "raw_stderr": "0.908", + "latency": "56.436", + "cost_per_test": "0.061301", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.908, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.122, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.122", + "raw_stderr": "1.758", + "latency": "55.944", + "cost_per_test": "0.052996", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.758, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.678, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.678", + "raw_stderr": "1.131", + "latency": "32.561", + "cost_per_test": "0.027652", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.131, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.296, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.296", + "raw_stderr": "1.346", + "latency": "98.218", + "cost_per_test": "0.101515", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.346, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.318, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.318", + "raw_stderr": "1.39", + "latency": "28.811", + "cost_per_test": "0.026623", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.39, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.591, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.591", + "raw_stderr": "2.201", + "latency": "31.974", + "cost_per_test": "0.029222", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.201, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 66.939, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.939", + "raw_stderr": "1.418", + "latency": "81.708", + "cost_per_test": "0.069952", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.418, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.042, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.042", + "raw_stderr": "0.696", + "latency": "64.992", + "cost_per_test": "0.064348", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.696, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.55, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.55", + "raw_stderr": "1.22", + "latency": "21.986", + "cost_per_test": "0.021591", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.22, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.865, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.865", + "raw_stderr": "0.361", + "latency": "47.087", + "cost_per_test": "0.045537", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.361, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.563, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.563", + "raw_stderr": "1.736", + "latency": "38.216", + "cost_per_test": "0.033281", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.736, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.838, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.838", + "raw_stderr": "0.844", + "latency": "59.061", + "cost_per_test": "0.060425", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.844, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.338, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.338", + "raw_stderr": "1.252", + "latency": "22.345", + "cost_per_test": "0.02236", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.252, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/d2ba7a8c-bfa0-465d-a545-05d08f8e9caa.json b/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/d2ba7a8c-bfa0-465d-a545-05d08f8e9caa.json new file mode 100644 index 0000000000000000000000000000000000000000..555094ecd5e34dd9bbe979d611b4634ccd3ad648 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/d2ba7a8c-bfa0-465d-a545-05d08f8e9caa.json @@ -0,0 +1,96 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/anthropic_claude-sonnet-4-20250514-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-20250514-thinking", + "id": "anthropic/claude-sonnet-4-20250514-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-20250514-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 69.353, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "69.353", + "raw_stderr": "2.212", + "latency": "39.572", + "cost_per_test": "0.053443", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.212, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/d6b155e8-d493-420c-bfe4-a7ab9b03d741.json b/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/d6b155e8-d493-420c-bfe4-a7ab9b03d741.json new file mode 100644 index 0000000000000000000000000000000000000000..ce63fbd4692ee0ce29c734360adb10578f3c464a --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/d6b155e8-d493-420c-bfe4-a7ab9b03d741.json @@ -0,0 +1,225 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/anthropic_claude-sonnet-4-20250514-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-20250514-thinking", + "id": "anthropic/claude-sonnet-4-20250514-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-20250514-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 5.167, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "5.167", + "raw_stderr": "1.892", + "latency": "1431.232", + "cost_per_test": "1.023628", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.892, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 4.0, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "4", + "raw_stderr": "1.269", + "latency": "3001.342", + "cost_per_test": "1.636035", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.269, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 4.584, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "4.584", + "raw_stderr": "1.139", + "latency": "2216.287", + "cost_per_test": "1.329832", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.139, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/eeafb08a-9f0c-4f57-95d1-cf4d8429cb56.json b/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/eeafb08a-9f0c-4f57-95d1-cf4d8429cb56.json new file mode 100644 index 0000000000000000000000000000000000000000..88797f915c64b25ad391b0f6237f832132254abe --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-20250514-thinking/eeafb08a-9f0c-4f57-95d1-cf4d8429cb56.json @@ -0,0 +1,276 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/anthropic_claude-sonnet-4-20250514-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-20250514-thinking", + "id": "anthropic/claude-sonnet-4-20250514-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-20250514-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 58.741, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.741", + "raw_stderr": "1.681", + "latency": "35.19", + "cost_per_test": "0.014194", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.681, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.49, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.49", + "raw_stderr": "1.669", + "latency": "351.337", + "cost_per_test": "0.463634", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.669, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.228, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.228", + "raw_stderr": "0.959", + "latency": "168.523", + "cost_per_test": "0.236746", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.959, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:anthropic/claude-sonnet-4-20250514-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 64.452, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.452", + "raw_stderr": "1.634", + "latency": "119.041", + "cost_per_test": "0.232409", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.634, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-20250514/020aebe9-b54e-4d00-98b6-dd01341b4225.json b/data/vals_ai/anthropic/claude-sonnet-4-20250514/020aebe9-b54e-4d00-98b6-dd01341b4225.json new file mode 100644 index 0000000000000000000000000000000000000000..a4d0c1e68f3797345cb73c7a073a9647c4a35eac --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-20250514/020aebe9-b54e-4d00-98b6-dd01341b4225.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/anthropic_claude-sonnet-4-20250514/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-20250514", + "id": "anthropic/claude-sonnet-4-20250514", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-20250514", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 96.894, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "96.894", + "raw_stderr": "0.967", + "latency": "8.598", + "cost_per_test": "0.008354", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.967, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 22.857, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "22.857", + "raw_stderr": "2.245", + "latency": "19.854", + "cost_per_test": "0.019844", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.245, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 59.269, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "59.269", + "raw_stderr": "2.511", + "latency": "13.721", + "cost_per_test": "0.013237", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.511, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 59.673, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "59.673", + "raw_stderr": "1.168", + "latency": "14.058", + "cost_per_test": "0.013812", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.168, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-20250514/2e0c032b-f950-499b-aaa0-742e6894f189.json b/data/vals_ai/anthropic/claude-sonnet-4-20250514/2e0c032b-f950-499b-aaa0-742e6894f189.json new file mode 100644 index 0000000000000000000000000000000000000000..1243d930e782953ab4dc339456e982a644ad991d --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-20250514/2e0c032b-f950-499b-aaa0-742e6894f189.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/anthropic_claude-sonnet-4-20250514/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-20250514", + "id": "anthropic/claude-sonnet-4-20250514", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-20250514", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 69.192, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.192", + "raw_stderr": "3.281", + "latency": "13.094", + "cost_per_test": "0.016271", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.281, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 69.697, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.697", + "raw_stderr": "3.266", + "latency": "12.541", + "cost_per_test": "0.010833", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.266, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 69.444, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.444", + "raw_stderr": "2.315", + "latency": "12.818", + "cost_per_test": "0.013552", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.315, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-20250514/59255a86-baa6-4abe-9aa5-fb589da9f7e6.json b/data/vals_ai/anthropic/claude-sonnet-4-20250514/59255a86-baa6-4abe-9aa5-fb589da9f7e6.json new file mode 100644 index 0000000000000000000000000000000000000000..94ad1e61d56a6c15145c2b0772458e48f427cfd9 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-20250514/59255a86-baa6-4abe-9aa5-fb589da9f7e6.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/anthropic_claude-sonnet-4-20250514/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-20250514", + "id": "anthropic/claude-sonnet-4-20250514", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-20250514", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 88.121, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.121", + "raw_stderr": "0.858", + "latency": "4.358", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.858, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.416, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.416", + "raw_stderr": "0.188", + "latency": "4.652", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.188, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 85.113, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.113", + "raw_stderr": "0.6", + "latency": "4.891", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.6, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.954, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.954", + "raw_stderr": "0.406", + "latency": "4.849", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.406, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.985, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.985", + "raw_stderr": "0.836", + "latency": "4.709", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.836, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.134, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.134", + "raw_stderr": "1.515", + "latency": "5.637", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.515, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-20250514/79656380-3eab-499e-97a5-ce2542164134.json b/data/vals_ai/anthropic/claude-sonnet-4-20250514/79656380-3eab-499e-97a5-ce2542164134.json new file mode 100644 index 0000000000000000000000000000000000000000..50d35701b716810930053083dba817903c9af91a --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-20250514/79656380-3eab-499e-97a5-ce2542164134.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/anthropic_claude-sonnet-4-20250514/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-20250514", + "id": "anthropic/claude-sonnet-4-20250514", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-20250514", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 72.411, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "72.411", + "raw_stderr": "1.929", + "latency": "25.674", + "cost_per_test": "0.038973", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.929, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-20250514/8bcba3a2-ca7b-473a-9e29-2a20b79eaf18.json b/data/vals_ai/anthropic/claude-sonnet-4-20250514/8bcba3a2-ca7b-473a-9e29-2a20b79eaf18.json new file mode 100644 index 0000000000000000000000000000000000000000..98cdf25e0508668a8a5a67711807aa3925a577ca --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-20250514/8bcba3a2-ca7b-473a-9e29-2a20b79eaf18.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/anthropic_claude-sonnet-4-20250514/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-20250514", + "id": "anthropic/claude-sonnet-4-20250514", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-20250514", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 57.89, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.89", + "raw_stderr": "1.412", + "latency": "11.847", + "cost_per_test": "0.0076", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.412, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 69.624, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "69.624", + "raw_stderr": "0.899", + "latency": "11.847", + "cost_per_test": "0.0076", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.899, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 81.357, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "81.357", + "raw_stderr": "1.114", + "latency": "11.847", + "cost_per_test": "0.0076", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.114, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-20250514/a25445c6-8114-4ce7-9e19-5e522597e96e.json b/data/vals_ai/anthropic/claude-sonnet-4-20250514/a25445c6-8114-4ce7-9e19-5e522597e96e.json new file mode 100644 index 0000000000000000000000000000000000000000..926d2376bd17e0992616ba5a328e5406f289e195 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-20250514/a25445c6-8114-4ce7-9e19-5e522597e96e.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/anthropic_claude-sonnet-4-20250514/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-20250514", + "id": "anthropic/claude-sonnet-4-20250514", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-20250514", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.213, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.213", + "raw_stderr": "1.057", + "latency": "7.935", + "cost_per_test": "0.010018", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.057, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 34.728, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "34.728", + "raw_stderr": "1.695", + "latency": "10.899", + "cost_per_test": "0.008309", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.695, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.777, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.777", + "raw_stderr": "1.038", + "latency": "12.501", + "cost_per_test": "0.013022", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.038, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.829, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.829", + "raw_stderr": "1.67", + "latency": "8.295", + "cost_per_test": "0.011196", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.67, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.152, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.152", + "raw_stderr": "1.112", + "latency": "10.177", + "cost_per_test": "0.008757", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.112, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 70.898, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.898", + "raw_stderr": "1.459", + "latency": "12.865", + "cost_per_test": "0.013775", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.459, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.685, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.685", + "raw_stderr": "1.38", + "latency": "7.345", + "cost_per_test": "0.008541", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.38, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.265, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.265", + "raw_stderr": "2.077", + "latency": "10.566", + "cost_per_test": "0.012016", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.077, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 69.664, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.664", + "raw_stderr": "1.385", + "latency": "11.73", + "cost_per_test": "0.010381", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.385, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.896, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.896", + "raw_stderr": "0.783", + "latency": "9.118", + "cost_per_test": "0.013312", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.783, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.71, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.71", + "raw_stderr": "1.272", + "latency": "6.537", + "cost_per_test": "0.007223", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.272, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.432, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.432", + "raw_stderr": "0.372", + "latency": "9.707", + "cost_per_test": "0.010229", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.372, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.762, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.762", + "raw_stderr": "1.765", + "latency": "10.531", + "cost_per_test": "0.008185", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.765, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.758, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.758", + "raw_stderr": "0.97", + "latency": "10.151", + "cost_per_test": "0.010341", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.97, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.714, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.714", + "raw_stderr": "1.239", + "latency": "7.25", + "cost_per_test": "0.008135", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.239, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-20250514/c8adee41-51ae-4262-be2d-824b21967727.json b/data/vals_ai/anthropic/claude-sonnet-4-20250514/c8adee41-51ae-4262-be2d-824b21967727.json new file mode 100644 index 0000000000000000000000000000000000000000..b24fd4b05afe0fff072abab4506aaac6610b23c5 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-20250514/c8adee41-51ae-4262-be2d-824b21967727.json @@ -0,0 +1,101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/anthropic_claude-sonnet-4-20250514/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-20250514", + "id": "anthropic/claude-sonnet-4-20250514", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-20250514", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 72.386, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.386", + "raw_stderr": "1.075", + "latency": "10.823", + "cost_per_test": "0.012845", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.075, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-20250514/e5a2578e-9433-4fbc-b239-221009f68f41.json b/data/vals_ai/anthropic/claude-sonnet-4-20250514/e5a2578e-9433-4fbc-b239-221009f68f41.json new file mode 100644 index 0000000000000000000000000000000000000000..fd97fd874c58d982776278dd66aa2ea688649763 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-20250514/e5a2578e-9433-4fbc-b239-221009f68f41.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/anthropic_claude-sonnet-4-20250514/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-20250514", + "id": "anthropic/claude-sonnet-4-20250514", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-20250514", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 66.511, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.511", + "raw_stderr": "1.317", + "latency": "8.517", + "cost_per_test": "0.0158", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.317, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 62.468, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.468", + "raw_stderr": "0.957", + "latency": "8.422", + "cost_per_test": "0.015974", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.957, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 58.426, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.426", + "raw_stderr": "1.39", + "latency": "8.328", + "cost_per_test": "0.016147", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.39, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-20250514/ef69fab8-8869-498e-a52e-63c46ec5ecbe.json b/data/vals_ai/anthropic/claude-sonnet-4-20250514/ef69fab8-8869-498e-a52e-63c46ec5ecbe.json new file mode 100644 index 0000000000000000000000000000000000000000..fb5e21bdfe5703af99b1135ace285c2ba72d6f40 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-20250514/ef69fab8-8869-498e-a52e-63c46ec5ecbe.json @@ -0,0 +1,346 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/anthropic_claude-sonnet-4-20250514/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-20250514", + "id": "anthropic/claude-sonnet-4-20250514", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-20250514", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 34.921, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "34.921", + "raw_stderr": "5.982", + "latency": "17.422", + "cost_per_test": "0.030408", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.982, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 23.81, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "23.81", + "raw_stderr": "5.414", + "latency": "17.56", + "cost_per_test": "0.044766", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.414, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 29.321, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "29.321", + "raw_stderr": "5.923", + "latency": "24.759", + "cost_per_test": "0.196354", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.923, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 46.27, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "46.27", + "raw_stderr": "5.601", + "latency": "16.733", + "cost_per_test": "0.062171", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.601, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 35.0, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "35", + "raw_stderr": "3.274", + "latency": "17.238", + "cost_per_test": "0.045782", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.274, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-20250514/fbe8f38f-5e4b-4ed0-b7af-48a2e402d6ae.json b/data/vals_ai/anthropic/claude-sonnet-4-20250514/fbe8f38f-5e4b-4ed0-b7af-48a2e402d6ae.json new file mode 100644 index 0000000000000000000000000000000000000000..5e1c5f9d64f0c818ffc0ce8180c3a889d7ae7290 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-20250514/fbe8f38f-5e4b-4ed0-b7af-48a2e402d6ae.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/anthropic_claude-sonnet-4-20250514/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-20250514", + "id": "anthropic/claude-sonnet-4-20250514", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-20250514", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 58.508, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.508", + "raw_stderr": "1.682", + "latency": "5.2", + "cost_per_test": "0.007941", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.682, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 52.681, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.681", + "raw_stderr": "1.705", + "latency": "449.724", + "cost_per_test": "0.452376", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.705, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 54.701, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "54.701", + "raw_stderr": "0.98", + "latency": "186.345", + "cost_per_test": "0.227257", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.98, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 52.914, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.914", + "raw_stderr": "1.704", + "latency": "104.11", + "cost_per_test": "0.221455", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.704, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-20250514/fd73e2d8-ff00-47f0-8288-49042d68b071.json b/data/vals_ai/anthropic/claude-sonnet-4-20250514/fd73e2d8-ff00-47f0-8288-49042d68b071.json new file mode 100644 index 0000000000000000000000000000000000000000..5a72b95317fedcdedefa62d3a4ca02661354a53c --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-20250514/fd73e2d8-ff00-47f0-8288-49042d68b071.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/anthropic_claude-sonnet-4-20250514/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-20250514", + "id": "anthropic/claude-sonnet-4-20250514", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-20250514", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 9.667, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "9.667", + "raw_stderr": "3.949", + "latency": "2018.192", + "cost_per_test": "2.080064", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.949, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 3.333, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "3.333", + "raw_stderr": "1.408", + "latency": "4488.047", + "cost_per_test": "2.697792", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.408, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:anthropic/claude-sonnet-4-20250514:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 6.5, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "6.5", + "raw_stderr": "2.096", + "latency": "3253.12", + "cost_per_test": "2.388928", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.096, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/00d6887b-4b76-43d0-a650-48defe7575fb.json b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/00d6887b-4b76-43d0-a650-48defe7575fb.json new file mode 100644 index 0000000000000000000000000000000000000000..3f23f726b1a31ac8727a6ef9f13d5e80836e8fe8 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/00d6887b-4b76-43d0-a650-48defe7575fb.json @@ -0,0 +1,346 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/anthropic_claude-sonnet-4-5-20250929-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-5-20250929-thinking", + "id": "anthropic/claude-sonnet-4-5-20250929-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-5-20250929-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 30.159, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "30.159", + "raw_stderr": "5.759", + "latency": "158.135", + "cost_per_test": "0.167902", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.759, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 22.222, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "22.222", + "raw_stderr": "5.327", + "latency": "240.19", + "cost_per_test": "0.09075", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.327, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 60.802, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "60.802", + "raw_stderr": "6.13", + "latency": "80.117", + "cost_per_test": "0.249092", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.13, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 55.813, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "55.813", + "raw_stderr": "5.578", + "latency": "232.901", + "cost_per_test": "0.154143", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.578, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 36.065, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "36.065", + "raw_stderr": "3.209", + "latency": "210.409", + "cost_per_test": "0.137598", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.209, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/02dbe39c-8e75-44ec-8e66-24bcf5d35f82.json b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/02dbe39c-8e75-44ec-8e66-24bcf5d35f82.json new file mode 100644 index 0000000000000000000000000000000000000000..2d0ea74abce1a3c992822678c355a1db98024ecd --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/02dbe39c-8e75-44ec-8e66-24bcf5d35f82.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/anthropic_claude-sonnet-4-5-20250929-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-5-20250929-thinking", + "id": "anthropic/claude-sonnet-4-5-20250929-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-5-20250929-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 81.633, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.633", + "raw_stderr": "3.363", + "latency": "108.446", + "cost_per_test": "0.108671", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.363, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 81.633, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.633", + "raw_stderr": "3.001", + "latency": "126.056", + "cost_per_test": "0.119852", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.001, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 81.633, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.633", + "raw_stderr": "2.254", + "latency": "117.251", + "cost_per_test": "0.114262", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.254, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/03b6f4bd-1819-4f9e-b889-27f3e9b0ffc3.json b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/03b6f4bd-1819-4f9e-b889-27f3e9b0ffc3.json new file mode 100644 index 0000000000000000000000000000000000000000..42c485a3392a1d75513d84730b04abbc3f470f3b --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/03b6f4bd-1819-4f9e-b889-27f3e9b0ffc3.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/anthropic_claude-sonnet-4-5-20250929-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-5-20250929-thinking", + "id": "anthropic/claude-sonnet-4-5-20250929-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-5-20250929-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.748, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.748", + "raw_stderr": "1.115", + "latency": "27.899", + "cost_per_test": "0.028668", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.115, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.888, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.888", + "raw_stderr": "1.219", + "latency": "26.047", + "cost_per_test": "0.028754", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.219, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.078, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.078", + "raw_stderr": "1.151", + "latency": "45.357", + "cost_per_test": "0.055856", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.151, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.512, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.512", + "raw_stderr": "2.142", + "latency": "40.257", + "cost_per_test": "0.043611", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.142, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.166, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.166", + "raw_stderr": "1.305", + "latency": "25.255", + "cost_per_test": "0.024978", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.305, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.456, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.456", + "raw_stderr": "1.606", + "latency": "81.642", + "cost_per_test": "0.093312", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.606, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.029, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.029", + "raw_stderr": "1.49", + "latency": "29.09", + "cost_per_test": "0.027233", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.49, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.89, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.89", + "raw_stderr": "2.159", + "latency": "29.478", + "cost_per_test": "0.029215", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.159, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.932, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.932", + "raw_stderr": "1.463", + "latency": "75.03", + "cost_per_test": "0.063558", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.463, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.967, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.967", + "raw_stderr": "1.123", + "latency": "37.162", + "cost_per_test": "0.048284", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.123, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.823, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.823", + "raw_stderr": "1.22", + "latency": "22.934", + "cost_per_test": "0.020915", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.22, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.357, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.357", + "raw_stderr": "0.391", + "latency": "38.2", + "cost_per_test": "0.040152", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.391, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.972, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.972", + "raw_stderr": "1.582", + "latency": "26.471", + "cost_per_test": "0.025073", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.582, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.686, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.686", + "raw_stderr": "1.171", + "latency": "43.672", + "cost_per_test": "0.050012", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.171, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.845, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.845", + "raw_stderr": "1.192", + "latency": "24.507", + "cost_per_test": "0.022662", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.192, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/355d54e9-0c3c-49cd-ab95-5061035fbab9.json b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/355d54e9-0c3c-49cd-ab95-5061035fbab9.json new file mode 100644 index 0000000000000000000000000000000000000000..aac0bed43307e3eaaad06facbb594343b4e73fc6 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/355d54e9-0c3c-49cd-ab95-5061035fbab9.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/anthropic_claude-sonnet-4-5-20250929-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-5-20250929-thinking", + "id": "anthropic/claude-sonnet-4-5-20250929-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-5-20250929-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 19.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "19", + "raw_stderr": "3.923", + "latency": "12300.813", + "cost_per_test": "0.65341", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.923, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/5f5c2f68-1c7b-403b-9cbd-dab7931d63e1.json b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/5f5c2f68-1c7b-403b-9cbd-dab7931d63e1.json new file mode 100644 index 0000000000000000000000000000000000000000..0a280fef8db1db319b1e46725fa28b89a3dff589 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/5f5c2f68-1c7b-403b-9cbd-dab7931d63e1.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/anthropic_claude-sonnet-4-5-20250929-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-5-20250929-thinking", + "id": "anthropic/claude-sonnet-4-5-20250929-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-5-20250929-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 88.926, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.926", + "raw_stderr": "0.764", + "latency": "16.072", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.764, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.126, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.126", + "raw_stderr": "0.265", + "latency": "12.969", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.265, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 88.569, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.569", + "raw_stderr": "0.584", + "latency": "12.52", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.584, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 84.084, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.084", + "raw_stderr": "0.447", + "latency": "13.286", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.447, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.725, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.725", + "raw_stderr": "1.149", + "latency": "9.747", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.149, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.076, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.076", + "raw_stderr": "1.637", + "latency": "15.124", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.637, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/866f6ea7-5b25-4952-8c8c-21807e9e9239.json b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/866f6ea7-5b25-4952-8c8c-21807e9e9239.json new file mode 100644 index 0000000000000000000000000000000000000000..20a90570440950c8e5bc0c13b31b1bf98d070619 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/866f6ea7-5b25-4952-8c8c-21807e9e9239.json @@ -0,0 +1,361 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/anthropic_claude-sonnet-4-5-20250929-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-5-20250929-thinking", + "id": "anthropic/claude-sonnet-4-5-20250929-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-5-20250929-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 30.952, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "30.952", + "raw_stderr": "7.22", + "latency": "454.571", + "cost_per_test": "1.112872", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 7.22, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 67.433, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "67.433", + "raw_stderr": "2.906", + "latency": "382.901", + "cost_per_test": "0.870678", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.906, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 82.99, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "82.99", + "raw_stderr": "2.705", + "latency": "308.744", + "cost_per_test": "0.651733", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.705, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "451.075", + "cost_per_test": "1.0515", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 70.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "70", + "raw_stderr": "2.051", + "latency": "360.558", + "cost_per_test": "0.807156", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.051, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/916f6e6b-e8ed-4c76-bcf9-60e578766118.json b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/916f6e6b-e8ed-4c76-bcf9-60e578766118.json new file mode 100644 index 0000000000000000000000000000000000000000..1f20da4e06967688731a32f936855b24e9b09d10 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/916f6e6b-e8ed-4c76-bcf9-60e578766118.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/anthropic_claude-sonnet-4-5-20250929-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-5-20250929-thinking", + "id": "anthropic/claude-sonnet-4-5-20250929-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-5-20250929-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 84.101, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "84.101", + "raw_stderr": "1.873", + "latency": "67.924", + "cost_per_test": "0.082281", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.873, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/b42c7f37-dc69-40e9-9394-18f8f04640fd.json b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/b42c7f37-dc69-40e9-9394-18f8f04640fd.json new file mode 100644 index 0000000000000000000000000000000000000000..77d054726fbc9d3cf7455f5e0af4b193d98bc0f5 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/b42c7f37-dc69-40e9-9394-18f8f04640fd.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/anthropic_claude-sonnet-4-5-20250929-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-5-20250929-thinking", + "id": "anthropic/claude-sonnet-4-5-20250929-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-5-20250929-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 66.852, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.852", + "raw_stderr": "1.327", + "latency": "69.751", + "cost_per_test": "0.071773", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.327, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 63.99, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.99", + "raw_stderr": "0.955", + "latency": "47.622", + "cost_per_test": "0.052096", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.955, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 61.129, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.129", + "raw_stderr": "1.374", + "latency": "25.492", + "cost_per_test": "0.032418", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.374, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/cdb187ea-a956-4688-8034-4806c0cd10b3.json b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/cdb187ea-a956-4688-8034-4806c0cd10b3.json new file mode 100644 index 0000000000000000000000000000000000000000..f2999a2338a9bf4e63bdc9f250fc956955295f3b --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/cdb187ea-a956-4688-8034-4806c0cd10b3.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/anthropic_claude-sonnet-4-5-20250929-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-5-20250929-thinking", + "id": "anthropic/claude-sonnet-4-5-20250929-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-5-20250929-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 22.621, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "22.621", + "raw_stderr": "3.728", + "latency": "2962.133", + "cost_per_test": "6.656984", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.728, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/cf91bc5b-26f3-4e12-ae6c-fcc3232ab7d0.json b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/cf91bc5b-26f3-4e12-ae6c-fcc3232ab7d0.json new file mode 100644 index 0000000000000000000000000000000000000000..b903d0dcad7352ba358be0c1e1bc983afa537877 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/cf91bc5b-26f3-4e12-ae6c-fcc3232ab7d0.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/anthropic_claude-sonnet-4-5-20250929-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-5-20250929-thinking", + "id": "anthropic/claude-sonnet-4-5-20250929-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-5-20250929-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 19.0, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "19", + "raw_stderr": "10.263", + "latency": "2583.917", + "cost_per_test": "2.711026", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 10.263, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 17.667, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "17.667", + "raw_stderr": "5.896", + "latency": "4046.414", + "cost_per_test": "3.434954", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.896, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 18.334, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "18.334", + "raw_stderr": "5.918", + "latency": "3315.166", + "cost_per_test": "3.07299", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.918, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/d318c417-c30f-4cb0-8b73-c856ccc9c63a.json b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/d318c417-c30f-4cb0-8b73-c856ccc9c63a.json new file mode 100644 index 0000000000000000000000000000000000000000..47f4452758619d927afd77ff0ec068ae1ac35756 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/d318c417-c30f-4cb0-8b73-c856ccc9c63a.json @@ -0,0 +1,101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/anthropic_claude-sonnet-4-5-20250929-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-5-20250929-thinking", + "id": "anthropic/claude-sonnet-4-5-20250929-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-5-20250929-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 79.306, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.306", + "raw_stderr": "0.974", + "latency": "56.361", + "cost_per_test": "0.058232", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.974, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/d7a0eb88-bcde-43d4-a1f4-5bf06e290c7e.json b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/d7a0eb88-bcde-43d4-a1f4-5bf06e290c7e.json new file mode 100644 index 0000000000000000000000000000000000000000..f8a0a3641f9621de4e94e6bbdbd95969c735c593 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/d7a0eb88-bcde-43d4-a1f4-5bf06e290c7e.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/anthropic_claude-sonnet-4-5-20250929-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-5-20250929-thinking", + "id": "anthropic/claude-sonnet-4-5-20250929-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-5-20250929-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 61.406, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.406", + "raw_stderr": "1.392", + "latency": "48.391", + "cost_per_test": "0.040814", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.392, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 73.303, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "73.303", + "raw_stderr": "0.861", + "latency": "48.391", + "cost_per_test": "0.040814", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.861, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 85.2, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "85.2", + "raw_stderr": "1.015", + "latency": "48.391", + "cost_per_test": "0.040814", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.015, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/f02660f1-8d59-4692-a6e2-969553d9a647.json b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/f02660f1-8d59-4692-a6e2-969553d9a647.json new file mode 100644 index 0000000000000000000000000000000000000000..76bb9c2b6e7ab1c5abab3a2698e1f5d11303e9e2 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/f02660f1-8d59-4692-a6e2-969553d9a647.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/anthropic_claude-sonnet-4-5-20250929-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-5-20250929-thinking", + "id": "anthropic/claude-sonnet-4-5-20250929-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-5-20250929-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.606, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.606", + "raw_stderr": "1.668", + "latency": "12.18", + "cost_per_test": "0.01402", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.668, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.072, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.072", + "raw_stderr": "1.665", + "latency": "27.456", + "cost_per_test": "0.464498", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.665, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.966, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.966", + "raw_stderr": "0.956", + "latency": "20.544", + "cost_per_test": "0.255071", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.956, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 64.219, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.219", + "raw_stderr": "1.636", + "latency": "21.997", + "cost_per_test": "0.286695", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.636, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/fa670826-6dfd-4a3c-89f8-de6ff41aed78.json b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/fa670826-6dfd-4a3c-89f8-de6ff41aed78.json new file mode 100644 index 0000000000000000000000000000000000000000..232750fb26d0f508aa55a87e061cc0d35cb82325 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929-thinking/fa670826-6dfd-4a3c-89f8-de6ff41aed78.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/anthropic_claude-sonnet-4-5-20250929-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-5-20250929-thinking", + "id": "anthropic/claude-sonnet-4-5-20250929-thinking", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-5-20250929-thinking", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 99.068, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "99.068", + "raw_stderr": "0.535", + "latency": "39.468", + "cost_per_test": "0.049942", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 40.286, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "40.286", + "raw_stderr": "2.622", + "latency": "174.048", + "cost_per_test": "0.213186", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.622, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 79.634, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "79.634", + "raw_stderr": "2.058", + "latency": "115.126", + "cost_per_test": "0.142184", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.058, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:anthropic/claude-sonnet-4-5-20250929-thinking:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 72.996, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "72.996", + "raw_stderr": "1.125", + "latency": "109.547", + "cost_per_test": "0.135104", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.125, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-5-20250929/a482c660-f42d-41c3-8c51-7e1b6a8610be.json b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929/a482c660-f42d-41c3-8c51-7e1b6a8610be.json new file mode 100644 index 0000000000000000000000000000000000000000..5c9bc33359260f9c417de676dd4bd1ce7f9eaa40 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929/a482c660-f42d-41c3-8c51-7e1b6a8610be.json @@ -0,0 +1,346 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/anthropic_claude-sonnet-4-5-20250929/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-5-20250929", + "id": "anthropic/claude-sonnet-4-5-20250929", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-5-20250929", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:anthropic/claude-sonnet-4-5-20250929:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 26.984, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "26.984", + "raw_stderr": "5.57", + "latency": "19.524", + "cost_per_test": "0.032822", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.57, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:anthropic/claude-sonnet-4-5-20250929:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 20.238, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "20.238", + "raw_stderr": "5.121", + "latency": "13.741", + "cost_per_test": "0.04246", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.121, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:anthropic/claude-sonnet-4-5-20250929:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 37.654, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "37.654", + "raw_stderr": "6.393", + "latency": "23.802", + "cost_per_test": "0.199573", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.393, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:anthropic/claude-sonnet-4-5-20250929:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 51.417, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "51.417", + "raw_stderr": "5.614", + "latency": "16.36", + "cost_per_test": "0.063363", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.614, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:anthropic/claude-sonnet-4-5-20250929:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 32.88, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "32.88", + "raw_stderr": "3.141", + "latency": "16.542", + "cost_per_test": "0.046215", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.141, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-5-20250929/c9350467-2bfb-43f1-88ff-7361c6c401f6.json b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929/c9350467-2bfb-43f1-88ff-7361c6c401f6.json new file mode 100644 index 0000000000000000000000000000000000000000..bfe24478096312d76ebc50a983b6d191e488ec19 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929/c9350467-2bfb-43f1-88ff-7361c6c401f6.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/anthropic_claude-sonnet-4-5-20250929/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-5-20250929", + "id": "anthropic/claude-sonnet-4-5-20250929", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-5-20250929", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:anthropic/claude-sonnet-4-5-20250929:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.373, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.373", + "raw_stderr": "1.67", + "latency": "9.701", + "cost_per_test": "0.007791", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.67, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:anthropic/claude-sonnet-4-5-20250929:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.538, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.538", + "raw_stderr": "1.661", + "latency": "16.257", + "cost_per_test": "0.483633", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.661, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:anthropic/claude-sonnet-4-5-20250929:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.8, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.8", + "raw_stderr": "0.962", + "latency": "13.383", + "cost_per_test": "0.237245", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.962, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:anthropic/claude-sonnet-4-5-20250929:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.49, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.49", + "raw_stderr": "1.669", + "latency": "14.192", + "cost_per_test": "0.22031", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.669, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-5-20250929/cf3e3ab9-f80e-480b-83aa-801961c7831b.json b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929/cf3e3ab9-f80e-480b-83aa-801961c7831b.json new file mode 100644 index 0000000000000000000000000000000000000000..57660d84966d3323915b414ce23f9e880c54fa7f --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-5-20250929/cf3e3ab9-f80e-480b-83aa-801961c7831b.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/anthropic_claude-sonnet-4-5-20250929/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-5-20250929", + "id": "anthropic/claude-sonnet-4-5-20250929", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-5-20250929", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:anthropic/claude-sonnet-4-5-20250929:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 84.515, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "84.515", + "raw_stderr": "1.929", + "latency": "44.419", + "cost_per_test": "0.054649", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.929, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-6-claude-code/c6ba4ae4-72b7-4656-b581-70a281be1656.json b/data/vals_ai/anthropic/claude-sonnet-4-6-claude-code/c6ba4ae4-72b7-4656-b581-70a281be1656.json new file mode 100644 index 0000000000000000000000000000000000000000..6341dfba416795f50fb4bcc445fb8595b55e2f02 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-6-claude-code/c6ba4ae4-72b7-4656-b581-70a281be1656.json @@ -0,0 +1,89 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/anthropic_claude-sonnet-4-6-claude-code/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-6-claude-code", + "id": "anthropic/claude-sonnet-4-6-claude-code", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-6-claude-code", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:anthropic/claude-sonnet-4-6-claude-code:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 55.774, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "55.774", + "raw_stderr": "4.682", + "latency": "2982.342", + "cost_per_test": "7.354471", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 4.682, + "method": "vals_reported" + } + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-6/0331d09e-a0fd-4172-aa74-b105ff350b12.json b/data/vals_ai/anthropic/claude-sonnet-4-6/0331d09e-a0fd-4172-aa74-b105ff350b12.json new file mode 100644 index 0000000000000000000000000000000000000000..c1f004ac9b69ec0588326ea6d0b31cfda1074760 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-6/0331d09e-a0fd-4172-aa74-b105ff350b12.json @@ -0,0 +1,320 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/programbench/anthropic_claude-sonnet-4-6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProgramBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-6", + "id": "anthropic/claude-sonnet-4-6", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-6", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "programbench:almost:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.programbench.almost", + "source_data": { + "dataset_name": "ProgramBench - Almost Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "almost", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Almost Resolved).", + "metric_id": "vals_ai.programbench.almost.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 2.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "almost", + "task_name": "Almost Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "2", + "raw_stderr": "0.992", + "latency": "4349.841", + "cost_per_test": "25.059427", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.992, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "programbench:overall:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.programbench.overall", + "source_data": { + "dataset_name": "ProgramBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Overall).", + "metric_id": "vals_ai.programbench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.5, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0.5", + "raw_stderr": "0.5", + "latency": "4349.841", + "cost_per_test": "25.059427", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.5, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "programbench:partial:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.programbench.partial", + "source_data": { + "dataset_name": "ProgramBench - Raw Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "partial", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Raw Pass Rate).", + "metric_id": "vals_ai.programbench.partial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 59.103, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "partial", + "task_name": "Raw Pass Rate", + "dataset_type": "public", + "industry": "coding", + "raw_score": "59.103", + "raw_stderr": "1.672", + "latency": "4349.841", + "cost_per_test": "25.059427", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.672, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "programbench:strict:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.programbench.strict", + "source_data": { + "dataset_name": "ProgramBench - Fully Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "strict", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Fully Resolved).", + "metric_id": "vals_ai.programbench.strict.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.5, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "strict", + "task_name": "Fully Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0.5", + "raw_stderr": "0.5", + "latency": "4349.841", + "cost_per_test": "25.059427", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.5, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-6/1da7585b-c765-4a5e-a254-752797ab7781.json b/data/vals_ai/anthropic/claude-sonnet-4-6/1da7585b-c765-4a5e-a254-752797ab7781.json new file mode 100644 index 0000000000000000000000000000000000000000..28a3caa443aafcd4293aa9f58e89440465ac540d --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-6/1da7585b-c765-4a5e-a254-752797ab7781.json @@ -0,0 +1,104 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/anthropic_claude-sonnet-4-6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-6", + "id": "anthropic/claude-sonnet-4-6", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-6", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 51.476, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "51.476", + "raw_stderr": "4.645", + "latency": "1572.118", + "cost_per_test": "5.907892", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 4.645, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-6/2643bda5-8d8f-43fc-907a-64dcb0a19f21.json b/data/vals_ai/anthropic/claude-sonnet-4-6/2643bda5-8d8f-43fc-907a-64dcb0a19f21.json new file mode 100644 index 0000000000000000000000000000000000000000..20d76cb0f9e5b2161f478c135ab65fe65c5c1264 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-6/2643bda5-8d8f-43fc-907a-64dcb0a19f21.json @@ -0,0 +1,374 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/anthropic_claude-sonnet-4-6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-6", + "id": "anthropic/claude-sonnet-4-6", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-6", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 41.27, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "41.27", + "raw_stderr": "6.185", + "latency": "196.409", + "cost_per_test": "0.278746", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.185, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "sage:automata:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 35.318, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "35.318", + "raw_stderr": "6.137", + "latency": "170.579", + "cost_per_test": "0.199244", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.137, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "sage:calculus:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 60.494, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "60.494", + "raw_stderr": "6.324", + "latency": "207.52", + "cost_per_test": "0.458947", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning": "true", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.324, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true" + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 63.158, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "63.158", + "raw_stderr": "5.474", + "latency": "123.106", + "cost_per_test": "0.225951", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.474, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "sage:overall:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 46.582, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "46.582", + "raw_stderr": "3.43", + "latency": "163.365", + "cost_per_test": "0.234647", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.43, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-6/354e7c1c-caae-4631-8719-c154fa5007b8.json b/data/vals_ai/anthropic/claude-sonnet-4-6/354e7c1c-caae-4631-8719-c154fa5007b8.json new file mode 100644 index 0000000000000000000000000000000000000000..87ddd990b8458439bb51b490c44adccad9997b2a --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-6/354e7c1c-caae-4631-8719-c154fa5007b8.json @@ -0,0 +1,240 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/anthropic_claude-sonnet-4-6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-6", + "id": "anthropic/claude-sonnet-4-6", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-6", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 74.96, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "74.96", + "raw_stderr": "1.221", + "latency": "45.711", + "cost_per_test": "0.062771", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.221, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 67.726, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.726", + "raw_stderr": "0.921", + "latency": "29.74", + "cost_per_test": "0.04467", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.921, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 60.493, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.493", + "raw_stderr": "1.378", + "latency": "13.769", + "cost_per_test": "0.02657", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.378, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-6/39bc6464-c2ae-4b81-b35b-5194fc079fb3.json b/data/vals_ai/anthropic/claude-sonnet-4-6/39bc6464-c2ae-4b81-b35b-5194fc079fb3.json new file mode 100644 index 0000000000000000000000000000000000000000..65e324883fc9cc8921674ae5e69e16c9a18ee5e2 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-6/39bc6464-c2ae-4b81-b35b-5194fc079fb3.json @@ -0,0 +1,320 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/anthropic_claude-sonnet-4-6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-6", + "id": "anthropic/claude-sonnet-4-6", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-6", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 100.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "100", + "raw_stderr": "0", + "latency": "15669.003", + "cost_per_test": "12.735719", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "false", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "false", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 34.444, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "34.444", + "raw_stderr": "2.94", + "latency": "2089.2", + "cost_per_test": "1.698096", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "false", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.94, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "false", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66.667", + "raw_stderr": "1.603", + "latency": "1139.564", + "cost_per_test": "0.926234", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "false", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.603, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "false", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 57.303, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "57.303", + "raw_stderr": "0.649", + "latency": "704.225", + "cost_per_test": "0.572392", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "false", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.649, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "false", + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-6/4c455bfb-5899-4e29-bbc7-2e4e188747f5.json b/data/vals_ai/anthropic/claude-sonnet-4-6/4c455bfb-5899-4e29-bbc7-2e4e188747f5.json new file mode 100644 index 0000000000000000000000000000000000000000..be726064787cd0f646ab7acbf2452a52042c64ed --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-6/4c455bfb-5899-4e29-bbc7-2e4e188747f5.json @@ -0,0 +1,3586 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/hlab/anthropic_claude-sonnet-4-6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Harvey's Legal Agent Benchmark", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-6", + "id": "anthropic/claude-sonnet-4-6", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-6", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "hlab:antitrust_competition:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "994.295", + "cost_per_test": "2.673062", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:arbitration_international_dispute_resolution:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1067.754", + "cost_per_test": "3.767811", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:banking_finance:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "982.437", + "cost_per_test": "2.744751", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:bankruptcy_restructuring:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1132.813", + "cost_per_test": "3.545522", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:capital_markets:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "923.605", + "cost_per_test": "2.396947", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:corporate_governance:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1323.158", + "cost_per_test": "3.276748", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:corporate_ma:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1143.37", + "cost_per_test": "3.051018", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Criteria Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Criteria Pass Rate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 86.655, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate", + "task_name": "Criteria Pass Rate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.655", + "raw_stderr": "0.409", + "latency": "1122.711", + "cost_per_test": "3.042346", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.409, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__antitrust_competition:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 87.319, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.319", + "raw_stderr": "2.007", + "latency": "994.295", + "cost_per_test": "2.673062", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.007, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__arbitration_international_dispute_resolution:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 85.663, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.663", + "raw_stderr": "2.102", + "latency": "1067.754", + "cost_per_test": "3.767811", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.102, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__banking_finance:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 91.45, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "91.45", + "raw_stderr": "1.708", + "latency": "982.437", + "cost_per_test": "2.744751", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.708, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__bankruptcy_restructuring:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 85.813, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.813", + "raw_stderr": "2.056", + "latency": "1132.813", + "cost_per_test": "3.545522", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.056, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__capital_markets:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 86.667, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.667", + "raw_stderr": "2.133", + "latency": "923.605", + "cost_per_test": "2.396947", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.133, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_governance:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 86.986, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.986", + "raw_stderr": "1.972", + "latency": "1323.158", + "cost_per_test": "3.276748", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.972, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_ma:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 91.026, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "91.026", + "raw_stderr": "1.621", + "latency": "1143.37", + "cost_per_test": "3.051018", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.621, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__data_privacy_cybersecurity:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 89.046, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.046", + "raw_stderr": "1.86", + "latency": "1034.441", + "cost_per_test": "3.1184", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.86, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__emerging_companies_venture_capital:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 84.983, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.983", + "raw_stderr": "2.091", + "latency": "1109.161", + "cost_per_test": "3.23361", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.091, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__employment_labor:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 78.652, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.652", + "raw_stderr": "2.512", + "latency": "1115.003", + "cost_per_test": "2.612081", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.512, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__energy_natural_resources:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 95.05, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "95.05", + "raw_stderr": "1.248", + "latency": "1396.089", + "cost_per_test": "4.475508", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.248, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__environmental_esg:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 85.663, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.663", + "raw_stderr": "2.102", + "latency": "1138.239", + "cost_per_test": "3.211685", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.102, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__funds_asset_management:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 82.507, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.507", + "raw_stderr": "2.054", + "latency": "1152.158", + "cost_per_test": "3.25694", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.054, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__healthcare_life_sciences:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 88.779, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.779", + "raw_stderr": "1.816", + "latency": "1277.58", + "cost_per_test": "3.819729", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.816, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__immigration:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 86.546, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.546", + "raw_stderr": "2.061", + "latency": "1046.023", + "cost_per_test": "2.059771", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.061, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__insurance:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 84.615, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.615", + "raw_stderr": "2.188", + "latency": "999.671", + "cost_per_test": "2.759668", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.188, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__intellectual_property:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 90.106, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.106", + "raw_stderr": "1.778", + "latency": "958.582", + "cost_per_test": "2.381656", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.778, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__international_trade_sanctions:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 81.746, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.746", + "raw_stderr": "2.438", + "latency": "1055.992", + "cost_per_test": "2.391323", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.438, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__litigation_dispute_resolution:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 89.513, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.513", + "raw_stderr": "1.879", + "latency": "1029.633", + "cost_per_test": "2.465192", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.879, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__real_estate:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 90.379, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.379", + "raw_stderr": "1.595", + "latency": "1116.431", + "cost_per_test": "2.958384", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.595, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__structured_finance_securitization:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 80.952, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.952", + "raw_stderr": "2.216", + "latency": "1117.966", + "cost_per_test": "3.023182", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.216, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__tax:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 85.953, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.953", + "raw_stderr": "2.013", + "latency": "1412.568", + "cost_per_test": "4.269538", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.013, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__trusts_estates_private_client:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 85.114, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.114", + "raw_stderr": "2.203", + "latency": "1290.311", + "cost_per_test": "2.740693", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.203, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__white_collar_defense_investigations:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 84.175, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.175", + "raw_stderr": "2.121", + "latency": "1127.789", + "cost_per_test": "2.783077", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.121, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:data_privacy_cybersecurity:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1034.441", + "cost_per_test": "3.1184", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:emerging_companies_venture_capital:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "1109.161", + "cost_per_test": "3.23361", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:employment_labor:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1115.003", + "cost_per_test": "2.612081", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:energy_natural_resources:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "1396.089", + "cost_per_test": "4.475508", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:environmental_esg:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1138.239", + "cost_per_test": "3.211685", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:funds_asset_management:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1152.158", + "cost_per_test": "3.25694", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:healthcare_life_sciences:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "1277.58", + "cost_per_test": "3.819729", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:immigration:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1046.023", + "cost_per_test": "2.059771", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:insurance:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "999.671", + "cost_per_test": "2.759668", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:intellectual_property:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "958.582", + "cost_per_test": "2.381656", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:international_trade_sanctions:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1055.992", + "cost_per_test": "2.391323", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:litigation_dispute_resolution:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1029.633", + "cost_per_test": "2.465192", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:overall:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.overall", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Overall).", + "metric_id": "vals_ai.hlab.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 5.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "5", + "raw_stderr": "1.646", + "latency": "1122.711", + "cost_per_test": "3.042346", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.646, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:real_estate:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1116.431", + "cost_per_test": "2.958384", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:structured_finance_securitization:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1117.966", + "cost_per_test": "3.023182", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:tax:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1412.568", + "cost_per_test": "4.269538", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:trusts_estates_private_client:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1290.311", + "cost_per_test": "2.740693", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:white_collar_defense_investigations:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.hlab.white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1127.789", + "cost_per_test": "2.783077", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-6/5666e114-8ac1-4fc3-80da-92168e2142d5.json b/data/vals_ai/anthropic/claude-sonnet-4-6/5666e114-8ac1-4fc3-80da-92168e2142d5.json new file mode 100644 index 0000000000000000000000000000000000000000..958bcdb6bab056efa02880c72dd59ed27a13b9cb --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-6/5666e114-8ac1-4fc3-80da-92168e2142d5.json @@ -0,0 +1,462 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/anthropic_claude-sonnet-4-6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-6", + "id": "anthropic/claude-sonnet-4-6", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-6", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.991, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.991", + "raw_stderr": "1.018", + "latency": "27.372", + "cost_per_test": "0.003786", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.018, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.469, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.469", + "raw_stderr": "0.191", + "latency": "19.328", + "cost_per_test": "0.002893", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.191, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 86.574, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.574", + "raw_stderr": "0.593", + "latency": "20.233", + "cost_per_test": "0.005224", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.593, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.12, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.12", + "raw_stderr": "0.476", + "latency": "21.628", + "cost_per_test": "0.004202", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.476, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.469, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.469", + "raw_stderr": "1.346", + "latency": "27.051", + "cost_per_test": "0.002685", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.346, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 74.097, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "74.097", + "raw_stderr": "1.56", + "latency": "14.158", + "cost_per_test": "0.006421", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.56, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-6/572b6817-0536-4f12-a56e-a6c70d9cb270.json b/data/vals_ai/anthropic/claude-sonnet-4-6/572b6817-0536-4f12-a56e-a6c70d9cb270.json new file mode 100644 index 0000000000000000000000000000000000000000..21ca801dbafbe5e6eda77b12b3d165c21628df74 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-6/572b6817-0536-4f12-a56e-a6c70d9cb270.json @@ -0,0 +1,104 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/anthropic_claude-sonnet-4-6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-6", + "id": "anthropic/claude-sonnet-4-6", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-6", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 45.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "45", + "raw_stderr": "4.975", + "latency": "5463.593", + "cost_per_test": "2.162108", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 4.975, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-6/59ea321c-8689-49af-9d3f-b4656512d63f.json b/data/vals_ai/anthropic/claude-sonnet-4-6/59ea321c-8689-49af-9d3f-b4656512d63f.json new file mode 100644 index 0000000000000000000000000000000000000000..cfa82c65349478a60b9c993836b518a1a597a7d4 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-6/59ea321c-8689-49af-9d3f-b4656512d63f.json @@ -0,0 +1,1101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/anthropic_claude-sonnet-4-6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-6", + "id": "anthropic/claude-sonnet-4-6", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-6", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.748, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.748", + "raw_stderr": "1.305", + "latency": "25.286", + "cost_per_test": "0.030244", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.305, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.888, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.888", + "raw_stderr": "1.573", + "latency": "41.95", + "cost_per_test": "0.052095", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.573, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.873, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.873", + "raw_stderr": "1.369", + "latency": "73.25", + "cost_per_test": "0.099424", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.369, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.0, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90", + "raw_stderr": "2.382", + "latency": "49.213", + "cost_per_test": "0.060432", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.382, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.64, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.64", + "raw_stderr": "1.528", + "latency": "24.868", + "cost_per_test": "0.029023", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.528, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.243, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.243", + "raw_stderr": "1.491", + "latency": "254.237", + "cost_per_test": "0.309322", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.491, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.396, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.396", + "raw_stderr": "1.553", + "latency": "29.104", + "cost_per_test": "0.03146", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.553, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.102, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.102", + "raw_stderr": "2.282", + "latency": "24.579", + "cost_per_test": "0.030658", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.282, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.569, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.569", + "raw_stderr": "1.442", + "latency": "72.778", + "cost_per_test": "0.073486", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.442, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.078, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.078", + "raw_stderr": "1.353", + "latency": "45.334", + "cost_per_test": "0.065662", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.353, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.117, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.117", + "raw_stderr": "1.401", + "latency": "21.251", + "cost_per_test": "0.022672", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.401, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.341, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.341", + "raw_stderr": "0.434", + "latency": "55.406", + "cost_per_test": "0.066412", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.434, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.172, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.172", + "raw_stderr": "1.97", + "latency": "20.743", + "cost_per_test": "0.02245", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.97, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.61, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.61", + "raw_stderr": "1.234", + "latency": "64.215", + "cost_per_test": "0.08129", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.234, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.343, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.343", + "raw_stderr": "1.339", + "latency": "28.875", + "cost_per_test": "0.021554", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.339, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-6/65f5195e-7008-4c81-b7ef-4b5d3a5851ae.json b/data/vals_ai/anthropic/claude-sonnet-4-6/65f5195e-7008-4c81-b7ef-4b5d3a5851ae.json new file mode 100644 index 0000000000000000000000000000000000000000..f25085e3c2fc6d3b936caabc11a406bbdf5d9600 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-6/65f5195e-7008-4c81-b7ef-4b5d3a5851ae.json @@ -0,0 +1,308 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/code-migration/anthropic_claude-sonnet-4-6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Code Migration", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-6", + "id": "anthropic/claude-sonnet-4-6", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-6", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "code-migration:cli:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.code-migration.cli", + "source_data": { + "dataset_name": "Code Migration - CLI", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cli", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (CLI).", + "metric_id": "vals_ai.code-migration.cli.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 30.333, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cli", + "task_name": "CLI", + "dataset_type": "private", + "industry": "coding", + "raw_score": "30.333", + "raw_stderr": "2.311", + "latency": "7067.428", + "cost_per_test": "48.506808", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.311, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "code-migration:cobol:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.code-migration.cobol", + "source_data": { + "dataset_name": "Code Migration - COBOL", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cobol", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (COBOL).", + "metric_id": "vals_ai.code-migration.cobol.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 68.571, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cobol", + "task_name": "COBOL", + "dataset_type": "private", + "industry": "coding", + "raw_score": "68.571", + "raw_stderr": "15.028", + "latency": "2713.85", + "cost_per_test": "8.510144", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 15.028, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "code-migration:code_quality:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.code-migration.code_quality", + "source_data": { + "dataset_name": "Code Migration - Code Quality", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "code_quality", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Code Quality).", + "metric_id": "vals_ai.code-migration.code_quality.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 3.997, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "code_quality", + "task_name": "Code Quality", + "dataset_type": "private", + "industry": "coding", + "raw_score": "3.997", + "raw_stderr": "0.032", + "latency": "5979.033", + "cost_per_test": "38.507642", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.032, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "code-migration:overall:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.code-migration.overall", + "source_data": { + "dataset_name": "Code Migration - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Overall).", + "metric_id": "vals_ai.code-migration.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 39.892, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "39.892", + "raw_stderr": "4.138", + "latency": "5979.033", + "cost_per_test": "38.507642", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 4.138, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-6/7e53100a-9fd8-4117-901f-d4e63f327879.json b/data/vals_ai/anthropic/claude-sonnet-4-6/7e53100a-9fd8-4117-901f-d4e63f327879.json new file mode 100644 index 0000000000000000000000000000000000000000..d1f64544dd3933095afbac58aed50e1457c1cfca --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-6/7e53100a-9fd8-4117-901f-d4e63f327879.json @@ -0,0 +1,428 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/anthropic_claude-sonnet-4-6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-6", + "id": "anthropic/claude-sonnet-4-6", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-6", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "66.667", + "raw_stderr": "1.609", + "latency": "20.037", + "cost_per_test": "0.287644", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.609, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 51.033, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "51.033", + "raw_stderr": "1.076", + "latency": "629.1", + "cost_per_test": "2.284299", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.076, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_index:overall:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 60.065, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "60.065", + "raw_stderr": "1.525", + "latency": "711.832", + "cost_per_test": "2.15182", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.525, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 77.451, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "77.451", + "raw_stderr": "4.138", + "latency": "633.682", + "cost_per_test": "1.706873" + }, + "uncertainty": { + "standard_error": { + "value": 4.138, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 57.303, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "57.303", + "raw_stderr": "0.649", + "latency": "704.225", + "cost_per_test": "0.572392", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "false", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.649, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "false", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 56.223, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "56.223", + "raw_stderr": "6.546", + "latency": "1572.118", + "cost_per_test": "5.907892", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.546, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-6/807b6c8c-f9c7-4848-82f7-2fba597c90f4.json b/data/vals_ai/anthropic/claude-sonnet-4-6/807b6c8c-f9c7-4848-82f7-2fba597c90f4.json new file mode 100644 index 0000000000000000000000000000000000000000..87203eb8b3ed5b654fed587c459036b503e4e07d --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-6/807b6c8c-f9c7-4848-82f7-2fba597c90f4.json @@ -0,0 +1,249 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/anthropic_claude-sonnet-4-6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-6", + "id": "anthropic/claude-sonnet-4-6", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-6", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 85.859, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.859", + "raw_stderr": "3.397", + "latency": "364.556", + "cost_per_test": "0.398667", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.397, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 85.354, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.354", + "raw_stderr": "3.408", + "latency": "391.171", + "cost_per_test": "0.418665", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.408, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 85.606, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.606", + "raw_stderr": "2.406", + "latency": "377.864", + "cost_per_test": "0.408666", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.406, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-6/80d2e530-28a7-4adb-a667-2f77bbc14570.json b/data/vals_ai/anthropic/claude-sonnet-4-6/80d2e530-28a7-4adb-a667-2f77bbc14570.json new file mode 100644 index 0000000000000000000000000000000000000000..ef64ba7f8ae8837fc13be9ed527a85bf3b0f8676 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-6/80d2e530-28a7-4adb-a667-2f77bbc14570.json @@ -0,0 +1,308 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/anthropic_claude-sonnet-4-6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-6", + "id": "anthropic/claude-sonnet-4-6", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-6", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 63.17, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.17", + "raw_stderr": "1.647", + "latency": "9.373", + "cost_per_test": "0.014257", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.647, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 66.084, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.084", + "raw_stderr": "1.616", + "latency": "26.038", + "cost_per_test": "0.620313", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.616, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 65.307, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.307", + "raw_stderr": "0.938", + "latency": "18.483", + "cost_per_test": "0.307405", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.938, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.667", + "raw_stderr": "1.609", + "latency": "20.037", + "cost_per_test": "0.287644", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.609, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-6/87343980-0a08-43fd-a7d5-61ff6ceeb992.json b/data/vals_ai/anthropic/claude-sonnet-4-6/87343980-0a08-43fd-a7d5-61ff6ceeb992.json new file mode 100644 index 0000000000000000000000000000000000000000..44a13d8a7739bc72e26a625a5b9ad79ce88eb355 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-6/87343980-0a08-43fd-a7d5-61ff6ceeb992.json @@ -0,0 +1,320 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/anthropic_claude-sonnet-4-6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-6", + "id": "anthropic/claude-sonnet-4-6", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-6", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.758, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.758", + "raw_stderr": "0.617", + "latency": "16.263", + "cost_per_test": "0.022982", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.617, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "lcb:hard:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 60.571, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "60.571", + "raw_stderr": "2.612", + "latency": "703.081", + "cost_per_test": "0.872008", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.612, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "lcb:medium:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 86.945, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "86.945", + "raw_stderr": "1.722", + "latency": "165.683", + "cost_per_test": "0.216735", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.722, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "lcb:overall:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 82.091, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "82.091", + "raw_stderr": "1.063", + "latency": "295.009", + "cost_per_test": "0.370575", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.063, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-6/88219c49-bcb9-46da-a05a-9d9b83111542.json b/data/vals_ai/anthropic/claude-sonnet-4-6/88219c49-bcb9-46da-a05a-9d9b83111542.json new file mode 100644 index 0000000000000000000000000000000000000000..f854b8e2cee4392956930fd72f167ff5b6e89aa1 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-6/88219c49-bcb9-46da-a05a-9d9b83111542.json @@ -0,0 +1,784 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/anthropic_claude-sonnet-4-6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-6", + "id": "anthropic/claude-sonnet-4-6", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-6", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 48.92, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "48.92", + "raw_stderr": "1.314", + "latency": "605.669", + "cost_per_test": "2.175111", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.314, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 38.795, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "38.795", + "raw_stderr": "0.232", + "latency": "695.588", + "cost_per_test": "2.409559", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.232, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 41.003, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "41.003", + "raw_stderr": "0.611", + "latency": "803.451", + "cost_per_test": "3.196849", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.611, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 65.77, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.77", + "raw_stderr": "1.75", + "latency": "764.736", + "cost_per_test": "2.519775", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.75, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 65.119, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.119", + "raw_stderr": "2.993", + "latency": "427.22", + "cost_per_test": "1.05356", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.993, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 17.561, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "17.561", + "raw_stderr": "1.726", + "latency": "942.334", + "cost_per_test": "2.769029", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.726, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 71.557, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.557", + "raw_stderr": "1.493", + "latency": "705.169", + "cost_per_test": "2.368893", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.493, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 72.004, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "72.004", + "raw_stderr": "1.524", + "latency": "418.676", + "cost_per_test": "2.178056", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.524, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 58.461, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.461", + "raw_stderr": "2.367", + "latency": "448.902", + "cost_per_test": "1.407303", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.367, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:overall:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 51.035, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "51.035", + "raw_stderr": "0.329", + "latency": "695.588", + "cost_per_test": "2.409559", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.329, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 18.923, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "18.923", + "raw_stderr": "1.821", + "latency": "1243.255", + "cost_per_test": "4.197491", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.821, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-6/8f607601-4b7d-42db-86c7-7793f9afc054.json b/data/vals_ai/anthropic/claude-sonnet-4-6/8f607601-4b7d-42db-86c7-7793f9afc054.json new file mode 100644 index 0000000000000000000000000000000000000000..52f665825dd7a16567ecdee1d4b9d9c19b7c44d3 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-6/8f607601-4b7d-42db-86c7-7793f9afc054.json @@ -0,0 +1,391 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/anthropic_claude-sonnet-4-6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-6", + "id": "anthropic/claude-sonnet-4-6", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-6", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 50.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "50", + "raw_stderr": "7.809", + "latency": "1095.035", + "cost_per_test": "3.079715", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 7.809, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 75.096, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "75.096", + "raw_stderr": "2.682", + "latency": "557.755", + "cost_per_test": "1.40845", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.682, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 87.113, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "87.113", + "raw_stderr": "2.412", + "latency": "319.967", + "cost_per_test": "0.766728", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.412, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "751.122", + "cost_per_test": "2.075859", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:overall:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 77.4, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "77.4", + "raw_stderr": "1.872", + "latency": "511.785", + "cost_per_test": "1.303853", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.872, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-6/93992fff-6d7d-4cb4-b86a-210c36612966.json b/data/vals_ai/anthropic/claude-sonnet-4-6/93992fff-6d7d-4cb4-b86a-210c36612966.json new file mode 100644 index 0000000000000000000000000000000000000000..3eea8433c7fccccd4e30fbb6471d0e19939dc172 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-6/93992fff-6d7d-4cb4-b86a-210c36612966.json @@ -0,0 +1,104 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/public-benefits-bench/anthropic_claude-sonnet-4-6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Public Benefits Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "social mobility", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-6", + "id": "anthropic/claude-sonnet-4-6", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-6", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "public-benefits-bench:overall:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.public-benefits-bench.overall", + "source_data": { + "dataset_name": "Public Benefits Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Public Benefits Bench v1.1 (Overall).", + "metric_id": "vals_ai.public-benefits-bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "score_details": { + "score": 62.449, + "details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "social mobility", + "raw_score": "62.449", + "raw_stderr": "1.26", + "latency": "1017.887", + "cost_per_test": "1.105228", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.26, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-6/b8208f1f-1c3f-42a6-b904-65639a47fdea.json b/data/vals_ai/anthropic/claude-sonnet-4-6/b8208f1f-1c3f-42a6-b904-65639a47fdea.json new file mode 100644 index 0000000000000000000000000000000000000000..1f7265218fb8272152f3546f527fcddefe744842 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-6/b8208f1f-1c3f-42a6-b904-65639a47fdea.json @@ -0,0 +1,107 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/anthropic_claude-sonnet-4-6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-6", + "id": "anthropic/claude-sonnet-4-6", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-6", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 83.584, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.584", + "raw_stderr": "0.891", + "latency": "131.728", + "cost_per_test": "0.144122", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.891, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-6/c918588f-7630-4d94-9806-2796c9a09b35.json b/data/vals_ai/anthropic/claude-sonnet-4-6/c918588f-7630-4d94-9806-2796c9a09b35.json new file mode 100644 index 0000000000000000000000000000000000000000..ee0338e42cc4ca9587befb3ed336488d111c92b1 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-6/c918588f-7630-4d94-9806-2796c9a09b35.json @@ -0,0 +1,1260 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_research/anthropic_claude-sonnet-4-6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Legal Research Bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-6", + "id": "anthropic/claude-sonnet-4-6", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-6", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_research:administrative_regulatory:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.legal_research.administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 43.939, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "43.939", + "raw_stderr": "6.156", + "latency": "901.446", + "cost_per_test": "2.410326", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.156, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:business_commercial:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.legal_research.business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 34.884, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "34.884", + "raw_stderr": "7.354", + "latency": "1023.821", + "cost_per_test": "2.037941", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 7.354, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:civil_litigation:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.legal_research.civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 29.63, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "29.63", + "raw_stderr": "8.955", + "latency": "1001.064", + "cost_per_test": "2.550366", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 8.955, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:constitutional_civil_rights:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.legal_research.constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 34.211, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "34.211", + "raw_stderr": "7.799", + "latency": "853.113", + "cost_per_test": "1.965994", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 7.799, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:criminal:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.legal_research.criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 36.842, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "36.842", + "raw_stderr": "6.446", + "latency": "989.26", + "cost_per_test": "1.99149", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.446, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:family:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.legal_research.family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 27.273, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "27.273", + "raw_stderr": "9.719", + "latency": "1189.264", + "cost_per_test": "2.408033", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 9.719, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:health:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.legal_research.health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 53.333, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "53.333", + "raw_stderr": "9.264", + "latency": "1047.406", + "cost_per_test": "3.45972", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 9.264, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:immigration:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.legal_research.immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 27.273, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "27.273", + "raw_stderr": "14.084", + "latency": "1048.83", + "cost_per_test": "2.703862", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 14.084, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:overall:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.legal_research.overall", + "source_data": { + "dataset_name": "Legal Research Bench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Overall).", + "metric_id": "vals_ai.legal_research.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 38.462, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "legal", + "raw_score": "38.462", + "raw_stderr": "3.381", + "latency": "989.604", + "cost_per_test": "2.302948", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.381, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass", + "source_data": { + "dataset_name": "Legal Research Bench - Weighted Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Weighted Pass).", + "metric_id": "vals_ai.legal_research.weighted_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 85.339, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass", + "task_name": "Weighted Pass", + "dataset_type": "private", + "industry": "legal", + "raw_score": "85.339", + "raw_stderr": "1.359", + "latency": "989.604", + "cost_per_test": "2.302948", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.359, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__administrative_regulatory:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.weighted_pass__administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 86.797, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "86.797", + "raw_stderr": "2.273", + "latency": "901.446", + "cost_per_test": "2.410326", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.273, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__business_commercial:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.weighted_pass__business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 87.972, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "87.972", + "raw_stderr": "2.977", + "latency": "1023.821", + "cost_per_test": "2.037941", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.977, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__civil_litigation:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.weighted_pass__civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 87.434, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "87.434", + "raw_stderr": "2.883", + "latency": "1001.064", + "cost_per_test": "2.550366", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.883, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__constitutional_civil_rights:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 87.871, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "87.871", + "raw_stderr": "2.478", + "latency": "853.113", + "cost_per_test": "1.965994", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.478, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__criminal:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.weighted_pass__criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 85.796, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "85.796", + "raw_stderr": "2.204", + "latency": "989.26", + "cost_per_test": "1.99149", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.204, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__family:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.weighted_pass__family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 76.751, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "76.751", + "raw_stderr": "4.6", + "latency": "1189.264", + "cost_per_test": "2.408033", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 4.6, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__health:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.weighted_pass__health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 90.283, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "90.283", + "raw_stderr": "2.565", + "latency": "1047.406", + "cost_per_test": "3.45972", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.565, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__immigration:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.weighted_pass__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 77.81, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "77.81", + "raw_stderr": "8.892", + "latency": "1048.83", + "cost_per_test": "2.703862", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 8.892, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-6/dfe66cc6-999a-4e72-8c0f-a210c3615815.json b/data/vals_ai/anthropic/claude-sonnet-4-6/dfe66cc6-999a-4e72-8c0f-a210c3615815.json new file mode 100644 index 0000000000000000000000000000000000000000..33d1392aabba318b751f710b2cae02455ba07fef --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-6/dfe66cc6-999a-4e72-8c0f-a210c3615815.json @@ -0,0 +1,564 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_multimodal_index/anthropic_claude-sonnet-4-6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Multimodal Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-6", + "id": "anthropic/claude-sonnet-4-6", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-6", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_multimodal_index:corp_fin_v2:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.vals_multimodal_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Multimodal Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (CorpFin v2).", + "metric_id": "vals_ai.vals_multimodal_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "66.667", + "raw_stderr": "1.609", + "latency": "20.037", + "cost_per_test": "0.287644", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.609, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:finance_agent:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.vals_multimodal_index.finance_agent", + "source_data": { + "dataset_name": "Vals Multimodal Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_multimodal_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 51.033, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "51.033", + "raw_stderr": "1.076", + "latency": "629.1", + "cost_per_test": "2.284299", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.076, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:mortgage_tax:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.vals_multimodal_index.mortgage_tax", + "source_data": { + "dataset_name": "Vals Multimodal Index - Mortgage Tax", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "mortgage_tax", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Mortgage Tax).", + "metric_id": "vals_ai.vals_multimodal_index.mortgage_tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 67.726, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "mortgage_tax", + "task_name": "Mortgage Tax", + "dataset_type": "private", + "industry": "index", + "raw_score": "67.726", + "raw_stderr": "0.921", + "latency": "29.74", + "cost_per_test": "0.04467", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.921, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:overall:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.vals_multimodal_index.overall", + "source_data": { + "dataset_name": "Vals Multimodal Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Overall).", + "metric_id": "vals_ai.vals_multimodal_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 60.571, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "60.571", + "raw_stderr": "1.385", + "latency": "536.038", + "cost_per_test": "1.576917", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.385, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:sage:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.vals_multimodal_index.sage", + "source_data": { + "dataset_name": "Vals Multimodal Index - SAGE", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "sage", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SAGE).", + "metric_id": "vals_ai.vals_multimodal_index.sage.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 46.582, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "sage", + "task_name": "SAGE", + "dataset_type": "private", + "industry": "index", + "raw_score": "46.582", + "raw_stderr": "3.43", + "latency": "163.365", + "cost_per_test": "0.234647", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.43, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:swebench:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.vals_multimodal_index.swebench", + "source_data": { + "dataset_name": "Vals Multimodal Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SWE-bench).", + "metric_id": "vals_ai.vals_multimodal_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 77.451, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "77.451", + "raw_stderr": "4.138", + "latency": "633.682", + "cost_per_test": "1.706873" + }, + "uncertainty": { + "standard_error": { + "value": 4.138, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:terminal_bench_2_1:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.vals_multimodal_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Multimodal Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_multimodal_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 57.303, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "57.303", + "raw_stderr": "0.649", + "latency": "704.225", + "cost_per_test": "0.572392", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "false", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.649, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "false", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:vibe_code_bench:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.vals_multimodal_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Multimodal Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_multimodal_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 56.223, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "56.223", + "raw_stderr": "6.546", + "latency": "1572.118", + "cost_per_test": "5.907892", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.546, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-6/e88ff6ef-d3aa-4b9c-a95d-8934ad010ff6.json b/data/vals_ai/anthropic/claude-sonnet-4-6/e88ff6ef-d3aa-4b9c-a95d-8934ad010ff6.json new file mode 100644 index 0000000000000000000000000000000000000000..b0e7569845b6c80f30a27296cf774f771b5c2028 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-6/e88ff6ef-d3aa-4b9c-a95d-8934ad010ff6.json @@ -0,0 +1,716 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/emb/anthropic_claude-sonnet-4-6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - EMB", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-6", + "id": "anthropic/claude-sonnet-4-6", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-6", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "emb:comps:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.emb.comps", + "source_data": { + "dataset_name": "EMB - Comps", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "comps", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Comps).", + "metric_id": "vals_ai.emb.comps.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 61.52, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "comps", + "task_name": "Comps", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.52", + "raw_stderr": "5.973", + "latency": "3683.558", + "cost_per_test": "8.994714", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.973, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:dataroom:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.emb.dataroom", + "source_data": { + "dataset_name": "EMB - Dataroom", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dataroom", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Dataroom).", + "metric_id": "vals_ai.emb.dataroom.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 78.493, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dataroom", + "task_name": "Dataroom", + "dataset_type": "private", + "industry": "finance", + "raw_score": "78.493", + "raw_stderr": "6.337", + "latency": "2415.188", + "cost_per_test": "5.102641", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.337, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:dcf:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.emb.dcf", + "source_data": { + "dataset_name": "EMB - DCF", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dcf", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (DCF).", + "metric_id": "vals_ai.emb.dcf.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 38.28, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dcf", + "task_name": "DCF", + "dataset_type": "private", + "industry": "finance", + "raw_score": "38.28", + "raw_stderr": "9.702", + "latency": "3293.001", + "cost_per_test": "7.355698", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 9.702, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:lbo:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.emb.lbo", + "source_data": { + "dataset_name": "EMB - LBO", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "lbo", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (LBO).", + "metric_id": "vals_ai.emb.lbo.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 52.536, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "lbo", + "task_name": "LBO", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.536", + "raw_stderr": "5.427", + "latency": "4347.368", + "cost_per_test": "11.216346", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.427, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:ma:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.emb.ma", + "source_data": { + "dataset_name": "EMB - M&A", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "ma", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (M&A).", + "metric_id": "vals_ai.emb.ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 70.503, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "ma", + "task_name": "M&A", + "dataset_type": "private", + "industry": "finance", + "raw_score": "70.503", + "raw_stderr": "6.534", + "latency": "2932.982", + "cost_per_test": "5.383689", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.534, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:misc:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.emb.misc", + "source_data": { + "dataset_name": "EMB - Misc", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "misc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Misc).", + "metric_id": "vals_ai.emb.misc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 61.691, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "misc", + "task_name": "Misc", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.691", + "raw_stderr": "7.18", + "latency": "3079.529", + "cost_per_test": "6.179184", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 7.18, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:operating:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.emb.operating", + "source_data": { + "dataset_name": "EMB - Operating", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "operating", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Operating).", + "metric_id": "vals_ai.emb.operating.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 58.056, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "operating", + "task_name": "Operating", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.056", + "raw_stderr": "7.22", + "latency": "2854.463", + "cost_per_test": "5.620452", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 7.22, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:overall:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.emb.overall", + "source_data": { + "dataset_name": "EMB - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Overall).", + "metric_id": "vals_ai.emb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 60.154, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.154", + "raw_stderr": "3.276", + "latency": "3255.853", + "cost_per_test": "7.22787", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.276, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:scratch:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.emb.scratch", + "source_data": { + "dataset_name": "EMB - Scratch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "scratch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Scratch).", + "metric_id": "vals_ai.emb.scratch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 64.751, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "scratch", + "task_name": "Scratch", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.751", + "raw_stderr": "2.791", + "latency": "3392.83", + "cost_per_test": "7.603845", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.791, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:template:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.emb.template", + "source_data": { + "dataset_name": "EMB - Template", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "template", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Template).", + "metric_id": "vals_ai.emb.template.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 55.17, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "template", + "task_name": "Template", + "dataset_type": "private", + "industry": "finance", + "raw_score": "55.17", + "raw_stderr": "4.807", + "latency": "3118.875", + "cost_per_test": "6.851896", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 4.807, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-4-6/ea89bc1c-b60e-4072-a865-c3df171f17cf.json b/data/vals_ai/anthropic/claude-sonnet-4-6/ea89bc1c-b60e-4072-a865-c3df171f17cf.json new file mode 100644 index 0000000000000000000000000000000000000000..60d4d35289d3416150f0019492b7f5f6fbbc64f5 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-4-6/ea89bc1c-b60e-4072-a865-c3df171f17cf.json @@ -0,0 +1,240 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/anthropic_claude-sonnet-4-6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-4-6", + "id": "anthropic/claude-sonnet-4-6", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-4-6", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 65.822, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.822", + "raw_stderr": "1.356", + "latency": "127.879", + "cost_per_test": "0.127973", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.356, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 77.106, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "77.106", + "raw_stderr": "0.818", + "latency": "127.879", + "cost_per_test": "0.127973", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.818, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:anthropic/claude-sonnet-4-6:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 88.389, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "88.389", + "raw_stderr": "0.916", + "latency": "127.879", + "cost_per_test": "0.127973", + "temperature": "1", + "max_output_tokens": "128000", + "reasoning": "true", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.916, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "true", + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-5/1149a07e-0df9-41f0-bc31-d2f6e649db3a.json b/data/vals_ai/anthropic/claude-sonnet-5/1149a07e-0df9-41f0-bc31-d2f6e649db3a.json new file mode 100644 index 0000000000000000000000000000000000000000..785564303a852c9aa3bb4501f5e6633eed347511 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-5/1149a07e-0df9-41f0-bc31-d2f6e649db3a.json @@ -0,0 +1,3486 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/hlab/anthropic_claude-sonnet-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Harvey's Legal Agent Benchmark", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-5", + "id": "anthropic/claude-sonnet-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "hlab:antitrust_competition:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "2025.345", + "cost_per_test": "10.26734", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:arbitration_international_dispute_resolution:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1998.75", + "cost_per_test": "17.475537", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:banking_finance:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1658.062", + "cost_per_test": "12.054357", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:bankruptcy_restructuring:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "3302.313", + "cost_per_test": "17.441412", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:capital_markets:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1292.072", + "cost_per_test": "7.692702", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:corporate_governance:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1958.872", + "cost_per_test": "16.332451", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:corporate_ma:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "2809.292", + "cost_per_test": "10.977606", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Criteria Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Criteria Pass Rate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 86.279, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate", + "task_name": "Criteria Pass Rate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.279", + "raw_stderr": "0.414", + "latency": "2331.654", + "cost_per_test": "13.430789", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.414, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__antitrust_competition:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 88.768, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.768", + "raw_stderr": "1.904", + "latency": "2025.345", + "cost_per_test": "10.26734", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.904, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__arbitration_international_dispute_resolution:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 87.455, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.455", + "raw_stderr": "1.987", + "latency": "1998.75", + "cost_per_test": "17.475537", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.987, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__banking_finance:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 89.963, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.963", + "raw_stderr": "1.836", + "latency": "1658.062", + "cost_per_test": "12.054357", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.836, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__bankruptcy_restructuring:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 83.737, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.737", + "raw_stderr": "2.175", + "latency": "3302.313", + "cost_per_test": "17.441412", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.175, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__capital_markets:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 89.02, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.02", + "raw_stderr": "1.962", + "latency": "1292.072", + "cost_per_test": "7.692702", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.962, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_governance:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 89.384, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.384", + "raw_stderr": "1.806", + "latency": "1958.872", + "cost_per_test": "16.332451", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.806, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_ma:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 88.462, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.462", + "raw_stderr": "1.812", + "latency": "2809.292", + "cost_per_test": "10.977606", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.812, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__data_privacy_cybersecurity:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 91.166, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "91.166", + "raw_stderr": "1.69", + "latency": "1857.569", + "cost_per_test": "14.094619", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.69, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__emerging_companies_venture_capital:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 85.666, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.666", + "raw_stderr": "2.051", + "latency": "3176.126", + "cost_per_test": "14.534865", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.051, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__employment_labor:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 82.022, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.022", + "raw_stderr": "2.354", + "latency": "1928.644", + "cost_per_test": "10.712721", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.354, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__energy_natural_resources:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 91.419, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "91.419", + "raw_stderr": "1.612", + "latency": "2085.857", + "cost_per_test": "20.449006", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.612, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__environmental_esg:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 83.871, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.871", + "raw_stderr": "2.206", + "latency": "2817.213", + "cost_per_test": "13.131835", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.206, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__funds_asset_management:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 83.382, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.382", + "raw_stderr": "2.013", + "latency": "2407.939", + "cost_per_test": "15.589852", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.013, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__healthcare_life_sciences:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 85.478, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.478", + "raw_stderr": "2.027", + "latency": "2340.064", + "cost_per_test": "12.925225", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.027, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__immigration:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 86.909, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.909", + "raw_stderr": "2.038", + "latency": "1971.076", + "cost_per_test": "12.122947", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.038, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__insurance:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 86.813, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.813", + "raw_stderr": "2.052", + "latency": "2865.775", + "cost_per_test": "10.748512", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.052, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__intellectual_property:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 88.693, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.693", + "raw_stderr": "1.886", + "latency": "3064.789", + "cost_per_test": "17.35086", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.886, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__international_trade_sanctions:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 85.318, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.318", + "raw_stderr": "2.234", + "latency": "2010.352", + "cost_per_test": "11.197071", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.234, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__litigation_dispute_resolution:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 85.768, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.768", + "raw_stderr": "2.142", + "latency": "1915.001", + "cost_per_test": "12.721548", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.142, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__real_estate:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 86.88, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.88", + "raw_stderr": "1.826", + "latency": "2920.547", + "cost_per_test": "15.69821", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.826, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__structured_finance_securitization:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 78.095, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.095", + "raw_stderr": "2.334", + "latency": "1378.036", + "cost_per_test": "8.975475", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.334, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__tax:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 82.274, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.274", + "raw_stderr": "2.212", + "latency": "2140.875", + "cost_per_test": "16.048557", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.212, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__trusts_estates_private_client:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 87.023, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.023", + "raw_stderr": "2.08", + "latency": "3221.687", + "cost_per_test": "14.531727", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.08, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__white_collar_defense_investigations:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 84.512, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.512", + "raw_stderr": "2.103", + "latency": "2813.449", + "cost_per_test": "9.264488", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.103, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:data_privacy_cybersecurity:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1857.569", + "cost_per_test": "14.094619", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:emerging_companies_venture_capital:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "3176.126", + "cost_per_test": "14.534865", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:employment_labor:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1928.644", + "cost_per_test": "10.712721", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:energy_natural_resources:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "2085.857", + "cost_per_test": "20.449006", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:environmental_esg:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "2817.213", + "cost_per_test": "13.131835", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:funds_asset_management:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "2407.939", + "cost_per_test": "15.589852", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:healthcare_life_sciences:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "2340.064", + "cost_per_test": "12.925225", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:immigration:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1971.076", + "cost_per_test": "12.122947", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:insurance:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "2865.775", + "cost_per_test": "10.748512", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:intellectual_property:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "3064.789", + "cost_per_test": "17.35086", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:international_trade_sanctions:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "2010.352", + "cost_per_test": "11.197071", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:litigation_dispute_resolution:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1915.001", + "cost_per_test": "12.721548", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:overall:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.overall", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Overall).", + "metric_id": "vals_ai.hlab.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 5.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "5", + "raw_stderr": "1.646", + "latency": "2331.654", + "cost_per_test": "13.430789", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.646, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:real_estate:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "2920.547", + "cost_per_test": "15.69821", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:structured_finance_securitization:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1378.036", + "cost_per_test": "8.975475", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:tax:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "2140.875", + "cost_per_test": "16.048557", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:trusts_estates_private_client:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "3221.687", + "cost_per_test": "14.531727", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:white_collar_defense_investigations:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.hlab.white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "2813.449", + "cost_per_test": "9.264488", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-5/1848bc4e-5c0f-430a-a309-45f8ab1ca0f7.json b/data/vals_ai/anthropic/claude-sonnet-5/1848bc4e-5c0f-430a-a309-45f8ab1ca0f7.json new file mode 100644 index 0000000000000000000000000000000000000000..759a9ee1dedff3b6d70775d4ac0a7ca7d8e2d360 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-5/1848bc4e-5c0f-430a-a309-45f8ab1ca0f7.json @@ -0,0 +1,696 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/emb/anthropic_claude-sonnet-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - EMB", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-5", + "id": "anthropic/claude-sonnet-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "emb:comps:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.emb.comps", + "source_data": { + "dataset_name": "EMB - Comps", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "comps", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Comps).", + "metric_id": "vals_ai.emb.comps.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 69.384, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "comps", + "task_name": "Comps", + "dataset_type": "private", + "industry": "finance", + "raw_score": "69.384", + "raw_stderr": "6.423", + "latency": "3392.345", + "cost_per_test": "19.384092", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.423, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:dataroom:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.emb.dataroom", + "source_data": { + "dataset_name": "EMB - Dataroom", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dataroom", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Dataroom).", + "metric_id": "vals_ai.emb.dataroom.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 84.909, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dataroom", + "task_name": "Dataroom", + "dataset_type": "private", + "industry": "finance", + "raw_score": "84.909", + "raw_stderr": "4.74", + "latency": "2740.679", + "cost_per_test": "12.285703", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 4.74, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:dcf:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.emb.dcf", + "source_data": { + "dataset_name": "EMB - DCF", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dcf", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (DCF).", + "metric_id": "vals_ai.emb.dcf.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 56.404, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dcf", + "task_name": "DCF", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.404", + "raw_stderr": "8.129", + "latency": "2886.854", + "cost_per_test": "14.534013", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 8.129, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:lbo:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.emb.lbo", + "source_data": { + "dataset_name": "EMB - LBO", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "lbo", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (LBO).", + "metric_id": "vals_ai.emb.lbo.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 59.967, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "lbo", + "task_name": "LBO", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.967", + "raw_stderr": "6.146", + "latency": "4048.244", + "cost_per_test": "23.03129", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.146, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:ma:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.emb.ma", + "source_data": { + "dataset_name": "EMB - M&A", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "ma", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (M&A).", + "metric_id": "vals_ai.emb.ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 75.832, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "ma", + "task_name": "M&A", + "dataset_type": "private", + "industry": "finance", + "raw_score": "75.832", + "raw_stderr": "4.698", + "latency": "2458.641", + "cost_per_test": "12.506382", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 4.698, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:misc:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.emb.misc", + "source_data": { + "dataset_name": "EMB - Misc", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "misc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Misc).", + "metric_id": "vals_ai.emb.misc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 52.26, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "misc", + "task_name": "Misc", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.26", + "raw_stderr": "9.595", + "latency": "2692.471", + "cost_per_test": "12.389921", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 9.595, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:operating:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.emb.operating", + "source_data": { + "dataset_name": "EMB - Operating", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "operating", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Operating).", + "metric_id": "vals_ai.emb.operating.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 65.455, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "operating", + "task_name": "Operating", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.455", + "raw_stderr": "6.282", + "latency": "2461.863", + "cost_per_test": "11.934898", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.282, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:overall:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.emb.overall", + "source_data": { + "dataset_name": "EMB - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Overall).", + "metric_id": "vals_ai.emb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 66.316, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.316", + "raw_stderr": "3.006", + "latency": "2985.704", + "cost_per_test": "15.440691", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.006, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:scratch:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.emb.scratch", + "source_data": { + "dataset_name": "EMB - Scratch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "scratch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Scratch).", + "metric_id": "vals_ai.emb.scratch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 65.037, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "scratch", + "task_name": "Scratch", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.037", + "raw_stderr": "3.212", + "latency": "3133.342", + "cost_per_test": "15.71521", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.212, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:template:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.emb.template", + "source_data": { + "dataset_name": "EMB - Template", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "template", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Template).", + "metric_id": "vals_ai.emb.template.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 67.983, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "template", + "task_name": "Template", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.983", + "raw_stderr": "4.194", + "latency": "2843.855", + "cost_per_test": "15.176937", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 4.194, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-5/2dbe37a5-f6e0-4ab5-95ce-ed31edc34205.json b/data/vals_ai/anthropic/claude-sonnet-5/2dbe37a5-f6e0-4ab5-95ce-ed31edc34205.json new file mode 100644 index 0000000000000000000000000000000000000000..cf38ac4b3dec8ae57384dfaa5e735dd5903eff44 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-5/2dbe37a5-f6e0-4ab5-95ce-ed31edc34205.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/anthropic_claude-sonnet-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-5", + "id": "anthropic/claude-sonnet-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 64.023, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.023", + "raw_stderr": "1.372", + "latency": "202.488", + "cost_per_test": "0.301619", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.372, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 75.634, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "75.634", + "raw_stderr": "0.836", + "latency": "202.488", + "cost_per_test": "0.301619", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.836, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 87.244, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "87.244", + "raw_stderr": "0.954", + "latency": "202.488", + "cost_per_test": "0.301619", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.954, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-5/2f27e9fa-a4ba-44e8-ad6d-95d361c7bf52.json b/data/vals_ai/anthropic/claude-sonnet-5/2f27e9fa-a4ba-44e8-ad6d-95d361c7bf52.json new file mode 100644 index 0000000000000000000000000000000000000000..b17bd29abc899365984356526920c393d1d5efe0 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-5/2f27e9fa-a4ba-44e8-ad6d-95d361c7bf52.json @@ -0,0 +1,1071 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/anthropic_claude-sonnet-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-5", + "id": "anthropic/claude-sonnet-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.584, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.584", + "raw_stderr": "1.443", + "latency": "33.745", + "cost_per_test": "0.05724", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.443, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.635, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.635", + "raw_stderr": "1.393", + "latency": "46.752", + "cost_per_test": "0.078843", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.393, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.813, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.813", + "raw_stderr": "1.062", + "latency": "9.099", + "cost_per_test": "0.022172", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.062, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.024, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.024", + "raw_stderr": "1.559", + "latency": "19.153", + "cost_per_test": "0.016148", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.559, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.455, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.455", + "raw_stderr": "1.189", + "latency": "27.727", + "cost_per_test": "0.045893", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.189, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.631, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.631", + "raw_stderr": "1.474", + "latency": "19.936", + "cost_per_test": "0.037502", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.474, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.274, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.274", + "raw_stderr": "1.692", + "latency": "40.906", + "cost_per_test": "0.065913", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.692, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.315, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.315", + "raw_stderr": "2.077", + "latency": "24.241", + "cost_per_test": "0.046377", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.077, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.747, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.747", + "raw_stderr": "1.266", + "latency": "6.521", + "cost_per_test": "0.015722", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.266, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 95.929, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "95.929", + "raw_stderr": "0.685", + "latency": "36.687", + "cost_per_test": "0.068593", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.685, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.39, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.39", + "raw_stderr": "1.22", + "latency": "26.94", + "cost_per_test": "0.043786", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.22, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.545, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.545", + "raw_stderr": "0.368", + "latency": "25.191", + "cost_per_test": "0.042996", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.368, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.573, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.573", + "raw_stderr": "1.536", + "latency": "25.082", + "cost_per_test": "0.039802", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.536, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.532, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.532", + "raw_stderr": "0.871", + "latency": "8.615", + "cost_per_test": "0.018511", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.871, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.722, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.722", + "raw_stderr": "1.239", + "latency": "27.27", + "cost_per_test": "0.045438", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.239, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-5/34dedaab-b271-4c97-93f8-3d2a72720ebf.json b/data/vals_ai/anthropic/claude-sonnet-5/34dedaab-b271-4c97-93f8-3d2a72720ebf.json new file mode 100644 index 0000000000000000000000000000000000000000..797fa53f1a7fe8a8d927dbfdeccf84f14ff216aa --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-5/34dedaab-b271-4c97-93f8-3d2a72720ebf.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/anthropic_claude-sonnet-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-5", + "id": "anthropic/claude-sonnet-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 81.325, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "81.325", + "raw_stderr": "3.046", + "latency": "4001.424", + "cost_per_test": "38.07813", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.046, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-5/4ac19c39-a06a-4990-8ef6-f46ae6fac41b.json b/data/vals_ai/anthropic/claude-sonnet-5/4ac19c39-a06a-4990-8ef6-f46ae6fac41b.json new file mode 100644 index 0000000000000000000000000000000000000000..a2b95434b47276cb6e4d0751052b082ccc680836 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-5/4ac19c39-a06a-4990-8ef6-f46ae6fac41b.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/anthropic_claude-sonnet-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-5", + "id": "anthropic/claude-sonnet-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 66.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "66", + "raw_stderr": "4.761", + "latency": "1014.587", + "cost_per_test": "2.248759", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 4.761, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-5/4c4ca75d-32f7-41e4-a85d-03fddfc2e5fd.json b/data/vals_ai/anthropic/claude-sonnet-5/4c4ca75d-32f7-41e4-a85d-03fddfc2e5fd.json new file mode 100644 index 0000000000000000000000000000000000000000..e05dd5d36f09e64763ddfe3d64a5080c3bf6cdc7 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-5/4c4ca75d-32f7-41e4-a85d-03fddfc2e5fd.json @@ -0,0 +1,550 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_multimodal_index/anthropic_claude-sonnet-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Multimodal Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-5", + "id": "anthropic/claude-sonnet-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_multimodal_index:corp_fin_v2:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Multimodal Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (CorpFin v2).", + "metric_id": "vals_ai.vals_multimodal_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 67.949, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "67.949", + "raw_stderr": "1.593", + "latency": "20.457", + "cost_per_test": "0.416108", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.593, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:finance_agent:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.finance_agent", + "source_data": { + "dataset_name": "Vals Multimodal Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_multimodal_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 51.979, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "51.979", + "raw_stderr": "0.534", + "latency": "867.393", + "cost_per_test": "4.021924", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.534, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:mortgage_tax:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.mortgage_tax", + "source_data": { + "dataset_name": "Vals Multimodal Index - Mortgage Tax", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "mortgage_tax", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Mortgage Tax).", + "metric_id": "vals_ai.vals_multimodal_index.mortgage_tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 70.032, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "mortgage_tax", + "task_name": "Mortgage Tax", + "dataset_type": "private", + "industry": "index", + "raw_score": "70.032", + "raw_stderr": "0.897", + "latency": "28.668", + "cost_per_test": "0.080204", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.897, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:overall:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.overall", + "source_data": { + "dataset_name": "Vals Multimodal Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Overall).", + "metric_id": "vals_ai.vals_multimodal_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 68.826, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "68.826", + "raw_stderr": "0.909", + "latency": "1012.043", + "cost_per_test": "6.532939", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.909, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:sage:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.sage", + "source_data": { + "dataset_name": "Vals Multimodal Index - SAGE", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "sage", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SAGE).", + "metric_id": "vals_ai.vals_multimodal_index.sage.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 48.925, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "sage", + "task_name": "SAGE", + "dataset_type": "private", + "industry": "index", + "raw_score": "48.925", + "raw_stderr": "3.405", + "latency": "434.2", + "cost_per_test": "0.588528", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.405, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:swebench:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.swebench", + "source_data": { + "dataset_name": "Vals Multimodal Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SWE-bench).", + "metric_id": "vals_ai.vals_multimodal_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 75.49, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "75.49", + "raw_stderr": "4.259", + "latency": "1097.115", + "cost_per_test": "1.749654" + }, + "uncertainty": { + "standard_error": { + "value": 4.259, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:terminal_bench_2_1:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Multimodal Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_multimodal_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 74.532, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "74.532", + "raw_stderr": "2.085", + "latency": "635.043", + "cost_per_test": "0.796027", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.085, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:vibe_code_bench:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Multimodal Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_multimodal_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 86.902, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "86.902", + "raw_stderr": "3.468", + "latency": "4001.424", + "cost_per_test": "38.07813", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.468, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-5/520c4d66-0d54-4dbf-9877-d1c2b4981fc6.json b/data/vals_ai/anthropic/claude-sonnet-5/520c4d66-0d54-4dbf-9877-d1c2b4981fc6.json new file mode 100644 index 0000000000000000000000000000000000000000..2ca9fa0bd23969d0a32dd451f067d37863f70e30 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-5/520c4d66-0d54-4dbf-9877-d1c2b4981fc6.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/anthropic_claude-sonnet-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-5", + "id": "anthropic/claude-sonnet-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 63.869, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.869", + "raw_stderr": "1.64", + "latency": "12.323", + "cost_per_test": "0.027074", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.64, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 69.114, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "69.114", + "raw_stderr": "1.577", + "latency": "23.105", + "cost_per_test": "1.105176", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.577, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 66.977, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.977", + "raw_stderr": "0.926", + "latency": "18.628", + "cost_per_test": "0.516119", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.926, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 67.949, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.949", + "raw_stderr": "1.593", + "latency": "20.457", + "cost_per_test": "0.416108", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.593, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-5/6dbc783b-fa84-4a0e-ab50-bc2a5f45df26.json b/data/vals_ai/anthropic/claude-sonnet-5/6dbc783b-fa84-4a0e-ab50-bc2a5f45df26.json new file mode 100644 index 0000000000000000000000000000000000000000..57d698467b9a1380aa0fea46cfbd0fa920e79a30 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-5/6dbc783b-fa84-4a0e-ab50-bc2a5f45df26.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/anthropic_claude-sonnet-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-5", + "id": "anthropic/claude-sonnet-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 100.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "100", + "raw_stderr": "0", + "latency": "14129.714", + "cost_per_test": "17.711595", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 57.778, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "57.778", + "raw_stderr": "2.222", + "latency": "1883.962", + "cost_per_test": "2.361546", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.222, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 81.818, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "81.818", + "raw_stderr": "2.777", + "latency": "1027.616", + "cost_per_test": "1.288116", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.777, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 74.532, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "74.532", + "raw_stderr": "2.085", + "latency": "635.043", + "cost_per_test": "0.796027", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.085, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-5/80d6b45d-f59b-488f-bc8f-9e9815f4f3d0.json b/data/vals_ai/anthropic/claude-sonnet-5/80d6b45d-f59b-488f-bc8f-9e9815f4f3d0.json new file mode 100644 index 0000000000000000000000000000000000000000..192719fe6ad6bbafb19ae90dab05658b2012c6de --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-5/80d6b45d-f59b-488f-bc8f-9e9815f4f3d0.json @@ -0,0 +1,450 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/anthropic_claude-sonnet-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-5", + "id": "anthropic/claude-sonnet-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 89.555, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.555", + "raw_stderr": "0.753", + "latency": "9.309", + "cost_per_test": "0.014349", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.753, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.412, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.412", + "raw_stderr": "0.205", + "latency": "2.575", + "cost_per_test": "0.004968", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.205, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 88.24, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.24", + "raw_stderr": "0.609", + "latency": "2.196", + "cost_per_test": "0.00861", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.609, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.922, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.922", + "raw_stderr": "0.458", + "latency": "4.901", + "cost_per_test": "0.008639", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.458, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.07, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.07", + "raw_stderr": "1.105", + "latency": "2.558", + "cost_per_test": "0.005061", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.105, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.331, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.331", + "raw_stderr": "1.742", + "latency": "7.866", + "cost_per_test": "0.010206", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.742, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-5/8373342c-684e-481c-86a3-2803a89f5c72.json b/data/vals_ai/anthropic/claude-sonnet-5/8373342c-684e-481c-86a3-2803a89f5c72.json new file mode 100644 index 0000000000000000000000000000000000000000..01eb4acd8c2c125a13bd745fcbf8a3049aa6ce8d --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-5/8373342c-684e-481c-86a3-2803a89f5c72.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/public-benefits-bench/anthropic_claude-sonnet-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Public Benefits Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "social mobility", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-5", + "id": "anthropic/claude-sonnet-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "public-benefits-bench:overall:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.public-benefits-bench.overall", + "source_data": { + "dataset_name": "Public Benefits Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Public Benefits Bench v1.1 (Overall).", + "metric_id": "vals_ai.public-benefits-bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "score_details": { + "score": 66.035, + "details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "social mobility", + "raw_score": "66.035", + "raw_stderr": "1.232", + "latency": "1460.435", + "cost_per_test": "1.936042", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.232, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-5/97a6dc7d-c273-4949-be6a-2a20e5e59c03.json b/data/vals_ai/anthropic/claude-sonnet-5/97a6dc7d-c273-4949-be6a-2a20e5e59c03.json new file mode 100644 index 0000000000000000000000000000000000000000..820f9c0338a51b5db687cc0df2f37872d874cfa4 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-5/97a6dc7d-c273-4949-be6a-2a20e5e59c03.json @@ -0,0 +1,762 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/anthropic_claude-sonnet-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-5", + "id": "anthropic/claude-sonnet-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 48.715, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "48.715", + "raw_stderr": "0.375", + "latency": "786.124", + "cost_per_test": "3.530087", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.375, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 42.084, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "42.084", + "raw_stderr": "0.357", + "latency": "792.086", + "cost_per_test": "1.126139", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.357, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 45.817, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "45.817", + "raw_stderr": "1.039", + "latency": "1108.045", + "cost_per_test": "4.636983", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.039, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 62.076, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.076", + "raw_stderr": "3.446", + "latency": "832.196", + "cost_per_test": "5.544368", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.446, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 72.588, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "72.588", + "raw_stderr": "0.589", + "latency": "434.462", + "cost_per_test": "1.072096", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.589, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 24.197, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "24.197", + "raw_stderr": "1.167", + "latency": "709.547", + "cost_per_test": "3.457935", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.167, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 64.67, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.67", + "raw_stderr": "1.631", + "latency": "1344.312", + "cost_per_test": "5.523658", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.631, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 74.268, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "74.268", + "raw_stderr": "1.12", + "latency": "847.356", + "cost_per_test": "4.089571", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.12, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 64.075, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.075", + "raw_stderr": "1.565", + "latency": "753.744", + "cost_per_test": "1.779045", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.565, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:overall:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 53.909, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "53.909", + "raw_stderr": "0.518", + "latency": "792.086", + "cost_per_test": "1.126139", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.518, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 28.774, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "28.774", + "raw_stderr": "1.863", + "latency": "1101.385", + "cost_per_test": "6.563577", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.863, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-5/9dadb888-2664-4f9a-85b8-d605d04d4e1c.json b/data/vals_ai/anthropic/claude-sonnet-5/9dadb888-2664-4f9a-85b8-d605d04d4e1c.json new file mode 100644 index 0000000000000000000000000000000000000000..059e3d2f45af22a2bec1f88489e25e49a05df99f --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-5/9dadb888-2664-4f9a-85b8-d605d04d4e1c.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/anthropic_claude-sonnet-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-5", + "id": "anthropic/claude-sonnet-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 78.696, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "78.696", + "raw_stderr": "1.154", + "latency": "47.997", + "cost_per_test": "0.110772", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.154, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 70.032, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "70.032", + "raw_stderr": "0.897", + "latency": "28.668", + "cost_per_test": "0.080204", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.897, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 61.367, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.367", + "raw_stderr": "1.373", + "latency": "9.339", + "cost_per_test": "0.049635", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.373, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-5/9e97c112-a04b-4f40-859e-235ab6f190e4.json b/data/vals_ai/anthropic/claude-sonnet-5/9e97c112-a04b-4f40-859e-235ab6f190e4.json new file mode 100644 index 0000000000000000000000000000000000000000..fe82783cdb4eb6ae99039fa87435d14369d40847 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-5/9e97c112-a04b-4f40-859e-235ab6f190e4.json @@ -0,0 +1,105 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/anthropic_claude-sonnet-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-5", + "id": "anthropic/claude-sonnet-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 83.006, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.006", + "raw_stderr": "0.903", + "latency": "18.817", + "cost_per_test": "0.033133", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.903, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-5/a15be168-0134-4ff3-9dbc-bc991bc4871d.json b/data/vals_ai/anthropic/claude-sonnet-5/a15be168-0134-4ff3-9dbc-bc991bc4871d.json new file mode 100644 index 0000000000000000000000000000000000000000..3d9ea328247f019cadd4e170d80fe07b61c53437 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-5/a15be168-0134-4ff3-9dbc-bc991bc4871d.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/anthropic_claude-sonnet-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-5", + "id": "anthropic/claude-sonnet-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 76.054, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "76.054", + "raw_stderr": "3.05", + "latency": "251.703", + "cost_per_test": "0.433684", + "temperature": "1", + "max_output_tokens": "30000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.05, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-5/b7e0802d-b546-4a47-b9ae-bc2801b0267b.json b/data/vals_ai/anthropic/claude-sonnet-5/b7e0802d-b546-4a47-b9ae-bc2801b0267b.json new file mode 100644 index 0000000000000000000000000000000000000000..8767f0120cce8b26e4a965e92f6239bdaa49505c --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-5/b7e0802d-b546-4a47-b9ae-bc2801b0267b.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/anthropic_claude-sonnet-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-5", + "id": "anthropic/claude-sonnet-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.447, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.447", + "raw_stderr": "0.689", + "latency": "16.815", + "cost_per_test": "0.022051", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.689, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "lcb:hard:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 66.857, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66.857", + "raw_stderr": "2.516", + "latency": "180.069", + "cost_per_test": "0.223535", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.516, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "lcb:medium:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 81.984, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "81.984", + "raw_stderr": "1.964", + "latency": "34.18", + "cost_per_test": "0.03838", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.964, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "lcb:overall:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 82.429, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "82.429", + "raw_stderr": "1.088", + "latency": "77.021", + "cost_per_test": "0.094655", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.088, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-5/b89fce6e-407f-43eb-b8a0-2d141eb3b3cd.json b/data/vals_ai/anthropic/claude-sonnet-5/b89fce6e-407f-43eb-b8a0-2d141eb3b3cd.json new file mode 100644 index 0000000000000000000000000000000000000000..334a64884f4cf0a2b42074792b6386b162a6c238 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-5/b89fce6e-407f-43eb-b8a0-2d141eb3b3cd.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/code-migration/anthropic_claude-sonnet-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Code Migration", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-5", + "id": "anthropic/claude-sonnet-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "code-migration:cli:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.code-migration.cli", + "source_data": { + "dataset_name": "Code Migration - CLI", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cli", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (CLI).", + "metric_id": "vals_ai.code-migration.cli.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 36.332, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cli", + "task_name": "CLI", + "dataset_type": "private", + "industry": "coding", + "raw_score": "36.332", + "raw_stderr": "2.636", + "latency": "8003.993", + "cost_per_test": "63.92939", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.636, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "code-migration:cobol:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.code-migration.cobol", + "source_data": { + "dataset_name": "Code Migration - COBOL", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cobol", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (COBOL).", + "metric_id": "vals_ai.code-migration.cobol.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 68.571, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cobol", + "task_name": "COBOL", + "dataset_type": "private", + "industry": "coding", + "raw_score": "68.571", + "raw_stderr": "15.028", + "latency": "4094.2", + "cost_per_test": "20.074458", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 15.028, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "code-migration:code_quality:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.code-migration.code_quality", + "source_data": { + "dataset_name": "Code Migration - Code Quality", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "code_quality", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Code Quality).", + "metric_id": "vals_ai.code-migration.code_quality.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 4.898, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "code_quality", + "task_name": "Code Quality", + "dataset_type": "private", + "industry": "coding", + "raw_score": "4.898", + "raw_stderr": "0.015", + "latency": "7026.544", + "cost_per_test": "52.965657", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.015, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "code-migration:overall:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.code-migration.overall", + "source_data": { + "dataset_name": "Code Migration - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Overall).", + "metric_id": "vals_ai.code-migration.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 44.392, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "44.392", + "raw_stderr": "4.246", + "latency": "7026.544", + "cost_per_test": "52.965657", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 4.246, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-5/bb04e3a4-f621-4689-a0ef-32fc1ea3f8c8.json b/data/vals_ai/anthropic/claude-sonnet-5/bb04e3a4-f621-4689-a0ef-32fc1ea3f8c8.json new file mode 100644 index 0000000000000000000000000000000000000000..c8f7455f8830362bd770132222a79b159e147ca1 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-5/bb04e3a4-f621-4689-a0ef-32fc1ea3f8c8.json @@ -0,0 +1,308 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/programbench/anthropic_claude-sonnet-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProgramBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-5", + "id": "anthropic/claude-sonnet-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "programbench:almost:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.programbench.almost", + "source_data": { + "dataset_name": "ProgramBench - Almost Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "almost", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Almost Resolved).", + "metric_id": "vals_ai.programbench.almost.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 13.5, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "almost", + "task_name": "Almost Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "13.5", + "raw_stderr": "2.422", + "latency": "5471.639", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.422, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "programbench:overall:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.programbench.overall", + "source_data": { + "dataset_name": "ProgramBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Overall).", + "metric_id": "vals_ai.programbench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "5471.639", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "programbench:partial:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.programbench.partial", + "source_data": { + "dataset_name": "ProgramBench - Raw Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "partial", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Raw Pass Rate).", + "metric_id": "vals_ai.programbench.partial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 72.067, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "partial", + "task_name": "Raw Pass Rate", + "dataset_type": "public", + "industry": "coding", + "raw_score": "72.067", + "raw_stderr": "1.789", + "latency": "5471.639", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.789, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "programbench:strict:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.programbench.strict", + "source_data": { + "dataset_name": "ProgramBench - Fully Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "strict", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Fully Resolved).", + "metric_id": "vals_ai.programbench.strict.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "strict", + "task_name": "Fully Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "5471.639", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-5/cc0c355b-0e6c-41cd-8021-785f314e9ea2.json b/data/vals_ai/anthropic/claude-sonnet-5/cc0c355b-0e6c-41cd-8021-785f314e9ea2.json new file mode 100644 index 0000000000000000000000000000000000000000..9780004ae365dad0b740a562a34a0c78a711dbb0 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-5/cc0c355b-0e6c-41cd-8021-785f314e9ea2.json @@ -0,0 +1,381 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/anthropic_claude-sonnet-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-5", + "id": "anthropic/claude-sonnet-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 76.19, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "76.19", + "raw_stderr": "6.652", + "latency": "1512.9", + "cost_per_test": "3.003622", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.652, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 77.395, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "77.395", + "raw_stderr": "2.594", + "latency": "1051.991", + "cost_per_test": "1.610154", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.594, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 83.505, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "83.505", + "raw_stderr": "2.671", + "latency": "704.586", + "cost_per_test": "0.968627", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.671, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66.667", + "raw_stderr": "33.333", + "latency": "2127.307", + "cost_per_test": "3.544588", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:overall:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 79.6, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "79.6", + "raw_stderr": "1.804", + "latency": "962.366", + "cost_per_test": "1.489899", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.804, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-5/d879bb31-90a4-4b56-9226-fea99b338ed2.json b/data/vals_ai/anthropic/claude-sonnet-5/d879bb31-90a4-4b56-9226-fea99b338ed2.json new file mode 100644 index 0000000000000000000000000000000000000000..5897af693c0c63fa6ef5b232c2980d032a53162b --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-5/d879bb31-90a4-4b56-9226-fea99b338ed2.json @@ -0,0 +1,418 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/anthropic_claude-sonnet-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-5", + "id": "anthropic/claude-sonnet-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 67.949, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "67.949", + "raw_stderr": "1.593", + "latency": "20.457", + "cost_per_test": "0.416108", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.593, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 51.979, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "51.979", + "raw_stderr": "0.534", + "latency": "867.393", + "cost_per_test": "4.021924", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.534, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_index:overall:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 68.608, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "68.608", + "raw_stderr": "0.996", + "latency": "1324.286", + "cost_per_test": "9.012369", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 0.996, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 75.49, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "75.49", + "raw_stderr": "4.259", + "latency": "1097.115", + "cost_per_test": "1.749654" + }, + "uncertainty": { + "standard_error": { + "value": 4.259, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 74.532, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "74.532", + "raw_stderr": "2.085", + "latency": "635.043", + "cost_per_test": "0.796027", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.085, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 86.902, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "86.902", + "raw_stderr": "3.468", + "latency": "4001.424", + "cost_per_test": "38.07813", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.468, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-5/e1e3b34d-9df7-40ef-9f0c-135b25b76b8c.json b/data/vals_ai/anthropic/claude-sonnet-5/e1e3b34d-9df7-40ef-9f0c-135b25b76b8c.json new file mode 100644 index 0000000000000000000000000000000000000000..776890f67990959f0b0f83984e67120a3964537a --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-5/e1e3b34d-9df7-40ef-9f0c-135b25b76b8c.json @@ -0,0 +1,1224 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_research/anthropic_claude-sonnet-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Legal Research Bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-5", + "id": "anthropic/claude-sonnet-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_research:administrative_regulatory:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.legal_research.administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 54.545, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "54.545", + "raw_stderr": "6.176", + "latency": "1285.722", + "cost_per_test": "3.761716", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.176, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:business_commercial:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.legal_research.business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 32.558, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "32.558", + "raw_stderr": "7.231", + "latency": "1790.227", + "cost_per_test": "4.547892", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 7.231, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:civil_litigation:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.legal_research.civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 44.444, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "44.444", + "raw_stderr": "9.745", + "latency": "1956.425", + "cost_per_test": "5.100811", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 9.745, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:constitutional_civil_rights:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.legal_research.constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 44.737, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "44.737", + "raw_stderr": "8.174", + "latency": "1566.47", + "cost_per_test": "3.716455", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 8.174, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:criminal:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.legal_research.criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 40.351, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "40.351", + "raw_stderr": "6.556", + "latency": "1506.644", + "cost_per_test": "3.778014", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.556, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:family:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.legal_research.family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 18.182, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "18.182", + "raw_stderr": "8.417", + "latency": "1564.903", + "cost_per_test": "4.138354", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 8.417, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:health:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.legal_research.health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 63.333, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "63.333", + "raw_stderr": "8.949", + "latency": "1364.3", + "cost_per_test": "3.986073", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 8.949, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:immigration:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.legal_research.immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 36.364, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "36.364", + "raw_stderr": "15.212", + "latency": "1629.748", + "cost_per_test": "4.132244", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 15.212, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:overall:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.legal_research.overall", + "source_data": { + "dataset_name": "Legal Research Bench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Overall).", + "metric_id": "vals_ai.legal_research.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 41.827, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "legal", + "raw_score": "41.827", + "raw_stderr": "3.428", + "latency": "1545.997", + "cost_per_test": "4.079169", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.428, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass", + "source_data": { + "dataset_name": "Legal Research Bench - Weighted Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Weighted Pass).", + "metric_id": "vals_ai.legal_research.weighted_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 84.592, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass", + "task_name": "Weighted Pass", + "dataset_type": "private", + "industry": "legal", + "raw_score": "84.592", + "raw_stderr": "1.431", + "latency": "1545.997", + "cost_per_test": "4.079169", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 1.431, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__administrative_regulatory:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.weighted_pass__administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 87.81, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "87.81", + "raw_stderr": "2.292", + "latency": "1285.722", + "cost_per_test": "3.761716", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.292, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__business_commercial:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.weighted_pass__business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 86.759, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "86.759", + "raw_stderr": "2.83", + "latency": "1790.227", + "cost_per_test": "4.547892", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.83, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__civil_litigation:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.weighted_pass__civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 88.718, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "88.718", + "raw_stderr": "2.99", + "latency": "1956.425", + "cost_per_test": "5.100811", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.99, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__constitutional_civil_rights:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 87.039, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "87.039", + "raw_stderr": "2.548", + "latency": "1566.47", + "cost_per_test": "3.716455", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.548, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__criminal:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.weighted_pass__criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 84.06, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "84.06", + "raw_stderr": "2.665", + "latency": "1506.644", + "cost_per_test": "3.778014", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.665, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__family:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.weighted_pass__family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 72.233, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "72.233", + "raw_stderr": "5.607", + "latency": "1564.903", + "cost_per_test": "4.138354", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.607, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__health:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.weighted_pass__health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 92.146, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "92.146", + "raw_stderr": "2.478", + "latency": "1364.3", + "cost_per_test": "3.986073", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.478, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__immigration:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.weighted_pass__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 79.199, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "79.199", + "raw_stderr": "7.413", + "latency": "1629.748", + "cost_per_test": "4.132244", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 7.413, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-5/ea1b326f-4647-4a67-bb32-a0bebd73eeea.json b/data/vals_ai/anthropic/claude-sonnet-5/ea1b326f-4647-4a67-bb32-a0bebd73eeea.json new file mode 100644 index 0000000000000000000000000000000000000000..48f6a1836cda0c618b8a5ae1d060f304f0b82b69 --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-5/ea1b326f-4647-4a67-bb32-a0bebd73eeea.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/anthropic_claude-sonnet-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-5", + "id": "anthropic/claude-sonnet-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 42.857, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "42.857", + "raw_stderr": "6.216", + "latency": "430.465", + "cost_per_test": "0.567587", + "temperature": "1", + "max_output_tokens": "30000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.216, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "sage:automata:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 34.92, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "34.92", + "raw_stderr": "6.195", + "latency": "677.504", + "cost_per_test": "0.801263", + "temperature": "1", + "max_output_tokens": "30000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 6.195, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 68.999, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "68.999", + "raw_stderr": "5.228", + "latency": "194.632", + "cost_per_test": "0.396734", + "temperature": "1", + "max_output_tokens": "30000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 5.228, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "sage:overall:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 48.925, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "48.925", + "raw_stderr": "3.405", + "latency": "434.2", + "cost_per_test": "0.588528", + "temperature": "1", + "max_output_tokens": "30000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.405, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/anthropic/claude-sonnet-5/eeb829aa-f9a1-4a31-9021-184e35a5a42e.json b/data/vals_ai/anthropic/claude-sonnet-5/eeb829aa-f9a1-4a31-9021-184e35a5a42e.json new file mode 100644 index 0000000000000000000000000000000000000000..dd8d648bb755a4238eac28758cc7cd6f56933c3b --- /dev/null +++ b/data/vals_ai/anthropic/claude-sonnet-5/eeb829aa-f9a1-4a31-9021-184e35a5a42e.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/anthropic_claude-sonnet-5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "claude-sonnet-5", + "id": "anthropic/claude-sonnet-5", + "developer": "anthropic", + "additional_details": { + "vals_model_id": "anthropic/claude-sonnet-5", + "vals_provider": "Anthropic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 87.374, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.374", + "raw_stderr": "3.456", + "latency": "55.38", + "cost_per_test": "0.084904", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 3.456, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 90.404, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.404", + "raw_stderr": "2.799", + "latency": "71.088", + "cost_per_test": "0.070514", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.799, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:anthropic/claude-sonnet-5:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 88.889, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.889", + "raw_stderr": "2.224", + "latency": "63.234", + "cost_per_test": "0.077709", + "temperature": "1", + "max_output_tokens": "128000", + "compute_effort": "max", + "provider": "Anthropic" + }, + "uncertainty": { + "standard_error": { + "value": 2.224, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "compute_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/arcee-ai/trinity-large-thinking/daa23f21-2cd8-4a94-a9be-ed2c885b1ac8.json b/data/vals_ai/arcee-ai/trinity-large-thinking/daa23f21-2cd8-4a94-a9be-ed2c885b1ac8.json new file mode 100644 index 0000000000000000000000000000000000000000..5a85c295cb3ae578bf4afda908067d8aa3b4bb00 --- /dev/null +++ b/data/vals_ai/arcee-ai/trinity-large-thinking/daa23f21-2cd8-4a94-a9be-ed2c885b1ac8.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/arcee-ai_trinity-large-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "trinity-large-thinking", + "id": "arcee-ai/trinity-large-thinking", + "developer": "arcee-ai", + "additional_details": { + "vals_model_id": "arcee-ai/trinity-large-thinking", + "vals_provider": "Arcee-Ai" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:arcee-ai/trinity-large-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.023, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.023", + "raw_stderr": "1.672", + "latency": "9.333", + "cost_per_test": "0.000879", + "temperature": "0.3", + "top_p": "0.95", + "max_output_tokens": "80000", + "provider": "Arcee-Ai" + }, + "uncertainty": { + "standard_error": { + "value": 1.672, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:arcee-ai/trinity-large-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 49.184, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "49.184", + "raw_stderr": "1.707", + "latency": "16.74", + "cost_per_test": "0.041363", + "temperature": "0.3", + "top_p": "0.95", + "max_output_tokens": "80000", + "provider": "Arcee-Ai" + }, + "uncertainty": { + "standard_error": { + "value": 1.707, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:arcee-ai/trinity-large-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 54.662, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "54.662", + "raw_stderr": "0.977", + "latency": "12.482", + "cost_per_test": "0.01939", + "temperature": "0.3", + "top_p": "0.95", + "max_output_tokens": "80000", + "provider": "Arcee-Ai" + }, + "uncertainty": { + "standard_error": { + "value": 0.977, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:arcee-ai/trinity-large-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 54.779, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "54.779", + "raw_stderr": "1.699", + "latency": "11.372", + "cost_per_test": "0.015928", + "temperature": "0.3", + "top_p": "0.95", + "max_output_tokens": "80000", + "provider": "Arcee-Ai" + }, + "uncertainty": { + "standard_error": { + "value": 1.699, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/aristotle/aristotle/6f81e329-9aa6-418d-8b63-d7f82c5bf10b.json b/data/vals_ai/aristotle/aristotle/6f81e329-9aa6-418d-8b63-d7f82c5bf10b.json new file mode 100644 index 0000000000000000000000000000000000000000..0b66668eb48f462680665fc8391abb6ce0a01fcc --- /dev/null +++ b/data/vals_ai/aristotle/aristotle/6f81e329-9aa6-418d-8b63-d7f82c5bf10b.json @@ -0,0 +1,88 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/aristotle_aristotle/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "aristotle", + "id": "aristotle/aristotle", + "developer": "aristotle", + "additional_details": { + "vals_model_id": "aristotle/aristotle", + "vals_provider": "Harmonic" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:aristotle/aristotle:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 71.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "71", + "raw_stderr": "4.5", + "latency": "5820", + "provider": "Harmonic" + }, + "uncertainty": { + "standard_error": { + "value": 4.5, + "method": "vals_reported" + } + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/cohere/command-a-03-2025/0e45a7eb-2131-488a-8242-97d0bd29bd2a.json b/data/vals_ai/cohere/command-a-03-2025/0e45a7eb-2131-488a-8242-97d0bd29bd2a.json new file mode 100644 index 0000000000000000000000000000000000000000..08d712bf17d2d630b9eed360a3ad4d19c6f485b5 --- /dev/null +++ b/data/vals_ai/cohere/command-a-03-2025/0e45a7eb-2131-488a-8242-97d0bd29bd2a.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/cohere_command-a-03-2025/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "command-a-03-2025", + "id": "cohere/command-a-03-2025", + "developer": "cohere", + "additional_details": { + "vals_model_id": "cohere/command-a-03-2025", + "vals_provider": "Cohere" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 50.0, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "50", + "raw_stderr": "3.363", + "latency": "11.887", + "cost_per_test": "0.012408", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 3.363, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 46.97, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "46.97", + "raw_stderr": "3.067", + "latency": "11.092", + "cost_per_test": "0.00778", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 3.067, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 48.485, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "48.485", + "raw_stderr": "2.276", + "latency": "11.49", + "cost_per_test": "0.010094", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 2.276, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/cohere/command-a-03-2025/1f89f2c2-362f-4d40-bc4a-ba7732aceb65.json b/data/vals_ai/cohere/command-a-03-2025/1f89f2c2-362f-4d40-bc4a-ba7732aceb65.json new file mode 100644 index 0000000000000000000000000000000000000000..147b08f8eb2419aee9a06508f0f303002823f842 --- /dev/null +++ b/data/vals_ai/cohere/command-a-03-2025/1f89f2c2-362f-4d40-bc4a-ba7732aceb65.json @@ -0,0 +1,361 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/cohere_command-a-03-2025/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "command-a-03-2025", + "id": "cohere/command-a-03-2025", + "developer": "cohere", + "additional_details": { + "vals_model_id": "cohere/command-a-03-2025", + "vals_provider": "Cohere" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "191.352", + "cost_per_test": "0.150556", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 4.598, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "4.598", + "raw_stderr": "1.299", + "latency": "259.738", + "cost_per_test": "0.171989", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.299, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 13.918, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "13.918", + "raw_stderr": "2.491", + "latency": "179.035", + "cost_per_test": "0.276573", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 2.491, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1068.024", + "cost_per_test": "0.084132", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 7.8, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "7.8", + "raw_stderr": "1.201", + "latency": "227.531", + "cost_per_test": "0.21024", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.201, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/cohere/command-a-03-2025/4c58a7e7-e856-4d4f-a50b-44d79879c26c.json b/data/vals_ai/cohere/command-a-03-2025/4c58a7e7-e856-4d4f-a50b-44d79879c26c.json new file mode 100644 index 0000000000000000000000000000000000000000..f0b8e8220963ba456754be6e5eb7f586634d5433 --- /dev/null +++ b/data/vals_ai/cohere/command-a-03-2025/4c58a7e7-e856-4d4f-a50b-44d79879c26c.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/cohere_command-a-03-2025/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "command-a-03-2025", + "id": "cohere/command-a-03-2025", + "developer": "cohere", + "additional_details": { + "vals_model_id": "cohere/command-a-03-2025", + "vals_provider": "Cohere" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 56.177, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.177", + "raw_stderr": "1.694", + "latency": "3.828", + "cost_per_test": "0.005637", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.694, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 41.958, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "41.958", + "raw_stderr": "1.685", + "latency": "27.747", + "cost_per_test": "0.445899", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.685, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 45.96, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "45.96", + "raw_stderr": "0.972", + "latency": "14.146", + "cost_per_test": "0.206138", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.972, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 39.744, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "39.744", + "raw_stderr": "1.671", + "latency": "10.862", + "cost_per_test": "0.166879", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.671, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/cohere/command-a-03-2025/595651ef-e871-4343-a996-d9802fe2acb2.json b/data/vals_ai/cohere/command-a-03-2025/595651ef-e871-4343-a996-d9802fe2acb2.json new file mode 100644 index 0000000000000000000000000000000000000000..0887e9eb1205da856151ffe632b6e82da30ba572 --- /dev/null +++ b/data/vals_ai/cohere/command-a-03-2025/595651ef-e871-4343-a996-d9802fe2acb2.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/cohere_command-a-03-2025/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "command-a-03-2025", + "id": "cohere/command-a-03-2025", + "developer": "cohere", + "additional_details": { + "vals_model_id": "cohere/command-a-03-2025", + "vals_provider": "Cohere" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 48.978, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "48.978", + "raw_stderr": "1.429", + "latency": "11.214", + "cost_per_test": "0.006484", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.429, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 61.366, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.366", + "raw_stderr": "0.952", + "latency": "11.214", + "cost_per_test": "0.006484", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.952, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 73.753, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "73.753", + "raw_stderr": "1.258", + "latency": "11.214", + "cost_per_test": "0.006484", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.258, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/cohere/command-a-03-2025/823a6c72-e725-4adc-89a2-2ace89a48dda.json b/data/vals_ai/cohere/command-a-03-2025/823a6c72-e725-4adc-89a2-2ace89a48dda.json new file mode 100644 index 0000000000000000000000000000000000000000..b3c285316dee8fcc50df4a9414502e98091e3fc9 --- /dev/null +++ b/data/vals_ai/cohere/command-a-03-2025/823a6c72-e725-4adc-89a2-2ace89a48dda.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/cohere_command-a-03-2025/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "command-a-03-2025", + "id": "cohere/command-a-03-2025", + "developer": "cohere", + "additional_details": { + "vals_model_id": "cohere/command-a-03-2025", + "vals_provider": "Cohere" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 78.882, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "78.882", + "raw_stderr": "2.275", + "latency": "20.386", + "cost_per_test": "0.004396", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 2.275, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 2.571, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "2.571", + "raw_stderr": "0.846", + "latency": "7.003", + "cost_per_test": "0.008464", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.846, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 23.76, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "23.76", + "raw_stderr": "2.175", + "latency": "7.849", + "cost_per_test": "0.006843", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 2.175, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 35.071, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "35.071", + "raw_stderr": "1.086", + "latency": "11.746", + "cost_per_test": "0.006568", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.086, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/cohere/command-a-03-2025/8c81a736-16ca-458a-a23f-ee42fa0a0cd6.json b/data/vals_ai/cohere/command-a-03-2025/8c81a736-16ca-458a-a23f-ee42fa0a0cd6.json new file mode 100644 index 0000000000000000000000000000000000000000..2c06ee609f1167376e0812d3e382c46c14f553ce --- /dev/null +++ b/data/vals_ai/cohere/command-a-03-2025/8c81a736-16ca-458a-a23f-ee42fa0a0cd6.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/cohere_command-a-03-2025/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "command-a-03-2025", + "id": "cohere/command-a-03-2025", + "developer": "cohere", + "additional_details": { + "vals_model_id": "cohere/command-a-03-2025", + "vals_provider": "Cohere" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.869, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.869", + "raw_stderr": "1.643", + "latency": "9.774", + "cost_per_test": "0.007289", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.643, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.384, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.384", + "raw_stderr": "1.678", + "latency": "7.866", + "cost_per_test": "0.006893", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.678, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 68.905, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "68.905", + "raw_stderr": "1.404", + "latency": "17.157", + "cost_per_test": "0.010878", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.404, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 70.976, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.976", + "raw_stderr": "2.368", + "latency": "8.315", + "cost_per_test": "0.008203", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 2.368, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.725, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.725", + "raw_stderr": "1.535", + "latency": "5.885", + "cost_per_test": "0.006694", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 55.521, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "55.521", + "raw_stderr": "1.606", + "latency": "18.253", + "cost_per_test": "0.011655", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.606, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 70.538, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.538", + "raw_stderr": "1.683", + "latency": "6.736", + "cost_per_test": "0.006182", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.683, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 64.042, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "64.042", + "raw_stderr": "2.474", + "latency": "5.903", + "cost_per_test": "0.009169", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 2.474, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 51.408, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "51.408", + "raw_stderr": "1.506", + "latency": "8.478", + "cost_per_test": "0.008141", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.506, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.46, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.46", + "raw_stderr": "1.133", + "latency": "13.483", + "cost_per_test": "0.009716", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.133, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 69.481, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.481", + "raw_stderr": "1.563", + "latency": "6.18", + "cost_per_test": "0.005133", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.563, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 69.172, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.172", + "raw_stderr": "0.463", + "latency": "9.853", + "cost_per_test": "0.007902", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.463, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 62.124, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "62.124", + "raw_stderr": "2.178", + "latency": "7.621", + "cost_per_test": "0.005916", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 2.178, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 69.284, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.284", + "raw_stderr": "1.298", + "latency": "15.203", + "cost_per_test": "0.008877", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.298, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.687, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.687", + "raw_stderr": "1.612", + "latency": "7.087", + "cost_per_test": "0.005887", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.612, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/cohere/command-a-03-2025/d4a46427-a057-4ab3-8aac-bf51dc2d14c3.json b/data/vals_ai/cohere/command-a-03-2025/d4a46427-a057-4ab3-8aac-bf51dc2d14c3.json new file mode 100644 index 0000000000000000000000000000000000000000..eb407f3ca55b8c985b07d0962de7634d08ff2780 --- /dev/null +++ b/data/vals_ai/cohere/command-a-03-2025/d4a46427-a057-4ab3-8aac-bf51dc2d14c3.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/cohere_command-a-03-2025/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "command-a-03-2025", + "id": "cohere/command-a-03-2025", + "developer": "cohere", + "additional_details": { + "vals_model_id": "cohere/command-a-03-2025", + "vals_provider": "Cohere" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.079, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.079", + "raw_stderr": "0.962", + "latency": "0.573", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.962, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.414, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.414", + "raw_stderr": "0.209", + "latency": "0.696", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.209, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 84.798, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.798", + "raw_stderr": "0.615", + "latency": "0.762", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.615, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.697, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.697", + "raw_stderr": "0.422", + "latency": "0.797", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.422, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.598, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.598", + "raw_stderr": "0.826", + "latency": "0.57", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.826, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:cohere/command-a-03-2025:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 73.598, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "73.598", + "raw_stderr": "1.556", + "latency": "1.385", + "temperature": "0.3", + "max_output_tokens": "8000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.556, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 8000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/cohere/command-a-plus-05-2026/36188c89-2ac8-4f55-ba53-d429c19fc8f4.json b/data/vals_ai/cohere/command-a-plus-05-2026/36188c89-2ac8-4f55-ba53-d429c19fc8f4.json new file mode 100644 index 0000000000000000000000000000000000000000..296a80d7f981551178e6f32f182d463ada3f5a71 --- /dev/null +++ b/data/vals_ai/cohere/command-a-plus-05-2026/36188c89-2ac8-4f55-ba53-d429c19fc8f4.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/cohere_command-a-plus-05-2026/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "command-a-plus-05-2026", + "id": "cohere/command-a-plus-05-2026", + "developer": "cohere", + "additional_details": { + "vals_model_id": "cohere/command-a-plus-05-2026", + "vals_provider": "Cohere" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.982, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.982", + "raw_stderr": "1.169", + "latency": "28.304", + "cost_per_test": "0.025781", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.169, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.411, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.411", + "raw_stderr": "1.291", + "latency": "66.971", + "cost_per_test": "0.04243", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.291, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.837, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.837", + "raw_stderr": "1.046", + "latency": "72.157", + "cost_per_test": "0.048928", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.046, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.634, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.634", + "raw_stderr": "1.826", + "latency": "64.078", + "cost_per_test": "0.039309", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.826, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.109, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.109", + "raw_stderr": "1.329", + "latency": "29.927", + "cost_per_test": "0.02356", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.329, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 68.008, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "68.008", + "raw_stderr": "1.508", + "latency": "212.024", + "cost_per_test": "0.128261", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.508, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.594, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.594", + "raw_stderr": "1.541", + "latency": "41.931", + "cost_per_test": "0.024181", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.541, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.667", + "raw_stderr": "2.438", + "latency": "58.211", + "cost_per_test": "0.034301", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 2.438, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 57.13, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "57.13", + "raw_stderr": "1.491", + "latency": "88.641", + "cost_per_test": "0.04333", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.491, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.746, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.746", + "raw_stderr": "0.722", + "latency": "63.038", + "cost_per_test": "0.035825", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.722, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.701, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.701", + "raw_stderr": "1.461", + "latency": "42.635", + "cost_per_test": "0.026316", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.461, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.151, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.151", + "raw_stderr": "0.406", + "latency": "66.351", + "cost_per_test": "0.040639", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.406, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 70.942, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.942", + "raw_stderr": "2.109", + "latency": "54.606", + "cost_per_test": "0.034931", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 2.109, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.913, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.913", + "raw_stderr": "0.943", + "latency": "75.825", + "cost_per_test": "0.043812", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.943, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.444, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.444", + "raw_stderr": "1.502", + "latency": "30.562", + "cost_per_test": "0.017984", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.502, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/cohere/command-a-plus-05-2026/371c6fa5-016e-4b9d-aff1-d74af4a3aa36.json b/data/vals_ai/cohere/command-a-plus-05-2026/371c6fa5-016e-4b9d-aff1-d74af4a3aa36.json new file mode 100644 index 0000000000000000000000000000000000000000..03f5b7b1dd591dd6868b5079ce662500894920a8 --- /dev/null +++ b/data/vals_ai/cohere/command-a-plus-05-2026/371c6fa5-016e-4b9d-aff1-d74af4a3aa36.json @@ -0,0 +1,103 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/cohere_command-a-plus-05-2026/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "command-a-plus-05-2026", + "id": "cohere/command-a-plus-05-2026", + "developer": "cohere", + "additional_details": { + "vals_model_id": "cohere/command-a-plus-05-2026", + "vals_provider": "Cohere" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 70.636, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.636", + "raw_stderr": "1.095", + "latency": "159.969", + "cost_per_test": "0.104105", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.095, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/cohere/command-a-plus-05-2026/3a88c3fe-f34e-4dc9-86b3-5704ceb02dd9.json b/data/vals_ai/cohere/command-a-plus-05-2026/3a88c3fe-f34e-4dc9-86b3-5704ceb02dd9.json new file mode 100644 index 0000000000000000000000000000000000000000..f60e5ec0d919656e029093fb9adf754600c981f2 --- /dev/null +++ b/data/vals_ai/cohere/command-a-plus-05-2026/3a88c3fe-f34e-4dc9-86b3-5704ceb02dd9.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/cohere_command-a-plus-05-2026/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "command-a-plus-05-2026", + "id": "cohere/command-a-plus-05-2026", + "developer": "cohere", + "additional_details": { + "vals_model_id": "cohere/command-a-plus-05-2026", + "vals_provider": "Cohere" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.137, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.137", + "raw_stderr": "0.754", + "latency": "32.294", + "cost_per_test": "0.023039", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.754, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 56.196, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "56.196", + "raw_stderr": "2.663", + "latency": "475.647", + "cost_per_test": "0.306506", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 2.663, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 83.812, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "83.812", + "raw_stderr": "1.882", + "latency": "208.984", + "cost_per_test": "0.137262", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.882, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 79.382, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "79.382", + "raw_stderr": "1.116", + "latency": "238.975", + "cost_per_test": "0.155602", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.116, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/cohere/command-a-plus-05-2026/4304bf9d-aa23-4133-985c-265fe76622b2.json b/data/vals_ai/cohere/command-a-plus-05-2026/4304bf9d-aa23-4133-985c-265fe76622b2.json new file mode 100644 index 0000000000000000000000000000000000000000..5a9d4b53fb2ddb3a1adf417d81f9e50eb5354ce5 --- /dev/null +++ b/data/vals_ai/cohere/command-a-plus-05-2026/4304bf9d-aa23-4133-985c-265fe76622b2.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/cohere_command-a-plus-05-2026/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "command-a-plus-05-2026", + "id": "cohere/command-a-plus-05-2026", + "developer": "cohere", + "additional_details": { + "vals_model_id": "cohere/command-a-plus-05-2026", + "vals_provider": "Cohere" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 41.128, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "41.128", + "raw_stderr": "1.407", + "latency": "141.943", + "cost_per_test": "0.106858", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.407, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 52.82, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.82", + "raw_stderr": "0.981", + "latency": "141.943", + "cost_per_test": "0.106858", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.981, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 64.513, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.513", + "raw_stderr": "1.368", + "latency": "141.943", + "cost_per_test": "0.106858", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.368, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/cohere/command-a-plus-05-2026/5af3096f-6586-4018-8602-aea8892eede4.json b/data/vals_ai/cohere/command-a-plus-05-2026/5af3096f-6586-4018-8602-aea8892eede4.json new file mode 100644 index 0000000000000000000000000000000000000000..7767a1228189d2bd5734b9ddd75599d6369fd595 --- /dev/null +++ b/data/vals_ai/cohere/command-a-plus-05-2026/5af3096f-6586-4018-8602-aea8892eede4.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/cohere_command-a-plus-05-2026/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "command-a-plus-05-2026", + "id": "cohere/command-a-plus-05-2026", + "developer": "cohere", + "additional_details": { + "vals_model_id": "cohere/command-a-plus-05-2026", + "vals_provider": "Cohere" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 50.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "50", + "raw_stderr": "0", + "latency": "0", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 3.333, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "3.333", + "raw_stderr": "0", + "latency": "0", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 23.03, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "23.03", + "raw_stderr": "0.606", + "latency": "0", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.606, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 17.603, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "17.603", + "raw_stderr": "0.375", + "latency": "0", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.375, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/cohere/command-a-plus-05-2026/785dd949-bc10-405d-b072-4163aacfbce7.json b/data/vals_ai/cohere/command-a-plus-05-2026/785dd949-bc10-405d-b072-4163aacfbce7.json new file mode 100644 index 0000000000000000000000000000000000000000..ccf535dd794b79a7129cf2da2512e5af98aaca59 --- /dev/null +++ b/data/vals_ai/cohere/command-a-plus-05-2026/785dd949-bc10-405d-b072-4163aacfbce7.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/cohere_command-a-plus-05-2026/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "command-a-plus-05-2026", + "id": "cohere/command-a-plus-05-2026", + "developer": "cohere", + "additional_details": { + "vals_model_id": "cohere/command-a-plus-05-2026", + "vals_provider": "Cohere" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 59.091, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.091", + "raw_stderr": "1.679", + "latency": "8.39", + "cost_per_test": "0.009481", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.679, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 46.037, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "46.037", + "raw_stderr": "1.702", + "latency": "152.239", + "cost_per_test": "0.222417", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.702, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 53.147, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "53.147", + "raw_stderr": "0.978", + "latency": "71.106", + "cost_per_test": "0.132683", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.978, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 54.312, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "54.312", + "raw_stderr": "1.701", + "latency": "52.688", + "cost_per_test": "0.166152", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.701, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/cohere/command-a-plus-05-2026/9348242f-baaf-4e2a-b57b-e2ce044a685f.json b/data/vals_ai/cohere/command-a-plus-05-2026/9348242f-baaf-4e2a-b57b-e2ce044a685f.json new file mode 100644 index 0000000000000000000000000000000000000000..fb1a6d815457fd8f3e8c7cbf0c35fef927dc565d --- /dev/null +++ b/data/vals_ai/cohere/command-a-plus-05-2026/9348242f-baaf-4e2a-b57b-e2ce044a685f.json @@ -0,0 +1,533 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_multimodal_index/cohere_command-a-plus-05-2026/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Multimodal Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "command-a-plus-05-2026", + "id": "cohere/command-a-plus-05-2026", + "developer": "cohere", + "additional_details": { + "vals_model_id": "cohere/command-a-plus-05-2026", + "vals_provider": "Cohere" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_multimodal_index:corp_fin_v2:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.vals_multimodal_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Multimodal Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (CorpFin v2).", + "metric_id": "vals_ai.vals_multimodal_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 54.312, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "54.312", + "raw_stderr": "1.701", + "latency": "52.688", + "cost_per_test": "0.166152", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.701, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:finance_agent:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.vals_multimodal_index.finance_agent", + "source_data": { + "dataset_name": "Vals Multimodal Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_multimodal_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 10.426, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "10.426", + "raw_stderr": "0.934", + "latency": "642.935", + "cost_per_test": "1.446489", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.934, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:mortgage_tax:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.vals_multimodal_index.mortgage_tax", + "source_data": { + "dataset_name": "Vals Multimodal Index - Mortgage Tax", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "mortgage_tax", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Mortgage Tax).", + "metric_id": "vals_ai.vals_multimodal_index.mortgage_tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 53.02, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "mortgage_tax", + "task_name": "Mortgage Tax", + "dataset_type": "private", + "industry": "index", + "raw_score": "53.02", + "raw_stderr": "0.967", + "latency": "46.542", + "cost_per_test": "0.05135", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.967, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:overall:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.vals_multimodal_index.overall", + "source_data": { + "dataset_name": "Vals Multimodal Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Overall).", + "metric_id": "vals_ai.vals_multimodal_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 27.257, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "27.257", + "raw_stderr": "0.623", + "latency": "231.362", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.623, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:sage:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.vals_multimodal_index.sage", + "source_data": { + "dataset_name": "Vals Multimodal Index - SAGE", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "sage", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SAGE).", + "metric_id": "vals_ai.vals_multimodal_index.sage.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 10.488, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "sage", + "task_name": "SAGE", + "dataset_type": "private", + "industry": "index", + "raw_score": "10.488", + "raw_stderr": "2.116", + "latency": "42.447", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 2.116, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:swebench:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.vals_multimodal_index.swebench", + "source_data": { + "dataset_name": "Vals Multimodal Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SWE-bench).", + "metric_id": "vals_ai.vals_multimodal_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 37.255, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "37.255", + "raw_stderr": "4.787", + "latency": "239.965", + "cost_per_test": "6.073739" + }, + "uncertainty": { + "standard_error": { + "value": 4.787, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:terminal_bench_2_1:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.vals_multimodal_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Multimodal Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_multimodal_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 17.603, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "17.603", + "raw_stderr": "0.375", + "latency": "0", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.375, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:vibe_code_bench:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.vals_multimodal_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Multimodal Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_multimodal_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "0", + "raw_stderr": "0", + "latency": "594.958", + "cost_per_test": "5.602728", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/cohere/command-a-plus-05-2026/957c504d-5fcf-435c-b87d-198f4492bff1.json b/data/vals_ai/cohere/command-a-plus-05-2026/957c504d-5fcf-435c-b87d-198f4492bff1.json new file mode 100644 index 0000000000000000000000000000000000000000..5b905852c173c7b45bfbb03c155f437359568bc6 --- /dev/null +++ b/data/vals_ai/cohere/command-a-plus-05-2026/957c504d-5fcf-435c-b87d-198f4492bff1.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/cohere_command-a-plus-05-2026/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "command-a-plus-05-2026", + "id": "cohere/command-a-plus-05-2026", + "developer": "cohere", + "additional_details": { + "vals_model_id": "cohere/command-a-plus-05-2026", + "vals_provider": "Cohere" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "447.356", + "cost_per_test": "0", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/cohere/command-a-plus-05-2026/a0c41f9a-80d0-41bb-8614-3a5fd6cca140.json b/data/vals_ai/cohere/command-a-plus-05-2026/a0c41f9a-80d0-41bb-8614-3a5fd6cca140.json new file mode 100644 index 0000000000000000000000000000000000000000..cc25b2a8f441f7ce341e63ab6f174747c34ed6ce --- /dev/null +++ b/data/vals_ai/cohere/command-a-plus-05-2026/a0c41f9a-80d0-41bb-8614-3a5fd6cca140.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/cohere_command-a-plus-05-2026/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "command-a-plus-05-2026", + "id": "cohere/command-a-plus-05-2026", + "developer": "cohere", + "additional_details": { + "vals_model_id": "cohere/command-a-plus-05-2026", + "vals_provider": "Cohere" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 64.785, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.785", + "raw_stderr": "1.347", + "latency": "62.619", + "cost_per_test": "0.061337", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.347, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 53.02, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "53.02", + "raw_stderr": "0.967", + "latency": "46.542", + "cost_per_test": "0.05135", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.967, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 41.256, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "41.256", + "raw_stderr": "1.388", + "latency": "30.465", + "cost_per_test": "0.041362", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.388, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/cohere/command-a-plus-05-2026/b20d9854-0492-4fc4-97ea-6af83354238a.json b/data/vals_ai/cohere/command-a-plus-05-2026/b20d9854-0492-4fc4-97ea-6af83354238a.json new file mode 100644 index 0000000000000000000000000000000000000000..1ab60114f34d9e8b6ac5a193d3cfdd463f0010d5 --- /dev/null +++ b/data/vals_ai/cohere/command-a-plus-05-2026/b20d9854-0492-4fc4-97ea-6af83354238a.json @@ -0,0 +1,740 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/cohere_command-a-plus-05-2026/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "command-a-plus-05-2026", + "id": "cohere/command-a-plus-05-2026", + "developer": "cohere", + "additional_details": { + "vals_model_id": "cohere/command-a-plus-05-2026", + "vals_provider": "Cohere" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 3.107, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "3.107", + "raw_stderr": "1.55", + "latency": "689.407", + "cost_per_test": "1.978852", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.55, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 2.761, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "2.761", + "raw_stderr": "0.539", + "latency": "566.745", + "cost_per_test": "1.63003", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.539, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 16.462, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "16.462", + "raw_stderr": "1.001", + "latency": "681.636", + "cost_per_test": "1.933787", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.001, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 7.584, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "7.584", + "raw_stderr": "0.695", + "latency": "581.011", + "cost_per_test": "1.512103", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.695, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 7.963, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "7.963", + "raw_stderr": "2.364", + "latency": "568.404", + "cost_per_test": "1.368424", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 2.364, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 0.763, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "0.763", + "raw_stderr": "0.144", + "latency": "506.787", + "cost_per_test": "2.31425", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.144, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 22.257, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "22.257", + "raw_stderr": "3.204", + "latency": "391", + "cost_per_test": "0.945316", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 3.204, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 8.899, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "8.899", + "raw_stderr": "2.465", + "latency": "836.425", + "cost_per_test": "1.899291", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 2.465, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 11.963, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "11.963", + "raw_stderr": "0.937", + "latency": "327.572", + "cost_per_test": "1.179474", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.937, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:overall:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 9.044, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "9.044", + "raw_stderr": "0.409", + "latency": "566.745", + "cost_per_test": "1.63003", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.409, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 2.395, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "2.395", + "raw_stderr": "0.666", + "latency": "585.071", + "cost_per_test": "1.469353", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.666, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/cohere/command-a-plus-05-2026/b6bc163e-7300-474e-bd90-47840d675cff.json b/data/vals_ai/cohere/command-a-plus-05-2026/b6bc163e-7300-474e-bd90-47840d675cff.json new file mode 100644 index 0000000000000000000000000000000000000000..aacb6c73df0a7ff7c5ee3c97500fa076b2e8bf3b --- /dev/null +++ b/data/vals_ai/cohere/command-a-plus-05-2026/b6bc163e-7300-474e-bd90-47840d675cff.json @@ -0,0 +1,438 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/cohere_command-a-plus-05-2026/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "command-a-plus-05-2026", + "id": "cohere/command-a-plus-05-2026", + "developer": "cohere", + "additional_details": { + "vals_model_id": "cohere/command-a-plus-05-2026", + "vals_provider": "Cohere" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.12, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.12", + "raw_stderr": "1.013", + "latency": "9.621", + "cost_per_test": "0.012736", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.013, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 54.446, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "54.446", + "raw_stderr": "0.288", + "latency": "7.838", + "cost_per_test": "0.011865", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.288, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 55.176, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "55.176", + "raw_stderr": "0.865", + "latency": "5.359", + "cost_per_test": "0.008963", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.865, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 61.422, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "61.422", + "raw_stderr": "0.627", + "latency": "22.648", + "cost_per_test": "0.033012", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.627, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 68.648, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "68.648", + "raw_stderr": "1.472", + "latency": "4.851", + "cost_per_test": "0.007517", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.472, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 46.72, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "46.72", + "raw_stderr": "2.408", + "latency": "85.57", + "cost_per_test": "0.123978", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 2.408, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/cohere/command-a-plus-05-2026/bc334ef1-a020-4d24-a502-07bdb10df4d6.json b/data/vals_ai/cohere/command-a-plus-05-2026/bc334ef1-a020-4d24-a502-07bdb10df4d6.json new file mode 100644 index 0000000000000000000000000000000000000000..06626a9ddd4900547133cc22af918f13347bbfde --- /dev/null +++ b/data/vals_ai/cohere/command-a-plus-05-2026/bc334ef1-a020-4d24-a502-07bdb10df4d6.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/cohere_command-a-plus-05-2026/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "command-a-plus-05-2026", + "id": "cohere/command-a-plus-05-2026", + "developer": "cohere", + "additional_details": { + "vals_model_id": "cohere/command-a-plus-05-2026", + "vals_provider": "Cohere" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 55.682, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "55.682", + "raw_stderr": "3.646", + "latency": "203.051", + "cost_per_test": "0.140316", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 3.646, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/cohere/command-a-plus-05-2026/cc228eb9-953f-45bc-a3ce-5fa701a26bbd.json b/data/vals_ai/cohere/command-a-plus-05-2026/cc228eb9-953f-45bc-a3ce-5fa701a26bbd.json new file mode 100644 index 0000000000000000000000000000000000000000..0a154a94b46a462d1f0a7766da6261b2f40def4c --- /dev/null +++ b/data/vals_ai/cohere/command-a-plus-05-2026/cc228eb9-953f-45bc-a3ce-5fa701a26bbd.json @@ -0,0 +1,406 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/cohere_command-a-plus-05-2026/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "command-a-plus-05-2026", + "id": "cohere/command-a-plus-05-2026", + "developer": "cohere", + "additional_details": { + "vals_model_id": "cohere/command-a-plus-05-2026", + "vals_provider": "Cohere" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 54.312, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "54.312", + "raw_stderr": "1.701", + "latency": "52.688", + "cost_per_test": "0.166152", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.701, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 10.426, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "10.426", + "raw_stderr": "0.934", + "latency": "642.935", + "cost_per_test": "1.446489", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.934, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:overall:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 24.688, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "24.688", + "raw_stderr": "0.755", + "latency": "306.109", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.755, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 37.255, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "37.255", + "raw_stderr": "4.787", + "latency": "239.965", + "cost_per_test": "6.073739" + }, + "uncertainty": { + "standard_error": { + "value": 4.787, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 17.603, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "17.603", + "raw_stderr": "0.375", + "latency": "0", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.375, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "0", + "raw_stderr": "0", + "latency": "594.958", + "cost_per_test": "5.602728", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/cohere/command-a-plus-05-2026/d5264c30-dba7-492b-8386-a6a0a74b0ca1.json b/data/vals_ai/cohere/command-a-plus-05-2026/d5264c30-dba7-492b-8386-a6a0a74b0ca1.json new file mode 100644 index 0000000000000000000000000000000000000000..1723e7cfc7ade4eb32e2e566a77d69d3812edbcf --- /dev/null +++ b/data/vals_ai/cohere/command-a-plus-05-2026/d5264c30-dba7-492b-8386-a6a0a74b0ca1.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/cohere_command-a-plus-05-2026/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "command-a-plus-05-2026", + "id": "cohere/command-a-plus-05-2026", + "developer": "cohere", + "additional_details": { + "vals_model_id": "cohere/command-a-plus-05-2026", + "vals_provider": "Cohere" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 71.212, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.212", + "raw_stderr": "3.472", + "latency": "99.605", + "cost_per_test": "0.131553", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 3.472, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 73.737, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.737", + "raw_stderr": "3.266", + "latency": "125.83", + "cost_per_test": "0.141597", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 3.266, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 72.474, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.474", + "raw_stderr": "2.383", + "latency": "112.718", + "cost_per_test": "0.136575", + "temperature": "0.9", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 2.383, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.9, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/cohere/command-a-plus-05-2026/eee51427-5132-4b0c-ac4d-3f01ef173069.json b/data/vals_ai/cohere/command-a-plus-05-2026/eee51427-5132-4b0c-ac4d-3f01ef173069.json new file mode 100644 index 0000000000000000000000000000000000000000..780c1642a2a8688717ba8194644e2c9669141f19 --- /dev/null +++ b/data/vals_ai/cohere/command-a-plus-05-2026/eee51427-5132-4b0c-ac4d-3f01ef173069.json @@ -0,0 +1,288 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/cohere_command-a-plus-05-2026/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "command-a-plus-05-2026", + "id": "cohere/command-a-plus-05-2026", + "developer": "cohere", + "additional_details": { + "vals_model_id": "cohere/command-a-plus-05-2026", + "vals_provider": "Cohere" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 15.079, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "15.079", + "raw_stderr": "4.456", + "latency": "49.348", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 4.456, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 13.492, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "13.492", + "raw_stderr": "4.048", + "latency": "38.268", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 4.048, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 2.892, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "2.892", + "raw_stderr": "2.011", + "latency": "39.724", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 2.011, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:cohere/command-a-plus-05-2026:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 10.488, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "10.488", + "raw_stderr": "2.116", + "latency": "42.447", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 2.116, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/cohere/command-r-plus/410f8914-57bf-4d6a-9e3d-573ba60cfe11.json b/data/vals_ai/cohere/command-r-plus/410f8914-57bf-4d6a-9e3d-573ba60cfe11.json new file mode 100644 index 0000000000000000000000000000000000000000..5fc25c97f343dc079b83dc99154970a7dd300bef --- /dev/null +++ b/data/vals_ai/cohere/command-r-plus/410f8914-57bf-4d6a-9e3d-573ba60cfe11.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/cohere_command-r-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "command-r-plus", + "id": "cohere/command-r-plus", + "developer": "cohere", + "additional_details": { + "vals_model_id": "cohere/command-r-plus", + "vals_provider": "Cohere" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:cohere/command-r-plus:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 48.306, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "48.306", + "raw_stderr": "1.196", + "latency": "0.238", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.196, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:cohere/command-r-plus:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 71.624, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "71.624", + "raw_stderr": "0.237", + "latency": "0.239", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.237, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:cohere/command-r-plus:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 76.319, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "76.319", + "raw_stderr": "0.708", + "latency": "0.318", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.708, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:cohere/command-r-plus:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 68.287, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "68.287", + "raw_stderr": "0.65", + "latency": "0.396", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.65, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:cohere/command-r-plus:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 75.635, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "75.635", + "raw_stderr": "0.842", + "latency": "0.247", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.842, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:cohere/command-r-plus:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 69.553, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "69.553", + "raw_stderr": "2.805", + "latency": "0.94", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 2.805, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/cohere/command-r-plus/a0b210fc-06c5-4482-9ab4-fdc9420640af.json b/data/vals_ai/cohere/command-r-plus/a0b210fc-06c5-4482-9ab4-fdc9420640af.json new file mode 100644 index 0000000000000000000000000000000000000000..670b934f9a15577b22db1cc8fec15145ba7b3c5e --- /dev/null +++ b/data/vals_ai/cohere/command-r-plus/a0b210fc-06c5-4482-9ab4-fdc9420640af.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/cohere_command-r-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "command-r-plus", + "id": "cohere/command-r-plus", + "developer": "cohere", + "additional_details": { + "vals_model_id": "cohere/command-r-plus", + "vals_provider": "Cohere" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:cohere/command-r-plus:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 28.283, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "28.283", + "raw_stderr": "3.183", + "latency": "8", + "cost_per_test": "0.009178", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 3.183, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:cohere/command-r-plus:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 33.838, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "33.838", + "raw_stderr": "3.266", + "latency": "8.387", + "cost_per_test": "0.005699", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 3.266, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:cohere/command-r-plus:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 31.06, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "31.06", + "raw_stderr": "2.28", + "latency": "8.194", + "cost_per_test": "0.007438", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 2.28, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/cohere/command-r-plus/b0cba6a4-2124-4d2c-9530-66a1fc36e225.json b/data/vals_ai/cohere/command-r-plus/b0cba6a4-2124-4d2c-9530-66a1fc36e225.json new file mode 100644 index 0000000000000000000000000000000000000000..de6501f000a27782c144d4ada08d8e551089fa0f --- /dev/null +++ b/data/vals_ai/cohere/command-r-plus/b0cba6a4-2124-4d2c-9530-66a1fc36e225.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/cohere_command-r-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "command-r-plus", + "id": "cohere/command-r-plus", + "developer": "cohere", + "additional_details": { + "vals_model_id": "cohere/command-r-plus", + "vals_provider": "Cohere" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:cohere/command-r-plus:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 62.483, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "62.483", + "raw_stderr": "1.808", + "latency": "5.069", + "cost_per_test": "0.00615", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.808, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:cohere/command-r-plus:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 41.572, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "41.572", + "raw_stderr": "1.754", + "latency": "5.995", + "cost_per_test": "0.00596", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.754, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:cohere/command-r-plus:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 28.71, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "28.71", + "raw_stderr": "1.345", + "latency": "6.703", + "cost_per_test": "0.007669", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.345, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:cohere/command-r-plus:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 46.585, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "46.585", + "raw_stderr": "2.463", + "latency": "4.718", + "cost_per_test": "0.006329", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 2.463, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:cohere/command-r-plus:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 52.251, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "52.251", + "raw_stderr": "1.72", + "latency": "4.52", + "cost_per_test": "0.005444", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.72, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:cohere/command-r-plus:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 26.109, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "26.109", + "raw_stderr": "1.411", + "latency": "7.918", + "cost_per_test": "0.00741", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.411, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:cohere/command-r-plus:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 46.088, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "46.088", + "raw_stderr": "1.743", + "latency": "5.687", + "cost_per_test": "0.005504", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.743, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:cohere/command-r-plus:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 49.344, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "49.344", + "raw_stderr": "2.561", + "latency": "1.934", + "cost_per_test": "0.00626", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 2.561, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:cohere/command-r-plus:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 36.058, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "36.058", + "raw_stderr": "1.446", + "latency": "3.06", + "cost_per_test": "0.005053", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.446, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:cohere/command-r-plus:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 39.008, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "39.008", + "raw_stderr": "1.327", + "latency": "4.218", + "cost_per_test": "0.006778", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.327, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:cohere/command-r-plus:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 51.082, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "51.082", + "raw_stderr": "1.645", + "latency": "1.729", + "cost_per_test": "0.003163", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.645, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:cohere/command-r-plus:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 44.003, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "44.003", + "raw_stderr": "0.479", + "latency": "4.419", + "cost_per_test": "0.00566", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.479, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:cohere/command-r-plus:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 42.084, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "42.084", + "raw_stderr": "2.209", + "latency": "2.517", + "cost_per_test": "0.003737", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 2.209, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:cohere/command-r-plus:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 34.642, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "34.642", + "raw_stderr": "1.32", + "latency": "4.669", + "cost_per_test": "0.005526", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.32, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:cohere/command-r-plus:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 60.025, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "60.025", + "raw_stderr": "1.734", + "latency": "3.13", + "cost_per_test": "0.004253", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.734, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/cohere/command-r-plus/e620fcf3-d135-4914-9645-9e50918aa346.json b/data/vals_ai/cohere/command-r-plus/e620fcf3-d135-4914-9645-9e50918aa346.json new file mode 100644 index 0000000000000000000000000000000000000000..ea8b8ca78c1702702b1ecf2eb0d1486364880a5b --- /dev/null +++ b/data/vals_ai/cohere/command-r-plus/e620fcf3-d135-4914-9645-9e50918aa346.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/cohere_command-r-plus/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "command-r-plus", + "id": "cohere/command-r-plus", + "developer": "cohere", + "additional_details": { + "vals_model_id": "cohere/command-r-plus", + "vals_provider": "Cohere" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:cohere/command-r-plus:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 48.447, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "48.447", + "raw_stderr": "2.785", + "latency": "3.194", + "cost_per_test": "0.002233", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 2.785, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:cohere/command-r-plus:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "6.583", + "cost_per_test": "0.004306", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:cohere/command-r-plus:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 6.266, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "6.266", + "raw_stderr": "1.238", + "latency": "6.612", + "cost_per_test": "0.003262", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.238, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:cohere/command-r-plus:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 18.238, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "18.238", + "raw_stderr": "1.016", + "latency": "5.463", + "cost_per_test": "0.003267", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.016, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/cohere/command-r/1ae18662-c7c8-4344-b1e9-a5eef45b24bc.json b/data/vals_ai/cohere/command-r/1ae18662-c7c8-4344-b1e9-a5eef45b24bc.json new file mode 100644 index 0000000000000000000000000000000000000000..7b392e8e2969896bc8edc174eeccbb4693dc78d4 --- /dev/null +++ b/data/vals_ai/cohere/command-r/1ae18662-c7c8-4344-b1e9-a5eef45b24bc.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/cohere_command-r/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "command-r", + "id": "cohere/command-r", + "developer": "cohere", + "additional_details": { + "vals_model_id": "cohere/command-r", + "vals_provider": "Cohere" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:cohere/command-r:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 13.98, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "13.98", + "raw_stderr": "0.699", + "latency": "0.168", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.699, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:cohere/command-r:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 38.716, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "38.716", + "raw_stderr": "0.208", + "latency": "0.186", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.208, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:cohere/command-r:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 8.888, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "8.888", + "raw_stderr": "0.357", + "latency": "0.22", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.357, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:cohere/command-r:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 32.965, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "32.965", + "raw_stderr": "0.71", + "latency": "0.244", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 0.71, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:cohere/command-r:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 71.005, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "71.005", + "raw_stderr": "1.375", + "latency": "0.163", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 1.375, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:cohere/command-r:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 32.235, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "32.235", + "raw_stderr": "3.169", + "latency": "0.485", + "temperature": "0.3", + "max_output_tokens": "4000", + "provider": "Cohere" + }, + "uncertainty": { + "standard_error": { + "value": 3.169, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 4000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/cursor/composer-2.5/106d0425-7bd7-435c-a283-8ea3af399259.json b/data/vals_ai/cursor/composer-2.5/106d0425-7bd7-435c-a283-8ea3af399259.json new file mode 100644 index 0000000000000000000000000000000000000000..1285202b55dfebeadaef265cf19eccb44b28897f --- /dev/null +++ b/data/vals_ai/cursor/composer-2.5/106d0425-7bd7-435c-a283-8ea3af399259.json @@ -0,0 +1,96 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/cursor_composer-2.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "composer-2.5", + "id": "cursor/composer-2.5", + "developer": "cursor", + "additional_details": { + "vals_model_id": "cursor/composer-2.5", + "vals_provider": "Cursor" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:cursor/composer-2.5:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 49.614, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "49.614", + "raw_stderr": "4.972", + "latency": "592.9", + "cost_per_test": "0.502233", + "max_output_tokens": "200000", + "provider": "Cursor" + }, + "uncertainty": { + "standard_error": { + "value": 4.972, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 200000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/cursor/composer-2.5/7dbc0219-0ef1-4876-9084-d47abe3e2c98.json b/data/vals_ai/cursor/composer-2.5/7dbc0219-0ef1-4876-9084-d47abe3e2c98.json new file mode 100644 index 0000000000000000000000000000000000000000..3b6b70057c2b52122a3171cc3dbc7c6d864cf29e --- /dev/null +++ b/data/vals_ai/cursor/composer-2.5/7dbc0219-0ef1-4876-9084-d47abe3e2c98.json @@ -0,0 +1,346 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/cursor_composer-2.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "composer-2.5", + "id": "cursor/composer-2.5", + "developer": "cursor", + "additional_details": { + "vals_model_id": "cursor/composer-2.5", + "vals_provider": "Cursor" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:cursor/composer-2.5:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 52.381, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "52.381", + "raw_stderr": "7.8", + "latency": "0", + "max_output_tokens": "200000", + "provider": "Cursor" + }, + "uncertainty": { + "standard_error": { + "value": 7.8, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 200000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:cursor/composer-2.5:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 78.927, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "78.927", + "raw_stderr": "2.529", + "latency": "0", + "max_output_tokens": "200000", + "provider": "Cursor" + }, + "uncertainty": { + "standard_error": { + "value": 2.529, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 200000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:cursor/composer-2.5:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 86.598, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "86.598", + "raw_stderr": "2.452", + "latency": "0", + "max_output_tokens": "200000", + "provider": "Cursor" + }, + "uncertainty": { + "standard_error": { + "value": 2.452, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 200000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:cursor/composer-2.5:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66.667", + "raw_stderr": "33.333", + "latency": "0", + "max_output_tokens": "200000", + "provider": "Cursor" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 200000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:cursor/composer-2.5:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 79.6, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "79.6", + "raw_stderr": "1.804", + "latency": "0", + "max_output_tokens": "200000", + "provider": "Cursor" + }, + "uncertainty": { + "standard_error": { + "value": 1.804, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 200000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/cursor/composer-2.5/fcdeea78-204a-492a-be8f-04445dc24055.json b/data/vals_ai/cursor/composer-2.5/fcdeea78-204a-492a-be8f-04445dc24055.json new file mode 100644 index 0000000000000000000000000000000000000000..720a6e3648ed90ebbfc709dab52e35d22fe5f863 --- /dev/null +++ b/data/vals_ai/cursor/composer-2.5/fcdeea78-204a-492a-be8f-04445dc24055.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/cursor_composer-2.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "composer-2.5", + "id": "cursor/composer-2.5", + "developer": "cursor", + "additional_details": { + "vals_model_id": "cursor/composer-2.5", + "vals_provider": "Cursor" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:cursor/composer-2.5:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 100.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "100", + "raw_stderr": "0", + "latency": "0", + "max_output_tokens": "200000", + "provider": "Cursor" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 200000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:cursor/composer-2.5:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 40.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "40", + "raw_stderr": "9.097", + "latency": "0", + "max_output_tokens": "200000", + "provider": "Cursor" + }, + "uncertainty": { + "standard_error": { + "value": 9.097, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 200000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:cursor/composer-2.5:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 65.455, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "65.455", + "raw_stderr": "6.471", + "latency": "0", + "max_output_tokens": "200000", + "provider": "Cursor" + }, + "uncertainty": { + "standard_error": { + "value": 6.471, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 200000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:cursor/composer-2.5:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 58.427, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "58.427", + "raw_stderr": "5.254", + "latency": "0", + "max_output_tokens": "200000", + "provider": "Cursor" + }, + "uncertainty": { + "standard_error": { + "value": 5.254, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 200000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/deepseek/deepseek-v4-flash/e8e0e0e0-f881-4f7d-a408-d6f507c9ea5d.json b/data/vals_ai/deepseek/deepseek-v4-flash/e8e0e0e0-f881-4f7d-a408-d6f507c9ea5d.json new file mode 100644 index 0000000000000000000000000000000000000000..6ab0c53b797f57b6f4e1188d482adb243f934c26 --- /dev/null +++ b/data/vals_ai/deepseek/deepseek-v4-flash/e8e0e0e0-f881-4f7d-a408-d6f507c9ea5d.json @@ -0,0 +1,195 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/cyber/deepseek_deepseek-v4-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CyberBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "beta", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v4-flash", + "id": "deepseek/deepseek-v4-flash", + "developer": "deepseek", + "additional_details": { + "vals_model_id": "deepseek/deepseek-v4-flash", + "vals_provider": "DeepSeek" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "cyber:overall:deepseek/deepseek-v4-flash:score", + "evaluation_name": "vals_ai.cyber.overall", + "source_data": { + "dataset_name": "CyberBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (Overall).", + "metric_id": "vals_ai.cyber.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 66.716, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "beta", + "raw_score": "66.716", + "raw_stderr": "6.109", + "latency": "857.597", + "cost_per_test": "0.25624", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 6.109, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "cyber:patch:deepseek/deepseek-v4-flash:score", + "evaluation_name": "vals_ai.cyber.patch", + "source_data": { + "dataset_name": "CyberBench - Patch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "patch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (Patch).", + "metric_id": "vals_ai.cyber.patch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 72.414, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "patch", + "task_name": "Patch", + "dataset_type": "private", + "industry": "beta", + "raw_score": "72.414", + "raw_stderr": "5.869", + "latency": "834.079", + "cost_per_test": "0.133932", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 5.869, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "cyber:poc:deepseek/deepseek-v4-flash:score", + "evaluation_name": "vals_ai.cyber.poc", + "source_data": { + "dataset_name": "CyberBench - PoC", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "poc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (PoC).", + "metric_id": "vals_ai.cyber.poc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 61.017, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "poc", + "task_name": "PoC", + "dataset_type": "private", + "industry": "beta", + "raw_score": "61.017", + "raw_stderr": "6.349", + "latency": "881.116", + "cost_per_test": "0.378549", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 6.349, + "method": "vals_reported" + } + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/deepseek/deepseek-v4-pro/142f1fd6-37e3-40da-908a-7d880cef41f3.json b/data/vals_ai/deepseek/deepseek-v4-pro/142f1fd6-37e3-40da-908a-7d880cef41f3.json new file mode 100644 index 0000000000000000000000000000000000000000..3f8327530eca2555a9f4d274d98eac9ac0e218bd --- /dev/null +++ b/data/vals_ai/deepseek/deepseek-v4-pro/142f1fd6-37e3-40da-908a-7d880cef41f3.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/deepseek_deepseek-v4-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v4-pro", + "id": "deepseek/deepseek-v4-pro", + "developer": "deepseek", + "additional_details": { + "vals_model_id": "deepseek/deepseek-v4-pro", + "vals_provider": "DeepSeek" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 75.144, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "75.144", + "raw_stderr": "2.002", + "latency": "345.723", + "cost_per_test": "0.016146", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.002, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/deepseek/deepseek-v4-pro/18e6a9fc-49ad-4580-b6c9-14238bee9ebd.json b/data/vals_ai/deepseek/deepseek-v4-pro/18e6a9fc-49ad-4580-b6c9-14238bee9ebd.json new file mode 100644 index 0000000000000000000000000000000000000000..d4c6e55b27d1cfa3b8663a1bdf9b9fc7775799e9 --- /dev/null +++ b/data/vals_ai/deepseek/deepseek-v4-pro/18e6a9fc-49ad-4580-b6c9-14238bee9ebd.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/deepseek_deepseek-v4-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v4-pro", + "id": "deepseek/deepseek-v4-pro", + "developer": "deepseek", + "additional_details": { + "vals_model_id": "deepseek/deepseek-v4-pro", + "vals_provider": "DeepSeek" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 59.207, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.207", + "raw_stderr": "1.678", + "latency": "84.032", + "cost_per_test": "0.007781", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.678, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 63.054, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.054", + "raw_stderr": "1.648", + "latency": "79.933", + "cost_per_test": "0.316589", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.648, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.383, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.383", + "raw_stderr": "0.959", + "latency": "79.331", + "cost_per_test": "0.148855", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.959, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.888, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.888", + "raw_stderr": "1.658", + "latency": "74.027", + "cost_per_test": "0.122195", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.658, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/deepseek/deepseek-v4-pro/2b654a62-d963-4242-aa29-7002daa24251.json b/data/vals_ai/deepseek/deepseek-v4-pro/2b654a62-d963-4242-aa29-7002daa24251.json new file mode 100644 index 0000000000000000000000000000000000000000..0ed3df662108ab19ad230b55bc4d6b31dd8ce4b5 --- /dev/null +++ b/data/vals_ai/deepseek/deepseek-v4-pro/2b654a62-d963-4242-aa29-7002daa24251.json @@ -0,0 +1,1188 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_research/deepseek_deepseek-v4-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Legal Research Bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v4-pro", + "id": "deepseek/deepseek-v4-pro", + "developer": "deepseek", + "additional_details": { + "vals_model_id": "deepseek/deepseek-v4-pro", + "vals_provider": "DeepSeek" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_research:administrative_regulatory:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.legal_research.administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 28.788, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "28.788", + "raw_stderr": "5.616", + "latency": "903.624", + "cost_per_test": "0.671217", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 5.616, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:business_commercial:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.legal_research.business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 23.256, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "23.256", + "raw_stderr": "6.519", + "latency": "917.178", + "cost_per_test": "0.622362", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 6.519, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:civil_litigation:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.legal_research.civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "33.333", + "raw_stderr": "9.245", + "latency": "1095.538", + "cost_per_test": "0.736017", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 9.245, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:constitutional_civil_rights:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.legal_research.constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 13.158, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "13.158", + "raw_stderr": "5.557", + "latency": "1096.62", + "cost_per_test": "0.614384", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 5.557, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:criminal:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.legal_research.criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 15.789, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "15.789", + "raw_stderr": "4.873", + "latency": "1004.085", + "cost_per_test": "0.615927", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 4.873, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:family:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.legal_research.family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 9.091, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "9.091", + "raw_stderr": "6.273", + "latency": "1317.912", + "cost_per_test": "0.816394", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 6.273, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:health:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.legal_research.health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 46.667, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "46.667", + "raw_stderr": "9.264", + "latency": "999.255", + "cost_per_test": "0.81017", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 9.264, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:immigration:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.legal_research.immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 9.091, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "9.091", + "raw_stderr": "9.091", + "latency": "902.658", + "cost_per_test": "0.700549", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 9.091, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:overall:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.legal_research.overall", + "source_data": { + "dataset_name": "Legal Research Bench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Overall).", + "metric_id": "vals_ai.legal_research.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 23.077, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "legal", + "raw_score": "23.077", + "raw_stderr": "2.928", + "latency": "1024.619", + "cost_per_test": "0.682218", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.928, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass", + "source_data": { + "dataset_name": "Legal Research Bench - Weighted Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Weighted Pass).", + "metric_id": "vals_ai.legal_research.weighted_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 80.689, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass", + "task_name": "Weighted Pass", + "dataset_type": "private", + "industry": "legal", + "raw_score": "80.689", + "raw_stderr": "1.537", + "latency": "1024.619", + "cost_per_test": "0.682218", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.537, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__administrative_regulatory:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.weighted_pass__administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 83.372, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "83.372", + "raw_stderr": "2.545", + "latency": "903.624", + "cost_per_test": "0.671217", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.545, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__business_commercial:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.weighted_pass__business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 86.491, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "86.491", + "raw_stderr": "2.924", + "latency": "917.178", + "cost_per_test": "0.622362", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.924, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__civil_litigation:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.weighted_pass__civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 85.96, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "85.96", + "raw_stderr": "3.534", + "latency": "1095.538", + "cost_per_test": "0.736017", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 3.534, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__constitutional_civil_rights:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 85.744, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "85.744", + "raw_stderr": "2.309", + "latency": "1096.62", + "cost_per_test": "0.614384", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.309, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__criminal:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.weighted_pass__criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 76.39, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "76.39", + "raw_stderr": "2.887", + "latency": "1004.085", + "cost_per_test": "0.615927", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.887, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__family:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.weighted_pass__family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 66.252, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "66.252", + "raw_stderr": "5.618", + "latency": "1317.912", + "cost_per_test": "0.816394", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 5.618, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__health:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.weighted_pass__health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 91.229, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "91.229", + "raw_stderr": "2.203", + "latency": "999.255", + "cost_per_test": "0.81017", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.203, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__immigration:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.weighted_pass__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 75.593, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "75.593", + "raw_stderr": "8.908", + "latency": "902.658", + "cost_per_test": "0.700549", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 8.908, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/deepseek/deepseek-v4-pro/3c79a3c2-9e5e-4a21-a5d6-0af99e089979.json b/data/vals_ai/deepseek/deepseek-v4-pro/3c79a3c2-9e5e-4a21-a5d6-0af99e089979.json new file mode 100644 index 0000000000000000000000000000000000000000..9d3320eb0a5f7ed69e5e00848919d921fd8aa320 --- /dev/null +++ b/data/vals_ai/deepseek/deepseek-v4-pro/3c79a3c2-9e5e-4a21-a5d6-0af99e089979.json @@ -0,0 +1,371 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/deepseek_deepseek-v4-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v4-pro", + "id": "deepseek/deepseek-v4-pro", + "developer": "deepseek", + "additional_details": { + "vals_model_id": "deepseek/deepseek-v4-pro", + "vals_provider": "DeepSeek" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 45.238, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "45.238", + "raw_stderr": "7.773", + "latency": "1048.718", + "cost_per_test": "0", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 7.773, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 75.862, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "75.862", + "raw_stderr": "2.654", + "latency": "716.772", + "cost_per_test": "0", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.654, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 87.629, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "87.629", + "raw_stderr": "2.37", + "latency": "430.413", + "cost_per_test": "0", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.37, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "893.873", + "cost_per_test": "0", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:overall:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 77.4, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "77.4", + "raw_stderr": "1.872", + "latency": "634.611", + "cost_per_test": "0.436763", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.872, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/deepseek/deepseek-v4-pro/51279a24-f1a5-4468-879b-e69cb5e3441b.json b/data/vals_ai/deepseek/deepseek-v4-pro/51279a24-f1a5-4468-879b-e69cb5e3441b.json new file mode 100644 index 0000000000000000000000000000000000000000..02ec8936d28548510b04cebf74d02347e2f779f2 --- /dev/null +++ b/data/vals_ai/deepseek/deepseek-v4-pro/51279a24-f1a5-4468-879b-e69cb5e3441b.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/deepseek_deepseek-v4-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v4-pro", + "id": "deepseek/deepseek-v4-pro", + "developer": "deepseek", + "additional_details": { + "vals_model_id": "deepseek/deepseek-v4-pro", + "vals_provider": "DeepSeek" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 49.931, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "49.931", + "raw_stderr": "4.767", + "latency": "3418.92", + "cost_per_test": "2.205991", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 4.767, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/deepseek/deepseek-v4-pro/5cf5db02-8bbf-4a1a-80b9-f24c44a5e5be.json b/data/vals_ai/deepseek/deepseek-v4-pro/5cf5db02-8bbf-4a1a-80b9-f24c44a5e5be.json new file mode 100644 index 0000000000000000000000000000000000000000..4b73af2529c52b71d9b9d1da1b58f9482d37d1b7 --- /dev/null +++ b/data/vals_ai/deepseek/deepseek-v4-pro/5cf5db02-8bbf-4a1a-80b9-f24c44a5e5be.json @@ -0,0 +1,676 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/emb/deepseek_deepseek-v4-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - EMB", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v4-pro", + "id": "deepseek/deepseek-v4-pro", + "developer": "deepseek", + "additional_details": { + "vals_model_id": "deepseek/deepseek-v4-pro", + "vals_provider": "DeepSeek" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "emb:comps:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.emb.comps", + "source_data": { + "dataset_name": "EMB - Comps", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "comps", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Comps).", + "metric_id": "vals_ai.emb.comps.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 57.434, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "comps", + "task_name": "Comps", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.434", + "raw_stderr": "6.116", + "latency": "867.448", + "cost_per_test": "1.747894", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 6.116, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:dataroom:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.emb.dataroom", + "source_data": { + "dataset_name": "EMB - Dataroom", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dataroom", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Dataroom).", + "metric_id": "vals_ai.emb.dataroom.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 62.572, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dataroom", + "task_name": "Dataroom", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.572", + "raw_stderr": "9.249", + "latency": "745.646", + "cost_per_test": "1.592812", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 9.249, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:dcf:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.emb.dcf", + "source_data": { + "dataset_name": "EMB - DCF", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dcf", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (DCF).", + "metric_id": "vals_ai.emb.dcf.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 44.435, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dcf", + "task_name": "DCF", + "dataset_type": "private", + "industry": "finance", + "raw_score": "44.435", + "raw_stderr": "6.242", + "latency": "816.693", + "cost_per_test": "1.114767", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 6.242, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:lbo:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.emb.lbo", + "source_data": { + "dataset_name": "EMB - LBO", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "lbo", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (LBO).", + "metric_id": "vals_ai.emb.lbo.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 43.938, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "lbo", + "task_name": "LBO", + "dataset_type": "private", + "industry": "finance", + "raw_score": "43.938", + "raw_stderr": "3.924", + "latency": "1042.97", + "cost_per_test": "1.640821", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 3.924, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:ma:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.emb.ma", + "source_data": { + "dataset_name": "EMB - M&A", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "ma", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (M&A).", + "metric_id": "vals_ai.emb.ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 48.156, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "ma", + "task_name": "M&A", + "dataset_type": "private", + "industry": "finance", + "raw_score": "48.156", + "raw_stderr": "9.249", + "latency": "764.843", + "cost_per_test": "1.076538", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 9.249, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:misc:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.emb.misc", + "source_data": { + "dataset_name": "EMB - Misc", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "misc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Misc).", + "metric_id": "vals_ai.emb.misc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 57.647, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "misc", + "task_name": "Misc", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.647", + "raw_stderr": "6.554", + "latency": "811.882", + "cost_per_test": "1.229579", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 6.554, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:operating:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.emb.operating", + "source_data": { + "dataset_name": "EMB - Operating", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "operating", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Operating).", + "metric_id": "vals_ai.emb.operating.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 47.176, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "operating", + "task_name": "Operating", + "dataset_type": "private", + "industry": "finance", + "raw_score": "47.176", + "raw_stderr": "6.647", + "latency": "685.554", + "cost_per_test": "0.875142", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 6.647, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:overall:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.emb.overall", + "source_data": { + "dataset_name": "EMB - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Overall).", + "metric_id": "vals_ai.emb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 51.623, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "51.623", + "raw_stderr": "3.053", + "latency": "822.144", + "cost_per_test": "1.332885", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 3.053, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:scratch:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.emb.scratch", + "source_data": { + "dataset_name": "EMB - Scratch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "scratch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Scratch).", + "metric_id": "vals_ai.emb.scratch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 56.692, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "scratch", + "task_name": "Scratch", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.692", + "raw_stderr": "2.818", + "latency": "866.492", + "cost_per_test": "1.275379", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.818, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:template:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.emb.template", + "source_data": { + "dataset_name": "EMB - Template", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "template", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Template).", + "metric_id": "vals_ai.emb.template.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 46.07, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "template", + "task_name": "Template", + "dataset_type": "private", + "industry": "finance", + "raw_score": "46.07", + "raw_stderr": "4.377", + "latency": "777.796", + "cost_per_test": "1.390391", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 4.377, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/deepseek/deepseek-v4-pro/5db13933-3321-4560-ad43-f011767b4dc1.json b/data/vals_ai/deepseek/deepseek-v4-pro/5db13933-3321-4560-ad43-f011767b4dc1.json new file mode 100644 index 0000000000000000000000000000000000000000..d1cc700324aa4ff71b145db080a0d74c27602da1 --- /dev/null +++ b/data/vals_ai/deepseek/deepseek-v4-pro/5db13933-3321-4560-ad43-f011767b4dc1.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/programbench/deepseek_deepseek-v4-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProgramBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v4-pro", + "id": "deepseek/deepseek-v4-pro", + "developer": "deepseek", + "additional_details": { + "vals_model_id": "deepseek/deepseek-v4-pro", + "vals_provider": "DeepSeek" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "programbench:almost:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.programbench.almost", + "source_data": { + "dataset_name": "ProgramBench - Almost Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "almost", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Almost Resolved).", + "metric_id": "vals_ai.programbench.almost.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "almost", + "task_name": "Almost Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "3778.098", + "cost_per_test": "2.80544", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "programbench:overall:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.programbench.overall", + "source_data": { + "dataset_name": "ProgramBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Overall).", + "metric_id": "vals_ai.programbench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "3778.098", + "cost_per_test": "2.80544", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "programbench:partial:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.programbench.partial", + "source_data": { + "dataset_name": "ProgramBench - Raw Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "partial", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Raw Pass Rate).", + "metric_id": "vals_ai.programbench.partial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 47.778, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "partial", + "task_name": "Raw Pass Rate", + "dataset_type": "public", + "industry": "coding", + "raw_score": "47.778", + "raw_stderr": "1.636", + "latency": "3778.098", + "cost_per_test": "2.80544", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.636, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "programbench:strict:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.programbench.strict", + "source_data": { + "dataset_name": "ProgramBench - Fully Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "strict", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Fully Resolved).", + "metric_id": "vals_ai.programbench.strict.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "strict", + "task_name": "Fully Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "3778.098", + "cost_per_test": "2.80544", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/deepseek/deepseek-v4-pro/5f4fee40-c7a4-47d6-84df-9c139c533174.json b/data/vals_ai/deepseek/deepseek-v4-pro/5f4fee40-c7a4-47d6-84df-9c139c533174.json new file mode 100644 index 0000000000000000000000000000000000000000..5e920c73300fc753d973acb324fea4cdf93a1513 --- /dev/null +++ b/data/vals_ai/deepseek/deepseek-v4-pro/5f4fee40-c7a4-47d6-84df-9c139c533174.json @@ -0,0 +1,3386 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/hlab/deepseek_deepseek-v4-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Harvey's Legal Agent Benchmark", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v4-pro", + "id": "deepseek/deepseek-v4-pro", + "developer": "deepseek", + "additional_details": { + "vals_model_id": "deepseek/deepseek-v4-pro", + "vals_provider": "DeepSeek" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "hlab:antitrust_competition:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "409.396", + "cost_per_test": "0.8503", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:arbitration_international_dispute_resolution:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "395.848", + "cost_per_test": "0.887292", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:banking_finance:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "311.901", + "cost_per_test": "0.609606", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:bankruptcy_restructuring:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "377.691", + "cost_per_test": "0.705278", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:capital_markets:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "350.408", + "cost_per_test": "0.764042", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:corporate_governance:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "355.346", + "cost_per_test": "0.585215", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:corporate_ma:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "451.454", + "cost_per_test": "0.855036", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Criteria Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Criteria Pass Rate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 84.397, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate", + "task_name": "Criteria Pass Rate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.397", + "raw_stderr": "0.437", + "latency": "401.527", + "cost_per_test": "0.737465", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.437, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__antitrust_competition:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 83.696, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.696", + "raw_stderr": "2.228", + "latency": "409.396", + "cost_per_test": "0.8503", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.228, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__arbitration_international_dispute_resolution:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 83.512, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.512", + "raw_stderr": "2.226", + "latency": "395.848", + "cost_per_test": "0.887292", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.226, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__banking_finance:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 86.245, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.245", + "raw_stderr": "2.104", + "latency": "311.901", + "cost_per_test": "0.609606", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.104, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__bankruptcy_restructuring:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 76.125, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "76.125", + "raw_stderr": "2.512", + "latency": "377.691", + "cost_per_test": "0.705278", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.512, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__capital_markets:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 86.274, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.274", + "raw_stderr": "2.159", + "latency": "350.408", + "cost_per_test": "0.764042", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.159, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_governance:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 85.959, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.959", + "raw_stderr": "2.037", + "latency": "355.346", + "cost_per_test": "0.585215", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.037, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_ma:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 88.462, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.462", + "raw_stderr": "1.812", + "latency": "451.454", + "cost_per_test": "0.855036", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.812, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__data_privacy_cybersecurity:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 85.866, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.866", + "raw_stderr": "2.075", + "latency": "375.351", + "cost_per_test": "0.741361", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.075, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__emerging_companies_venture_capital:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 87.372, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.372", + "raw_stderr": "1.944", + "latency": "415.085", + "cost_per_test": "0.76098", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.944, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__employment_labor:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 74.906, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "74.906", + "raw_stderr": "2.658", + "latency": "465.325", + "cost_per_test": "0.691985", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.658, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__energy_natural_resources:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 93.399, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "93.399", + "raw_stderr": "1.429", + "latency": "388.472", + "cost_per_test": "0.983444", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.429, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__environmental_esg:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 83.871, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.871", + "raw_stderr": "2.206", + "latency": "412.296", + "cost_per_test": "0.64771", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.206, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__funds_asset_management:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 80.758, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.758", + "raw_stderr": "2.132", + "latency": "523.881", + "cost_per_test": "0.955566", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.132, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__healthcare_life_sciences:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 86.139, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.139", + "raw_stderr": "1.988", + "latency": "435.381", + "cost_per_test": "0.848233", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.988, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__immigration:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 84.364, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.364", + "raw_stderr": "2.194", + "latency": "333.847", + "cost_per_test": "0.519471", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.194, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__insurance:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 80.586, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.586", + "raw_stderr": "2.398", + "latency": "405.384", + "cost_per_test": "0.628366", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.398, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__intellectual_property:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 88.339, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.339", + "raw_stderr": "1.911", + "latency": "380.943", + "cost_per_test": "0.704723", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.911, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__international_trade_sanctions:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 82.936, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.936", + "raw_stderr": "2.374", + "latency": "461.643", + "cost_per_test": "0.642446", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.374, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__litigation_dispute_resolution:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 84.644, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.644", + "raw_stderr": "2.211", + "latency": "344.871", + "cost_per_test": "0.597517", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.211, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__real_estate:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 89.213, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.213", + "raw_stderr": "1.677", + "latency": "415.105", + "cost_per_test": "0.804107", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.677, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__structured_finance_securitization:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 82.222, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.222", + "raw_stderr": "2.158", + "latency": "393.298", + "cost_per_test": "0.786036", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.158, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__tax:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 77.926, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.926", + "raw_stderr": "2.403", + "latency": "381.65", + "cost_per_test": "0.767561", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.403, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__trusts_estates_private_client:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 88.168, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.168", + "raw_stderr": "1.999", + "latency": "462.126", + "cost_per_test": "0.765564", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.999, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__white_collar_defense_investigations:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 83.502, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.502", + "raw_stderr": "2.157", + "latency": "389.949", + "cost_per_test": "0.597321", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.157, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:data_privacy_cybersecurity:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "375.351", + "cost_per_test": "0.741361", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:emerging_companies_venture_capital:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "415.085", + "cost_per_test": "0.76098", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:employment_labor:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "465.325", + "cost_per_test": "0.691985", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:energy_natural_resources:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "388.472", + "cost_per_test": "0.983444", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:environmental_esg:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "412.296", + "cost_per_test": "0.64771", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:funds_asset_management:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "523.881", + "cost_per_test": "0.955566", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:healthcare_life_sciences:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "435.381", + "cost_per_test": "0.848233", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:immigration:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "333.847", + "cost_per_test": "0.519471", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:insurance:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "405.384", + "cost_per_test": "0.628366", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:intellectual_property:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "380.943", + "cost_per_test": "0.704723", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:international_trade_sanctions:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "461.643", + "cost_per_test": "0.642446", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:litigation_dispute_resolution:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "344.871", + "cost_per_test": "0.597517", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:overall:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.overall", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Overall).", + "metric_id": "vals_ai.hlab.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 3.75, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "3.75", + "raw_stderr": "1.431", + "latency": "401.527", + "cost_per_test": "0.737465", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.431, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:real_estate:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "415.105", + "cost_per_test": "0.804107", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:structured_finance_securitization:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "393.298", + "cost_per_test": "0.786036", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:tax:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "381.65", + "cost_per_test": "0.767561", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:trusts_estates_private_client:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "462.126", + "cost_per_test": "0.765564", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:white_collar_defense_investigations:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.hlab.white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "389.949", + "cost_per_test": "0.597321", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/deepseek/deepseek-v4-pro/79cd01f5-5512-4681-838a-80cc39579e7d.json b/data/vals_ai/deepseek/deepseek-v4-pro/79cd01f5-5512-4681-838a-80cc39579e7d.json new file mode 100644 index 0000000000000000000000000000000000000000..370a56e8d22e6b236e2dfa824a7a6ea5dceba6ef --- /dev/null +++ b/data/vals_ai/deepseek/deepseek-v4-pro/79cd01f5-5512-4681-838a-80cc39579e7d.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/deepseek_deepseek-v4-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v4-pro", + "id": "deepseek/deepseek-v4-pro", + "developer": "deepseek", + "additional_details": { + "vals_model_id": "deepseek/deepseek-v4-pro", + "vals_provider": "DeepSeek" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 60.589, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.589", + "raw_stderr": "1.397", + "latency": "189.695", + "cost_per_test": "0.002185", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.397, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 72.077, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "72.077", + "raw_stderr": "0.877", + "latency": "189.695", + "cost_per_test": "0.002185", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.877, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 83.565, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "83.565", + "raw_stderr": "1.06", + "latency": "189.695", + "cost_per_test": "0.002185", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.06, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/deepseek/deepseek-v4-pro/7a8a79ab-de09-4a92-899b-f2e58db68fe5.json b/data/vals_ai/deepseek/deepseek-v4-pro/7a8a79ab-de09-4a92-899b-f2e58db68fe5.json new file mode 100644 index 0000000000000000000000000000000000000000..28db1f17323faff6e4129121e5f500d9ea0eda65 --- /dev/null +++ b/data/vals_ai/deepseek/deepseek-v4-pro/7a8a79ab-de09-4a92-899b-f2e58db68fe5.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/deepseek_deepseek-v4-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v4-pro", + "id": "deepseek/deepseek-v4-pro", + "developer": "deepseek", + "additional_details": { + "vals_model_id": "deepseek/deepseek-v4-pro", + "vals_provider": "DeepSeek" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 10.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "10", + "raw_stderr": "3", + "latency": "518.756", + "cost_per_test": "0", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 3.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/deepseek/deepseek-v4-pro/8a181540-8824-43b5-b5ee-926fec569a37.json b/data/vals_ai/deepseek/deepseek-v4-pro/8a181540-8824-43b5-b5ee-926fec569a37.json new file mode 100644 index 0000000000000000000000000000000000000000..c4bee17ef9c04cb81e0cb4498838f92df26b858c --- /dev/null +++ b/data/vals_ai/deepseek/deepseek-v4-pro/8a181540-8824-43b5-b5ee-926fec569a37.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/deepseek_deepseek-v4-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v4-pro", + "id": "deepseek/deepseek-v4-pro", + "developer": "deepseek", + "additional_details": { + "vals_model_id": "deepseek/deepseek-v4-pro", + "vals_provider": "DeepSeek" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 31.333, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "31.333", + "raw_stderr": "10.426", + "latency": "5262.744", + "cost_per_test": "0.580726", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 10.426, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 40.333, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "40.333", + "raw_stderr": "9.754", + "latency": "5485.007", + "cost_per_test": "0.556402", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 9.754, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "ioi:overall:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 35.833, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "35.833", + "raw_stderr": "7.139", + "latency": "5373.876", + "cost_per_test": "0.568564", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 7.139, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/deepseek/deepseek-v4-pro/8f5789c8-c8bb-431a-8e5c-9d788d0cf264.json b/data/vals_ai/deepseek/deepseek-v4-pro/8f5789c8-c8bb-431a-8e5c-9d788d0cf264.json new file mode 100644 index 0000000000000000000000000000000000000000..4e00fe326f3dc62cbd4fce2919e7d3b020639c39 --- /dev/null +++ b/data/vals_ai/deepseek/deepseek-v4-pro/8f5789c8-c8bb-431a-8e5c-9d788d0cf264.json @@ -0,0 +1,408 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/deepseek_deepseek-v4-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v4-pro", + "id": "deepseek/deepseek-v4-pro", + "developer": "deepseek", + "additional_details": { + "vals_model_id": "deepseek/deepseek-v4-pro", + "vals_provider": "DeepSeek" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 61.888, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "61.888", + "raw_stderr": "1.658", + "latency": "74.027", + "cost_per_test": "0.148855", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.658, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 44.081, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "44.081", + "raw_stderr": "0.767", + "latency": "1450.039", + "cost_per_test": "1.100162", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.767, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_index:overall:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 55.615, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "55.615", + "raw_stderr": "1.685", + "latency": "1319.052", + "cost_per_test": "0.826192", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.685, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 75.49, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "75.49", + "raw_stderr": "4.259", + "latency": "574.117", + "cost_per_test": "0.436763" + }, + "uncertainty": { + "standard_error": { + "value": 4.259, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 50.187, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "50.187", + "raw_stderr": "1.498", + "latency": "1078.158", + "cost_per_test": "0.239191", + "max_output_tokens": "128000", + "reasoning": "false", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.498, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "false" + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 55.905, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "55.905", + "raw_stderr": "7.421", + "latency": "3418.92", + "cost_per_test": "2.205991", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 7.421, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/deepseek/deepseek-v4-pro/961ec8f8-2a5f-41e8-b4c7-e6d42547ef02.json b/data/vals_ai/deepseek/deepseek-v4-pro/961ec8f8-2a5f-41e8-b4c7-e6d42547ef02.json new file mode 100644 index 0000000000000000000000000000000000000000..1b6b04676e3417a079eb878e0d3ea2fa87030018 --- /dev/null +++ b/data/vals_ai/deepseek/deepseek-v4-pro/961ec8f8-2a5f-41e8-b4c7-e6d42547ef02.json @@ -0,0 +1,195 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/cyber/deepseek_deepseek-v4-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CyberBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "beta", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v4-pro", + "id": "deepseek/deepseek-v4-pro", + "developer": "deepseek", + "additional_details": { + "vals_model_id": "deepseek/deepseek-v4-pro", + "vals_provider": "DeepSeek" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "cyber:overall:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.cyber.overall", + "source_data": { + "dataset_name": "CyberBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (Overall).", + "metric_id": "vals_ai.cyber.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 68.644, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "beta", + "raw_score": "68.644", + "raw_stderr": "5.837", + "latency": "1120.199", + "cost_per_test": "1.620525", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 5.837, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "cyber:patch:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.cyber.patch", + "source_data": { + "dataset_name": "CyberBench - Patch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "patch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (Patch).", + "metric_id": "vals_ai.cyber.patch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 79.661, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "patch", + "task_name": "Patch", + "dataset_type": "private", + "industry": "beta", + "raw_score": "79.661", + "raw_stderr": "5.24", + "latency": "1102.852", + "cost_per_test": "0.888072", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 5.24, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "cyber:poc:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.cyber.poc", + "source_data": { + "dataset_name": "CyberBench - PoC", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "poc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (PoC).", + "metric_id": "vals_ai.cyber.poc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 57.627, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "poc", + "task_name": "PoC", + "dataset_type": "private", + "industry": "beta", + "raw_score": "57.627", + "raw_stderr": "6.433", + "latency": "1137.546", + "cost_per_test": "2.352979", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 6.433, + "method": "vals_reported" + } + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/deepseek/deepseek-v4-pro/b58c0ae9-5050-4ca2-a12f-8ae1234561cf.json b/data/vals_ai/deepseek/deepseek-v4-pro/b58c0ae9-5050-4ca2-a12f-8ae1234561cf.json new file mode 100644 index 0000000000000000000000000000000000000000..135aa8f5acd66c304d016134fe26a0627597f2cb --- /dev/null +++ b/data/vals_ai/deepseek/deepseek-v4-pro/b58c0ae9-5050-4ca2-a12f-8ae1234561cf.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/public-benefits-bench/deepseek_deepseek-v4-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Public Benefits Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "social mobility", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v4-pro", + "id": "deepseek/deepseek-v4-pro", + "developer": "deepseek", + "additional_details": { + "vals_model_id": "deepseek/deepseek-v4-pro", + "vals_provider": "DeepSeek" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "public-benefits-bench:overall:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.public-benefits-bench.overall", + "source_data": { + "dataset_name": "Public Benefits Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Public Benefits Bench v1.1 (Overall).", + "metric_id": "vals_ai.public-benefits-bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "score_details": { + "score": 62.923, + "details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "social mobility", + "raw_score": "62.923", + "raw_stderr": "1.256", + "latency": "1058.649", + "cost_per_test": "0.460453", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.256, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/deepseek/deepseek-v4-pro/b8288823-4e07-46c3-85cd-cf5de6fee7e4.json b/data/vals_ai/deepseek/deepseek-v4-pro/b8288823-4e07-46c3-85cd-cf5de6fee7e4.json new file mode 100644 index 0000000000000000000000000000000000000000..dbebd2f1acf2973271f8f419cadb3138f99976a7 --- /dev/null +++ b/data/vals_ai/deepseek/deepseek-v4-pro/b8288823-4e07-46c3-85cd-cf5de6fee7e4.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/deepseek_deepseek-v4-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v4-pro", + "id": "deepseek/deepseek-v4-pro", + "developer": "deepseek", + "additional_details": { + "vals_model_id": "deepseek/deepseek-v4-pro", + "vals_provider": "DeepSeek" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66.667", + "raw_stderr": "8.333", + "latency": "23989.013", + "cost_per_test": "5.322002", + "max_output_tokens": "128000", + "reasoning": "false", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 8.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "false" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 31.111, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "31.111", + "raw_stderr": "2.222", + "latency": "3198.535", + "cost_per_test": "0.7096", + "max_output_tokens": "128000", + "reasoning": "false", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.222, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "false" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 59.394, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "59.394", + "raw_stderr": "1.603", + "latency": "1744.656", + "cost_per_test": "0.387055", + "max_output_tokens": "128000", + "reasoning": "false", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.603, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "false" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 50.187, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "50.187", + "raw_stderr": "1.498", + "latency": "1078.158", + "cost_per_test": "0.239191", + "max_output_tokens": "128000", + "reasoning": "false", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.498, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning": "false" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/deepseek/deepseek-v4-pro/c94e9ddb-6615-4945-a40b-dff11eeef1b8.json b/data/vals_ai/deepseek/deepseek-v4-pro/c94e9ddb-6615-4945-a40b-dff11eeef1b8.json new file mode 100644 index 0000000000000000000000000000000000000000..1034040666fa0622ce233667b70c8c40867e232e --- /dev/null +++ b/data/vals_ai/deepseek/deepseek-v4-pro/c94e9ddb-6615-4945-a40b-dff11eeef1b8.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/deepseek_deepseek-v4-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v4-pro", + "id": "deepseek/deepseek-v4-pro", + "developer": "deepseek", + "additional_details": { + "vals_model_id": "deepseek/deepseek-v4-pro", + "vals_provider": "DeepSeek" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.758, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.758", + "raw_stderr": "0.617", + "latency": "73.491", + "cost_per_test": "0.009784", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.617, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "lcb:hard:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 72.571, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "72.571", + "raw_stderr": "2.385", + "latency": "1487.685", + "cost_per_test": "0.20727", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.385, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "lcb:medium:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 91.123, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "91.123", + "raw_stderr": "1.453", + "latency": "585.336", + "cost_per_test": "0.073297", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.453, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "lcb:overall:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 87.484, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "87.484", + "raw_stderr": "0.953", + "latency": "715.504", + "cost_per_test": "0.096784", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.953, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/deepseek/deepseek-v4-pro/d8079aed-9338-4d7e-9e1c-3e03a505b579.json b/data/vals_ai/deepseek/deepseek-v4-pro/d8079aed-9338-4d7e-9e1c-3e03a505b579.json new file mode 100644 index 0000000000000000000000000000000000000000..ae8c6b705b3edf0a5e4342fa6a3ca0c4fe9990b0 --- /dev/null +++ b/data/vals_ai/deepseek/deepseek-v4-pro/d8079aed-9338-4d7e-9e1c-3e03a505b579.json @@ -0,0 +1,740 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/deepseek_deepseek-v4-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v4-pro", + "id": "deepseek/deepseek-v4-pro", + "developer": "deepseek", + "additional_details": { + "vals_model_id": "deepseek/deepseek-v4-pro", + "vals_provider": "DeepSeek" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 41.586, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "41.586", + "raw_stderr": "1.185", + "latency": "1486.77", + "cost_per_test": "1.170222", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.185, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 31.448, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "31.448", + "raw_stderr": "0.715", + "latency": "1526.183", + "cost_per_test": "1.128416", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.715, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 39.632, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "39.632", + "raw_stderr": "1.323", + "latency": "1805.003", + "cost_per_test": "1.504718", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.323, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 52.649, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.649", + "raw_stderr": "2.117", + "latency": "1666.135", + "cost_per_test": "1.375749", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.117, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 59.84, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.84", + "raw_stderr": "1.503", + "latency": "1036.974", + "cost_per_test": "0.424147", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.503, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 10.548, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "10.548", + "raw_stderr": "2.633", + "latency": "1617.918", + "cost_per_test": "1.066968", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.633, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 60.341, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.341", + "raw_stderr": "1.505", + "latency": "1682.108", + "cost_per_test": "1.190101", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.505, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 65.75, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.75", + "raw_stderr": "1.396", + "latency": "1415.095", + "cost_per_test": "1.225772", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.396, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 50.801, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "50.801", + "raw_stderr": "1.742", + "latency": "889.762", + "cost_per_test": "0.526255", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.742, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:overall:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 44.083, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "44.083", + "raw_stderr": "0.653", + "latency": "1526.183", + "cost_per_test": "1.128416", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.653, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 15.6, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "15.6", + "raw_stderr": "0.961", + "latency": "2630.599", + "cost_per_test": "1.701832", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.961, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/deepseek/deepseek-v4-pro/e1ecf2cc-055f-438d-a928-3593be5ebb19.json b/data/vals_ai/deepseek/deepseek-v4-pro/e1ecf2cc-055f-438d-a928-3593be5ebb19.json new file mode 100644 index 0000000000000000000000000000000000000000..f935e22c453eb8c4079f73cb0c49006aea8b3982 --- /dev/null +++ b/data/vals_ai/deepseek/deepseek-v4-pro/e1ecf2cc-055f-438d-a928-3593be5ebb19.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/deepseek_deepseek-v4-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v4-pro", + "id": "deepseek/deepseek-v4-pro", + "developer": "deepseek", + "additional_details": { + "vals_model_id": "deepseek/deepseek-v4-pro", + "vals_provider": "DeepSeek" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 90.404, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.404", + "raw_stderr": "2.277", + "latency": "823.036", + "cost_per_test": "0.101321", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.277, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 88.384, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.384", + "raw_stderr": "2.4", + "latency": "977.721", + "cost_per_test": "0.122206", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.4, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 89.394, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.394", + "raw_stderr": "1.654", + "latency": "900.378", + "cost_per_test": "0.111764", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.654, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/deepseek/deepseek-v4-pro/e6c9d8e4-392e-4e6f-9c72-c1af19e47d5f.json b/data/vals_ai/deepseek/deepseek-v4-pro/e6c9d8e4-392e-4e6f-9c72-c1af19e47d5f.json new file mode 100644 index 0000000000000000000000000000000000000000..492e5a63ca29f11e718a0d7b061e391760e66d26 --- /dev/null +++ b/data/vals_ai/deepseek/deepseek-v4-pro/e6c9d8e4-392e-4e6f-9c72-c1af19e47d5f.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/deepseek_deepseek-v4-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v4-pro", + "id": "deepseek/deepseek-v4-pro", + "developer": "deepseek", + "additional_details": { + "vals_model_id": "deepseek/deepseek-v4-pro", + "vals_provider": "DeepSeek" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.887, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.887", + "raw_stderr": "0.96", + "latency": "636.88", + "cost_per_test": "0.01296", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.96, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.888, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.888", + "raw_stderr": "0.972", + "latency": "399.135", + "cost_per_test": "0.025086", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.972, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.491, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.491", + "raw_stderr": "0.787", + "latency": "588.835", + "cost_per_test": "0.03771", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.787, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.024, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.024", + "raw_stderr": "1.559", + "latency": "753.864", + "cost_per_test": "0.028258", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.559, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.863, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.863", + "raw_stderr": "1.083", + "latency": "373.31", + "cost_per_test": "0.013878", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.083, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.687, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.687", + "raw_stderr": "1.139", + "latency": "877.3", + "cost_per_test": "0.089616", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.139, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.396, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.396", + "raw_stderr": "1.342", + "latency": "678.193", + "cost_per_test": "0.015574", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.342, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.69, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.69", + "raw_stderr": "2.193", + "latency": "635.878", + "cost_per_test": "0.012817", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 2.193, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.024, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.024", + "raw_stderr": "1.345", + "latency": "535.644", + "cost_per_test": "0.027025", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.345, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 95.337, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "95.337", + "raw_stderr": "0.578", + "latency": "474.812", + "cost_per_test": "0.022501", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.578, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.199, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.199", + "raw_stderr": "1.2", + "latency": "420.745", + "cost_per_test": "0.013436", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.2, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.249, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.249", + "raw_stderr": "0.335", + "latency": "577.617", + "cost_per_test": "0.025579", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.335, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.569, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.569", + "raw_stderr": "1.617", + "latency": "657.825", + "cost_per_test": "0.014196", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.617, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.457, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.457", + "raw_stderr": "0.701", + "latency": "561.23", + "cost_per_test": "0.035315", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.701, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.97, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.97", + "raw_stderr": "1.167", + "latency": "492.984", + "cost_per_test": "0.009737", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.167, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/deepseek/deepseek-v4-pro/f3070bc2-d95f-40fe-8835-89b246b53316.json b/data/vals_ai/deepseek/deepseek-v4-pro/f3070bc2-d95f-40fe-8835-89b246b53316.json new file mode 100644 index 0000000000000000000000000000000000000000..c45c4c836505bd43e5d6db94ac55dafa9857350f --- /dev/null +++ b/data/vals_ai/deepseek/deepseek-v4-pro/f3070bc2-d95f-40fe-8835-89b246b53316.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/code-migration/deepseek_deepseek-v4-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Code Migration", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v4-pro", + "id": "deepseek/deepseek-v4-pro", + "developer": "deepseek", + "additional_details": { + "vals_model_id": "deepseek/deepseek-v4-pro", + "vals_provider": "DeepSeek" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "code-migration:cli:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.code-migration.cli", + "source_data": { + "dataset_name": "Code Migration - CLI", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cli", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (CLI).", + "metric_id": "vals_ai.code-migration.cli.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 15.47, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cli", + "task_name": "CLI", + "dataset_type": "private", + "industry": "coding", + "raw_score": "15.47", + "raw_stderr": "1.689", + "latency": "1746.505", + "cost_per_test": "2.486718", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.689, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "code-migration:cobol:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.code-migration.cobol", + "source_data": { + "dataset_name": "Code Migration - COBOL", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cobol", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (COBOL).", + "metric_id": "vals_ai.code-migration.cobol.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 58.393, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cobol", + "task_name": "COBOL", + "dataset_type": "private", + "industry": "coding", + "raw_score": "58.393", + "raw_stderr": "15.334", + "latency": "1712.395", + "cost_per_test": "2.058485", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 15.334, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "code-migration:code_quality:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.code-migration.code_quality", + "source_data": { + "dataset_name": "Code Migration - Code Quality", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "code_quality", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Code Quality).", + "metric_id": "vals_ai.code-migration.code_quality.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 3.401, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "code_quality", + "task_name": "Code Quality", + "dataset_type": "private", + "industry": "coding", + "raw_score": "3.401", + "raw_stderr": "0.036", + "latency": "1737.978", + "cost_per_test": "2.37966", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.036, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "code-migration:overall:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.code-migration.overall", + "source_data": { + "dataset_name": "Code Migration - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Overall).", + "metric_id": "vals_ai.code-migration.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 26.201, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "26.201", + "raw_stderr": "4.037", + "latency": "1737.978", + "cost_per_test": "2.37966", + "max_output_tokens": "384000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 4.037, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 384000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/deepseek/deepseek-v4-pro/f560a3bf-b7bb-4765-81e1-22b8e4f532ba.json b/data/vals_ai/deepseek/deepseek-v4-pro/f560a3bf-b7bb-4765-81e1-22b8e4f532ba.json new file mode 100644 index 0000000000000000000000000000000000000000..d99e261a2b0b0d2519866b1d3f8dff6afa7f6c54 --- /dev/null +++ b/data/vals_ai/deepseek/deepseek-v4-pro/f560a3bf-b7bb-4765-81e1-22b8e4f532ba.json @@ -0,0 +1,438 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/deepseek_deepseek-v4-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v4-pro", + "id": "deepseek/deepseek-v4-pro", + "developer": "deepseek", + "additional_details": { + "vals_model_id": "deepseek/deepseek-v4-pro", + "vals_provider": "DeepSeek" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 89.117, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.117", + "raw_stderr": "0.704", + "latency": "115.976", + "cost_per_test": "0", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.704, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 75.169, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "75.169", + "raw_stderr": "0.226", + "latency": "242.068", + "cost_per_test": "0", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.226, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 84.587, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.587", + "raw_stderr": "0.671", + "latency": "283.655", + "cost_per_test": "0", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.671, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.323, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.323", + "raw_stderr": "0.468", + "latency": "184.46", + "cost_per_test": "0", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 0.468, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 75.316, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "75.316", + "raw_stderr": "1.259", + "latency": "209.987", + "cost_per_test": "0", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.259, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:deepseek/deepseek-v4-pro:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.425, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.425", + "raw_stderr": "1.699", + "latency": "70.613", + "cost_per_test": "0", + "max_output_tokens": "256000", + "reasoning_effort": "max", + "provider": "DeepSeek" + }, + "uncertainty": { + "standard_error": { + "value": 1.699, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/devin/swe-1-6-fast/163bad1e-8da7-428d-83b4-1e5133a7c20c.json b/data/vals_ai/devin/swe-1-6-fast/163bad1e-8da7-428d-83b4-1e5133a7c20c.json new file mode 100644 index 0000000000000000000000000000000000000000..254678742e95d4d0bc4bc60e335f65d3aadb6233 --- /dev/null +++ b/data/vals_ai/devin/swe-1-6-fast/163bad1e-8da7-428d-83b4-1e5133a7c20c.json @@ -0,0 +1,89 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/devin_swe-1-6-fast/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "swe-1-6-fast", + "id": "devin/swe-1-6-fast", + "developer": "devin", + "additional_details": { + "vals_model_id": "devin/swe-1-6-fast", + "vals_provider": "Devin" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:devin/swe-1-6-fast:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 7.22, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "7.22", + "raw_stderr": "2.635", + "latency": "189.132", + "max_output_tokens": "0", + "provider": "Devin" + }, + "uncertainty": { + "standard_error": { + "value": 2.635, + "method": "vals_reported" + } + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-r1/23a7f703-69d6-42d0-932d-bdd25f63088b.json b/data/vals_ai/fireworks/deepseek-r1/23a7f703-69d6-42d0-932d-bdd25f63088b.json new file mode 100644 index 0000000000000000000000000000000000000000..bebd51d97c7ce31cb5cf20bb0962c09daf7ed6c3 --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-r1/23a7f703-69d6-42d0-932d-bdd25f63088b.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/fireworks_deepseek-r1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-r1", + "id": "fireworks/deepseek-r1", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-r1", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 60.752, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.752", + "raw_stderr": "1.396", + "latency": "152.958", + "cost_per_test": "0.012239", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.396, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 72.281, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "72.281", + "raw_stderr": "0.874", + "latency": "152.958", + "cost_per_test": "0.012239", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.874, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 83.81, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "83.81", + "raw_stderr": "1.053", + "latency": "152.958", + "cost_per_test": "0.012239", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.053, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-r1/28ac6116-508b-40ce-9350-d3e7ac60a14a.json b/data/vals_ai/fireworks/deepseek-r1/28ac6116-508b-40ce-9350-d3e7ac60a14a.json new file mode 100644 index 0000000000000000000000000000000000000000..3618d09350faef4ae4d403ea60934cbaed2a02fd --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-r1/28ac6116-508b-40ce-9350-d3e7ac60a14a.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/fireworks_deepseek-r1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-r1", + "id": "fireworks/deepseek-r1", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-r1", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.983, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.983", + "raw_stderr": "0.775", + "latency": "21.182", + "temperature": "1", + "max_output_tokens": "2048", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.775, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 2048, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 66.411, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "66.411", + "raw_stderr": "0.234", + "latency": "19.442", + "temperature": "1", + "max_output_tokens": "2048", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.234, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 2048, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.163, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.163", + "raw_stderr": "0.666", + "latency": "12.795", + "temperature": "1", + "max_output_tokens": "2048", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.666, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 2048, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 67.323, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "67.323", + "raw_stderr": "0.535", + "latency": "18.549", + "temperature": "1", + "max_output_tokens": "2048", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 2048, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 68.729, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "68.729", + "raw_stderr": "0.728", + "latency": "11.562", + "temperature": "1", + "max_output_tokens": "2048", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.728, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 2048, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 42.328, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "42.328", + "raw_stderr": "2.349", + "latency": "27.764", + "temperature": "1", + "max_output_tokens": "2048", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.349, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 2048, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-r1/2e2fafd0-42b7-4ec6-b60c-a84311fa6b14.json b/data/vals_ai/fireworks/deepseek-r1/2e2fafd0-42b7-4ec6-b60c-a84311fa6b14.json new file mode 100644 index 0000000000000000000000000000000000000000..ab3a6ca9035cd5c698dbd2bd4705d462b63b4422 --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-r1/2e2fafd0-42b7-4ec6-b60c-a84311fa6b14.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/fireworks_deepseek-r1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-r1", + "id": "fireworks/deepseek-r1", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-r1", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 59.557, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.557", + "raw_stderr": "1.675", + "latency": "20.406", + "cost_per_test": "0.00529", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.675, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 50.233, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "50.233", + "raw_stderr": "1.707", + "latency": "83.286", + "cost_per_test": "0.166784", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.707, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 54.118, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "54.118", + "raw_stderr": "0.979", + "latency": "46.774", + "cost_per_test": "0.086732", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.979, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 52.564, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.564", + "raw_stderr": "1.705", + "latency": "36.63", + "cost_per_test": "0.088122", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.705, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-r1/46ef97b1-8be6-42ef-9cfc-b917b38e27ad.json b/data/vals_ai/fireworks/deepseek-r1/46ef97b1-8be6-42ef-9cfc-b917b38e27ad.json new file mode 100644 index 0000000000000000000000000000000000000000..7c73e711a52b0920f531422b2167eacd64b27211 --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-r1/46ef97b1-8be6-42ef-9cfc-b917b38e27ad.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/fireworks_deepseek-r1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-r1", + "id": "fireworks/deepseek-r1", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-r1", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.795, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.795", + "raw_stderr": "1.634", + "latency": "21.636", + "cost_per_test": "0.010148", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.634, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.973, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.973", + "raw_stderr": "1.438", + "latency": "26.612", + "cost_per_test": "0.011762", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.438, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.428, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.428", + "raw_stderr": "1.38", + "latency": "38.388", + "cost_per_test": "0.017823", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.38, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.366, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.366", + "raw_stderr": "2.298", + "latency": "26.57", + "cost_per_test": "0.012518", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.298, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.678, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.678", + "raw_stderr": "1.474", + "latency": "21.108", + "cost_per_test": "0.009453", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.474, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.289, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.289", + "raw_stderr": "1.52", + "latency": "63.374", + "cost_per_test": "0.027306", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.52, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.139, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.139", + "raw_stderr": "1.586", + "latency": "19.731", + "cost_per_test": "0.008846", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.586, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.016, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.016", + "raw_stderr": "2.385", + "latency": "16.728", + "cost_per_test": "0.009046", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.385, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 68.756, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "68.756", + "raw_stderr": "1.476", + "latency": "30.476", + "cost_per_test": "0.013571", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.476, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.968, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.968", + "raw_stderr": "1.043", + "latency": "26.179", + "cost_per_test": "0.012333", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.043, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.385, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.385", + "raw_stderr": "1.494", + "latency": "18.175", + "cost_per_test": "0.008235", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.494, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.184, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.184", + "raw_stderr": "0.444", + "latency": "27.278", + "cost_per_test": "0.012339", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.444, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.553, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.553", + "raw_stderr": "2.072", + "latency": "17.808", + "cost_per_test": "0.008096", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.072, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.53, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.53", + "raw_stderr": "1.265", + "latency": "38.631", + "cost_per_test": "0.016036", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.265, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.707, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.707", + "raw_stderr": "1.603", + "latency": "16.482", + "cost_per_test": "0.007569", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.603, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-r1/62501156-0206-49e5-a22c-7f7480b50523.json b/data/vals_ai/fireworks/deepseek-r1/62501156-0206-49e5-a22c-7f7480b50523.json new file mode 100644 index 0000000000000000000000000000000000000000..931e0c540d009c7979524ae365bc8cdd014d53c7 --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-r1/62501156-0206-49e5-a22c-7f7480b50523.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/fireworks_deepseek-r1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-r1", + "id": "fireworks/deepseek-r1", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-r1", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.137, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.137", + "raw_stderr": "0.754", + "latency": "26.505", + "cost_per_test": "0.013517", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.754, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 36.286, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "36.286", + "raw_stderr": "2.57", + "latency": "150.553", + "cost_per_test": "0.067513", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.57, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 76.24, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "76.24", + "raw_stderr": "2.175", + "latency": "81.165", + "cost_per_test": "0.041335", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.175, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:fireworks/deepseek-r1:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 70.221, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "70.221", + "raw_stderr": "1.15", + "latency": "86.074", + "cost_per_test": "0.040788", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.15, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-v3-0324/1affa777-0730-48c3-b497-4f9c530fafa1.json b/data/vals_ai/fireworks/deepseek-v3-0324/1affa777-0730-48c3-b497-4f9c530fafa1.json new file mode 100644 index 0000000000000000000000000000000000000000..44f04342116673f93e46fc84fbe0132476c1277e --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-v3-0324/1affa777-0730-48c3-b497-4f9c530fafa1.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/fireworks_deepseek-v3-0324/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v3-0324", + "id": "fireworks/deepseek-v3-0324", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-v3-0324", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 58.953, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.953", + "raw_stderr": "1.407", + "latency": "34.009", + "cost_per_test": "0.000652", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.407, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 71.096, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.096", + "raw_stderr": "0.883", + "latency": "34.009", + "cost_per_test": "0.000652", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.883, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 83.238, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "83.238", + "raw_stderr": "1.068", + "latency": "34.009", + "cost_per_test": "0.000652", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.068, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-v3-0324/41789e45-1578-4736-9bbe-e310ed8029b3.json b/data/vals_ai/fireworks/deepseek-v3-0324/41789e45-1578-4736-9bbe-e310ed8029b3.json new file mode 100644 index 0000000000000000000000000000000000000000..fef81ee1d2bc51162e542f1baae74dccd7bbb8af --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-v3-0324/41789e45-1578-4736-9bbe-e310ed8029b3.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/fireworks_deepseek-v3-0324/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v3-0324", + "id": "fireworks/deepseek-v3-0324", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-v3-0324", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.67, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.67", + "raw_stderr": "0.927", + "latency": "10.726", + "temperature": "1", + "max_output_tokens": "131072", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.927, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.276, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.276", + "raw_stderr": "0.201", + "latency": "6.96", + "temperature": "1", + "max_output_tokens": "131072", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.201, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 86.204, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.204", + "raw_stderr": "0.609", + "latency": "7.862", + "temperature": "1", + "max_output_tokens": "131072", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.609, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.727, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.727", + "raw_stderr": "0.418", + "latency": "7.712", + "temperature": "1", + "max_output_tokens": "131072", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.418, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.163, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.163", + "raw_stderr": "0.942", + "latency": "6.982", + "temperature": "1", + "max_output_tokens": "131072", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.942, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 59.321, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "59.321", + "raw_stderr": "1.487", + "latency": "6.029", + "temperature": "1", + "max_output_tokens": "131072", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.487, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-v3-0324/8d9435ab-5448-45c4-aeed-86823d665b5c.json b/data/vals_ai/fireworks/deepseek-v3-0324/8d9435ab-5448-45c4-aeed-86823d665b5c.json new file mode 100644 index 0000000000000000000000000000000000000000..e4ae8524db1326d62c0193f54f1601f0c4c46795 --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-v3-0324/8d9435ab-5448-45c4-aeed-86823d665b5c.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/fireworks_deepseek-v3-0324/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v3-0324", + "id": "fireworks/deepseek-v3-0324", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-v3-0324", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.373, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.373", + "raw_stderr": "1.67", + "latency": "12.461", + "cost_per_test": "0.001418", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.67, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 50.932, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "50.932", + "raw_stderr": "1.707", + "latency": "66.39", + "cost_per_test": "0.098456", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.707, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 54.74, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "54.74", + "raw_stderr": "0.978", + "latency": "36.732", + "cost_per_test": "0.052049", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.978, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 52.914, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.914", + "raw_stderr": "1.704", + "latency": "31.344", + "cost_per_test": "0.056273", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.704, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-v3-0324/9acc914b-84ab-46b7-9c14-7077e3f14250.json b/data/vals_ai/fireworks/deepseek-v3-0324/9acc914b-84ab-46b7-9c14-7077e3f14250.json new file mode 100644 index 0000000000000000000000000000000000000000..5d8cf11d2456297c686129d050f93aff4ab08e92 --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-v3-0324/9acc914b-84ab-46b7-9c14-7077e3f14250.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/fireworks_deepseek-v3-0324/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v3-0324", + "id": "fireworks/deepseek-v3-0324", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-v3-0324", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 1.667, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "1.667", + "raw_stderr": "1.521", + "latency": "144.646", + "cost_per_test": "0.106655", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.521, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 1.667, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "1.667", + "raw_stderr": "0.962", + "latency": "198.312", + "cost_per_test": "1.106313", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.962, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 1.667, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "1.667", + "raw_stderr": "0.9", + "latency": "171.479", + "cost_per_test": "0.606484", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.9, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-v3-0324/ad53c5c6-f80c-41c6-bdad-4f49c271250c.json b/data/vals_ai/fireworks/deepseek-v3-0324/ad53c5c6-f80c-41c6-bdad-4f49c271250c.json new file mode 100644 index 0000000000000000000000000000000000000000..f2877a6bbfefc8c64b01918122bf6807d6f62a8b --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-v3-0324/ad53c5c6-f80c-41c6-bdad-4f49c271250c.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/fireworks_deepseek-v3-0324/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v3-0324", + "id": "fireworks/deepseek-v3-0324", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-v3-0324", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 96.584, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "96.584", + "raw_stderr": "1.012", + "latency": "13.189", + "cost_per_test": "0.000798", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.012, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 28.571, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "28.571", + "raw_stderr": "2.415", + "latency": "46.773", + "cost_per_test": "0.002913", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.415, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 71.279, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "71.279", + "raw_stderr": "2.312", + "latency": "20.827", + "cost_per_test": "0.001617", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.312, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 65.478, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "65.478", + "raw_stderr": "1.164", + "latency": "26.93", + "cost_per_test": "0.001776", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.164, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-v3-0324/bad126e2-47bb-4eb2-b9bb-88b76181e00a.json b/data/vals_ai/fireworks/deepseek-v3-0324/bad126e2-47bb-4eb2-b9bb-88b76181e00a.json new file mode 100644 index 0000000000000000000000000000000000000000..4ac723613be6d81263fe2ee4eb9aadb987091ff1 --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-v3-0324/bad126e2-47bb-4eb2-b9bb-88b76181e00a.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/fireworks_deepseek-v3-0324/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v3-0324", + "id": "fireworks/deepseek-v3-0324", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-v3-0324", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 58.586, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "58.586", + "raw_stderr": "3.501", + "latency": "19.603", + "cost_per_test": "0.002135", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.501, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 64.646, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "64.646", + "raw_stderr": "3.419", + "latency": "27.347", + "cost_per_test": "0.000969", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.419, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 61.616, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "61.616", + "raw_stderr": "2.447", + "latency": "23.475", + "cost_per_test": "0.001552", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.447, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-v3-0324/f28287c8-8cf0-4c78-809b-46f677ebab05.json b/data/vals_ai/fireworks/deepseek-v3-0324/f28287c8-8cf0-4c78-809b-46f677ebab05.json new file mode 100644 index 0000000000000000000000000000000000000000..9b059262995c1120a7cb3795a9512e864f8189d8 --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-v3-0324/f28287c8-8cf0-4c78-809b-46f677ebab05.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/fireworks_deepseek-v3-0324/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v3-0324", + "id": "fireworks/deepseek-v3-0324", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-v3-0324", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.679, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.679", + "raw_stderr": "1.143", + "latency": "23.473", + "cost_per_test": "0.001441", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.143, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.044, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.044", + "raw_stderr": "1.283", + "latency": "18.53", + "cost_per_test": "0.001292", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.283, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.951, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.951", + "raw_stderr": "1.14", + "latency": "47.977", + "cost_per_test": "0.001779", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.14, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.195, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.195", + "raw_stderr": "1.948", + "latency": "29.083", + "cost_per_test": "0.001713", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.948, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.479, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.479", + "raw_stderr": "1.246", + "latency": "12.961", + "cost_per_test": "0.001274", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.246, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 72.446, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.446", + "raw_stderr": "1.465", + "latency": "67.828", + "cost_per_test": "0.001946", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.465, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.65, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.65", + "raw_stderr": "1.479", + "latency": "22.284", + "cost_per_test": "0.001227", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.479, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 70.079, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.079", + "raw_stderr": "2.346", + "latency": "7.615", + "cost_per_test": "0.002016", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.346, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 56.403, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "56.403", + "raw_stderr": "1.494", + "latency": "29.997", + "cost_per_test": "0.001441", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.494, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.044, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.044", + "raw_stderr": "0.8", + "latency": "23.543", + "cost_per_test": "0.001944", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.8, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.247, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.247", + "raw_stderr": "1.357", + "latency": "9.885", + "cost_per_test": "0.000955", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.357, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.474, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.474", + "raw_stderr": "0.396", + "latency": "25.051", + "cost_per_test": "0.00147", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.396, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.752, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.752", + "raw_stderr": "1.919", + "latency": "24.442", + "cost_per_test": "0.000964", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.919, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.22, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.22", + "raw_stderr": "1.016", + "latency": "22.957", + "cost_per_test": "0.001517", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.016, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:fireworks/deepseek-v3-0324:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.454, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.454", + "raw_stderr": "1.376", + "latency": "10.137", + "cost_per_test": "0.001074", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.376, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-v3/14e70a23-00ab-4c40-b153-7401f8904ec1.json b/data/vals_ai/fireworks/deepseek-v3/14e70a23-00ab-4c40-b153-7401f8904ec1.json new file mode 100644 index 0000000000000000000000000000000000000000..9388b4d812edf24b1ba17d1275ca41df34420758 --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-v3/14e70a23-00ab-4c40-b153-7401f8904ec1.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/fireworks_deepseek-v3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v3", + "id": "fireworks/deepseek-v3", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-v3", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:fireworks/deepseek-v3:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 57.073, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.073", + "raw_stderr": "1.415", + "latency": "27.696", + "cost_per_test": "0.00056", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.415, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:fireworks/deepseek-v3:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 67.907, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.907", + "raw_stderr": "0.918", + "latency": "27.696", + "cost_per_test": "0.00056", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.918, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:fireworks/deepseek-v3:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 78.741, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "78.741", + "raw_stderr": "1.17", + "latency": "27.696", + "cost_per_test": "0.00056", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.17, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-v3/62ea6b12-4a55-40b1-8472-1ce2ba3eebd0.json b/data/vals_ai/fireworks/deepseek-v3/62ea6b12-4a55-40b1-8472-1ce2ba3eebd0.json new file mode 100644 index 0000000000000000000000000000000000000000..d322636cd9c52067eea7d67f38380b16748aaac4 --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-v3/62ea6b12-4a55-40b1-8472-1ce2ba3eebd0.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/fireworks_deepseek-v3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v3", + "id": "fireworks/deepseek-v3", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-v3", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:fireworks/deepseek-v3:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 58.741, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.741", + "raw_stderr": "1.681", + "latency": "6.717", + "cost_per_test": "0.001386", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.681, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:fireworks/deepseek-v3:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 47.669, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "47.669", + "raw_stderr": "1.705", + "latency": "50.565", + "cost_per_test": "0.098432", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.705, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:fireworks/deepseek-v3:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 52.486, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.486", + "raw_stderr": "0.98", + "latency": "28.322", + "cost_per_test": "0.052026", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.98, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:fireworks/deepseek-v3:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 51.049, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "51.049", + "raw_stderr": "1.707", + "latency": "27.684", + "cost_per_test": "0.05626", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.707, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-v3/9c04895b-9836-4de7-be85-7a7551df085c.json b/data/vals_ai/fireworks/deepseek-v3/9c04895b-9836-4de7-be85-7a7551df085c.json new file mode 100644 index 0000000000000000000000000000000000000000..3ae12bf22c72996306b43c1c8853d9b25fb18e38 --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-v3/9c04895b-9836-4de7-be85-7a7551df085c.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/fireworks_deepseek-v3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v3", + "id": "fireworks/deepseek-v3", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-v3", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:fireworks/deepseek-v3:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 84.255, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.255", + "raw_stderr": "0.864", + "latency": "5.812", + "temperature": "1", + "max_output_tokens": "131072", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.864, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:fireworks/deepseek-v3:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.976, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.976", + "raw_stderr": "0.206", + "latency": "4.133", + "temperature": "1", + "max_output_tokens": "131072", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.206, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:fireworks/deepseek-v3:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 85.619, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.619", + "raw_stderr": "0.608", + "latency": "2.697", + "temperature": "1", + "max_output_tokens": "131072", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.608, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:fireworks/deepseek-v3:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.762, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.762", + "raw_stderr": "0.429", + "latency": "4.395", + "temperature": "1", + "max_output_tokens": "131072", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.429, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:fireworks/deepseek-v3:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.062, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.062", + "raw_stderr": "0.853", + "latency": "4.108", + "temperature": "1", + "max_output_tokens": "131072", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.853, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:fireworks/deepseek-v3:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 73.899, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "73.899", + "raw_stderr": "1.648", + "latency": "5.227", + "temperature": "1", + "max_output_tokens": "131072", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.648, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-v3/a41a521a-9429-4170-a5dc-3e4c749ab15c.json b/data/vals_ai/fireworks/deepseek-v3/a41a521a-9429-4170-a5dc-3e4c749ab15c.json new file mode 100644 index 0000000000000000000000000000000000000000..d0314bff3d24d512930a35e02805dd7319e6bb81 --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-v3/a41a521a-9429-4170-a5dc-3e4c749ab15c.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/fireworks_deepseek-v3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v3", + "id": "fireworks/deepseek-v3", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-v3", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:fireworks/deepseek-v3:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.495, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.495", + "raw_stderr": "1.315", + "latency": "16.683", + "cost_per_test": "0.001331", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.315, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:fireworks/deepseek-v3:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.271, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.271", + "raw_stderr": "1.556", + "latency": "4.416", + "cost_per_test": "0.001053", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.556, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:fireworks/deepseek-v3:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.972, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.972", + "raw_stderr": "1.27", + "latency": "17.212", + "cost_per_test": "0.001588", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.27, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:fireworks/deepseek-v3:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.317, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.317", + "raw_stderr": "2.068", + "latency": "10.47", + "cost_per_test": "0.001581", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.068, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:fireworks/deepseek-v3:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.806, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.806", + "raw_stderr": "1.356", + "latency": "4.392", + "cost_per_test": "0.001227", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.356, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:fireworks/deepseek-v3:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 60.681, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "60.681", + "raw_stderr": "1.571", + "latency": "28.367", + "cost_per_test": "0.001626", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.571, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:fireworks/deepseek-v3:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.572, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.572", + "raw_stderr": "1.523", + "latency": "18.211", + "cost_per_test": "0.001068", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.523, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:fireworks/deepseek-v3:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 67.979, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "67.979", + "raw_stderr": "2.39", + "latency": "2.571", + "cost_per_test": "0.001981", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.39, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:fireworks/deepseek-v3:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 51.045, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "51.045", + "raw_stderr": "1.507", + "latency": "14.726", + "cost_per_test": "0.001382", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.507, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:fireworks/deepseek-v3:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.826, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.826", + "raw_stderr": "0.976", + "latency": "16.389", + "cost_per_test": "0.001766", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.976, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:fireworks/deepseek-v3:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.134, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.134", + "raw_stderr": "1.441", + "latency": "4.106", + "cost_per_test": "0.000891", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.441, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:fireworks/deepseek-v3:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.82, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.82", + "raw_stderr": "0.424", + "latency": "11.36", + "cost_per_test": "0.001338", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.424, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:fireworks/deepseek-v3:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 70.341, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.341", + "raw_stderr": "2.045", + "latency": "11.205", + "cost_per_test": "0.000936", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.045, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:fireworks/deepseek-v3:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.213, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.213", + "raw_stderr": "1.164", + "latency": "7.147", + "cost_per_test": "0.001282", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.164, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:fireworks/deepseek-v3:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.822, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.822", + "raw_stderr": "1.446", + "latency": "3.148", + "cost_per_test": "0.001019", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.446, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-v3/b28a08a0-f76d-4099-8c80-983df70a397c.json b/data/vals_ai/fireworks/deepseek-v3/b28a08a0-f76d-4099-8c80-983df70a397c.json new file mode 100644 index 0000000000000000000000000000000000000000..c0bbac84e12da8246984fb0403c32e6fba85adaf --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-v3/b28a08a0-f76d-4099-8c80-983df70a397c.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/fireworks_deepseek-v3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v3", + "id": "fireworks/deepseek-v3", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-v3", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:fireworks/deepseek-v3:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 53.535, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "53.535", + "raw_stderr": "3.544", + "latency": "19.423", + "cost_per_test": "0.002142", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.544, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:fireworks/deepseek-v3:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 55.556, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "55.556", + "raw_stderr": "3.539", + "latency": "23.141", + "cost_per_test": "0.000795", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.539, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:fireworks/deepseek-v3:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 54.546, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "54.546", + "raw_stderr": "2.504", + "latency": "21.282", + "cost_per_test": "0.001468", + "temperature": "1", + "max_output_tokens": "8192", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.504, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-v3p1/0ffb036b-bad4-4472-8fcc-a959715a4be5.json b/data/vals_ai/fireworks/deepseek-v3p1/0ffb036b-bad4-4472-8fcc-a959715a4be5.json new file mode 100644 index 0000000000000000000000000000000000000000..b91804a7243add309dc3621f6959f67ae9b65080 --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-v3p1/0ffb036b-bad4-4472-8fcc-a959715a4be5.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/fireworks_deepseek-v3p1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v3p1", + "id": "fireworks/deepseek-v3p1", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-v3p1", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:fireworks/deepseek-v3p1:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 56.876, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.876", + "raw_stderr": "1.691", + "latency": "31.727", + "cost_per_test": "0.000982", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.691, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:fireworks/deepseek-v3p1:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 48.252, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "48.252", + "raw_stderr": "1.706", + "latency": "96.378", + "cost_per_test": "0.075218", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.706, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:fireworks/deepseek-v3p1:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 51.476, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "51.476", + "raw_stderr": "0.982", + "latency": "62.999", + "cost_per_test": "0.037369", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.982, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:fireworks/deepseek-v3p1:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 49.301, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "49.301", + "raw_stderr": "1.707", + "latency": "60.891", + "cost_per_test": "0.035907", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.707, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-v3p2-thinking/042b909d-b938-48a9-9950-3526600a5b4b.json b/data/vals_ai/fireworks/deepseek-v3p2-thinking/042b909d-b938-48a9-9950-3526600a5b4b.json new file mode 100644 index 0000000000000000000000000000000000000000..6991b77186f5d319f5f70f4684b1b7866d0a6dc6 --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-v3p2-thinking/042b909d-b938-48a9-9950-3526600a5b4b.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/fireworks_deepseek-v3p2-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v3p2-thinking", + "id": "fireworks/deepseek-v3p2-thinking", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-v3p2-thinking", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 5.108, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "5.108", + "raw_stderr": "2.127", + "latency": "3365.726", + "cost_per_test": "2.469325", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.127, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-v3p2-thinking/1ac7ee0b-0e9c-45a1-9416-38663fbf2cf9.json b/data/vals_ai/fireworks/deepseek-v3p2-thinking/1ac7ee0b-0e9c-45a1-9416-38663fbf2cf9.json new file mode 100644 index 0000000000000000000000000000000000000000..557c037c958ccc31d9b29f948823c592a453f811 --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-v3p2-thinking/1ac7ee0b-0e9c-45a1-9416-38663fbf2cf9.json @@ -0,0 +1,450 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/fireworks_deepseek-v3p2-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v3p2-thinking", + "id": "fireworks/deepseek-v3p2-thinking", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-v3p2-thinking", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.709, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.709", + "raw_stderr": "0.815", + "latency": "29.776", + "cost_per_test": "0.000425", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.815, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 75.892, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "75.892", + "raw_stderr": "0.231", + "latency": "32.315", + "cost_per_test": "0.000468", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.231, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 76.035, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "76.035", + "raw_stderr": "0.808", + "latency": "33.565", + "cost_per_test": "0.000512", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.808, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 76.076, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "76.076", + "raw_stderr": "0.493", + "latency": "31.468", + "cost_per_test": "0.000454", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.493, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 72.096, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "72.096", + "raw_stderr": "1.389", + "latency": "29.918", + "cost_per_test": "0.000433", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.389, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 68.646, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "68.646", + "raw_stderr": "1.667", + "latency": "31.764", + "cost_per_test": "0.000434", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.667, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-v3p2-thinking/3978e2cb-c9cb-432b-8d67-8a9ae3387f37.json b/data/vals_ai/fireworks/deepseek-v3p2-thinking/3978e2cb-c9cb-432b-8d67-8a9ae3387f37.json new file mode 100644 index 0000000000000000000000000000000000000000..70480973c909f534335cc5bcec2d45fc631a1a47 --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-v3p2-thinking/3978e2cb-c9cb-432b-8d67-8a9ae3387f37.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/fireworks_deepseek-v3p2-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v3p2-thinking", + "id": "fireworks/deepseek-v3p2-thinking", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-v3p2-thinking", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 78.788, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.788", + "raw_stderr": "2.905", + "latency": "246.265", + "cost_per_test": "0.003266", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.905, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 81.818, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.818", + "raw_stderr": "2.741", + "latency": "274.231", + "cost_per_test": "0.003664", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.741, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 80.303, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.303", + "raw_stderr": "1.997", + "latency": "260.248", + "cost_per_test": "0.003465", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.997, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-v3p2-thinking/5c846690-2781-40b6-b49e-d37a7d85697d.json b/data/vals_ai/fireworks/deepseek-v3p2-thinking/5c846690-2781-40b6-b49e-d37a7d85697d.json new file mode 100644 index 0000000000000000000000000000000000000000..7db81d997767b1f03ea523fb8081a4289d06af11 --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-v3p2-thinking/5c846690-2781-40b6-b49e-d37a7d85697d.json @@ -0,0 +1,381 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/fireworks_deepseek-v3p2-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v3p2-thinking", + "id": "fireworks/deepseek-v3p2-thinking", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-v3p2-thinking", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 28.571, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "28.571", + "raw_stderr": "7.055", + "latency": "630.532", + "cost_per_test": "0.908617", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 7.055, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 62.069, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "62.069", + "raw_stderr": "3.009", + "latency": "566.085", + "cost_per_test": "0.735042", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.009, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 84.021, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "84.021", + "raw_stderr": "2.638", + "latency": "465.962", + "cost_per_test": "0.533824", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.638, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "630.294", + "cost_per_test": "0.949807", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:overall:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 67.6, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "67.6", + "raw_stderr": "2.095", + "latency": "533.036", + "cost_per_test": "0.672838", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.095, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-v3p2-thinking/615a449f-9905-4627-8c7c-b7782f2e146a.json b/data/vals_ai/fireworks/deepseek-v3p2-thinking/615a449f-9905-4627-8c7c-b7782f2e146a.json new file mode 100644 index 0000000000000000000000000000000000000000..eecb67e4143d5837cf24d979a49525c2c8480f2f --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-v3p2-thinking/615a449f-9905-4627-8c7c-b7782f2e146a.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/fireworks_deepseek-v3p2-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v3p2-thinking", + "id": "fireworks/deepseek-v3p2-thinking", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-v3p2-thinking", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 11.167, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "11.167", + "raw_stderr": "5.32", + "latency": "1374.876", + "cost_per_test": "0.050024", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 5.32, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 10.167, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "10.167", + "raw_stderr": "3.14", + "latency": "1742.362", + "cost_per_test": "0.067402", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.14, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "ioi:overall:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 10.667, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "10.667", + "raw_stderr": "3.089", + "latency": "1558.619", + "cost_per_test": "0.058713", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.089, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-v3p2-thinking/641e649d-a5ce-4aa7-8cb2-a99025fd823f.json b/data/vals_ai/fireworks/deepseek-v3p2-thinking/641e649d-a5ce-4aa7-8cb2-a99025fd823f.json new file mode 100644 index 0000000000000000000000000000000000000000..28ea24f965b9b1034c825286daf47c90c4dad306 --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-v3p2-thinking/641e649d-a5ce-4aa7-8cb2-a99025fd823f.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/fireworks_deepseek-v3p2-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v3p2-thinking", + "id": "fireworks/deepseek-v3p2-thinking", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-v3p2-thinking", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 4.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "4", + "raw_stderr": "1.96", + "latency": "3441.448", + "cost_per_test": "0.128469", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.96, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-v3p2-thinking/66f124f8-2bcf-4d2b-b591-8d69b6d210c8.json b/data/vals_ai/fireworks/deepseek-v3p2-thinking/66f124f8-2bcf-4d2b-b591-8d69b6d210c8.json new file mode 100644 index 0000000000000000000000000000000000000000..4d57e588dbd44fd41daadf1335b24863ef8d9501 --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-v3p2-thinking/66f124f8-2bcf-4d2b-b591-8d69b6d210c8.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/fireworks_deepseek-v3p2-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v3p2-thinking", + "id": "fireworks/deepseek-v3p2-thinking", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-v3p2-thinking", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 44.755, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "44.755", + "raw_stderr": "1.698", + "latency": "173.546", + "cost_per_test": "0.075006", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.698, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 58.858, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.858", + "raw_stderr": "1.68", + "latency": "15.178", + "cost_per_test": "0.0018", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.68, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 50.971, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "50.971", + "raw_stderr": "0.979", + "latency": "99.485", + "cost_per_test": "0.038215", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.979, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 49.301, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "49.301", + "raw_stderr": "1.707", + "latency": "109.73", + "cost_per_test": "0.037839", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.707, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-v3p2-thinking/8d91b0f7-0b96-4273-8f98-7178864309e8.json b/data/vals_ai/fireworks/deepseek-v3p2-thinking/8d91b0f7-0b96-4273-8f98-7178864309e8.json new file mode 100644 index 0000000000000000000000000000000000000000..b244421f74c0ee016868dd9c4dbc6584ef670567 --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-v3p2-thinking/8d91b0f7-0b96-4273-8f98-7178864309e8.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/fireworks_deepseek-v3p2-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v3p2-thinking", + "id": "fireworks/deepseek-v3p2-thinking", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-v3p2-thinking", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 97.826, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "97.826", + "raw_stderr": "0.813", + "latency": "47.866", + "cost_per_test": "0.000681", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.813, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:hard:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 59.143, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "59.143", + "raw_stderr": "2.628", + "latency": "474.722", + "cost_per_test": "0.006164", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.628, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:medium:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 85.117, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "85.117", + "raw_stderr": "1.819", + "latency": "240.241", + "cost_per_test": "0.002954", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.819, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:overall:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 80.695, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "80.695", + "raw_stderr": "1.099", + "latency": "254.276", + "cost_per_test": "0.003266", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.099, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-v3p2-thinking/e0c0b4ae-4317-48e6-9d2f-cd018aacef32.json b/data/vals_ai/fireworks/deepseek-v3p2-thinking/e0c0b4ae-4317-48e6-9d2f-cd018aacef32.json new file mode 100644 index 0000000000000000000000000000000000000000..d9abde0b99502fd769b7a063d9c020014b0eadec --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-v3p2-thinking/e0c0b4ae-4317-48e6-9d2f-cd018aacef32.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/fireworks_deepseek-v3p2-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v3p2-thinking", + "id": "fireworks/deepseek-v3p2-thinking", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-v3p2-thinking", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 57.89, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.89", + "raw_stderr": "1.412", + "latency": "98.919", + "cost_per_test": "0.002376", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.412, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 68.152, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.152", + "raw_stderr": "0.919", + "latency": "98.919", + "cost_per_test": "0.002376", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.919, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 78.414, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "78.414", + "raw_stderr": "1.176", + "latency": "98.919", + "cost_per_test": "0.002376", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.176, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-v3p2-thinking/e914be53-9283-4bc7-9061-5241f68d4dd9.json b/data/vals_ai/fireworks/deepseek-v3p2-thinking/e914be53-9283-4bc7-9061-5241f68d4dd9.json new file mode 100644 index 0000000000000000000000000000000000000000..256381a8cfb26a3b1936cb5e5754a92bdae8e413 --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-v3p2-thinking/e914be53-9283-4bc7-9061-5241f68d4dd9.json @@ -0,0 +1,1071 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/fireworks_deepseek-v3p2-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v3p2-thinking", + "id": "fireworks/deepseek-v3p2-thinking", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-v3p2-thinking", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.213, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.213", + "raw_stderr": "1.057", + "latency": "65.509", + "cost_per_test": "0.000683", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.057, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.114, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.114", + "raw_stderr": "1.063", + "latency": "97.784", + "cost_per_test": "0.001082", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.063, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.841, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.841", + "raw_stderr": "0.898", + "latency": "127.304", + "cost_per_test": "0.001651", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.898, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.317, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.317", + "raw_stderr": "1.643", + "latency": "87.233", + "cost_per_test": "0.001014", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.643, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.507, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.507", + "raw_stderr": "1.098", + "latency": "62.365", + "cost_per_test": "0.000679", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.098, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.282, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.282", + "raw_stderr": "1.199", + "latency": "222.254", + "cost_per_test": "0.003153", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.199, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.362, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.362", + "raw_stderr": "1.44", + "latency": "67.657", + "cost_per_test": "0.000779", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.44, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.115, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.115", + "raw_stderr": "2.184", + "latency": "61.056", + "cost_per_test": "0.000701", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.184, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 72.025, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.025", + "raw_stderr": "1.353", + "latency": "105.186", + "cost_per_test": "0.001281", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.353, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 95.041, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "95.041", + "raw_stderr": "0.591", + "latency": "73.223", + "cost_per_test": "0.000995", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.591, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.468, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.468", + "raw_stderr": "1.251", + "latency": "59.92", + "cost_per_test": "0.000662", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.251, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.917, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.917", + "raw_stderr": "0.353", + "latency": "90.148", + "cost_per_test": "0.001102", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.353, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.561, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.561", + "raw_stderr": "1.772", + "latency": "61.44", + "cost_per_test": "0.000613", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.772, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.53, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.53", + "raw_stderr": "0.849", + "latency": "112.283", + "cost_per_test": "0.001525", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.849, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:fireworks/deepseek-v3p2-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.461, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.461", + "raw_stderr": "1.282", + "latency": "58.865", + "cost_per_test": "0.000616", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "high", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.282, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-v3p2/0d3e0dce-a348-474c-b5ae-5cee6dc9396e.json b/data/vals_ai/fireworks/deepseek-v3p2/0d3e0dce-a348-474c-b5ae-5cee6dc9396e.json new file mode 100644 index 0000000000000000000000000000000000000000..e0f000d44556fab594a690d28e0fec69b3ebf0cc --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-v3p2/0d3e0dce-a348-474c-b5ae-5cee6dc9396e.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/fireworks_deepseek-v3p2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v3p2", + "id": "fireworks/deepseek-v3p2", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-v3p2", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 43.473, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "43.473", + "raw_stderr": "1.692", + "latency": "355.63", + "cost_per_test": "0.073712", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.692, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 54.895, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "54.895", + "raw_stderr": "1.699", + "latency": "5.336", + "cost_per_test": "0.000945", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.699, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 47.941, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "47.941", + "raw_stderr": "0.98", + "latency": "133.787", + "cost_per_test": "0.036826", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.98, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 45.455, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "45.455", + "raw_stderr": "1.7", + "latency": "40.394", + "cost_per_test": "0.035821", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.7, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-v3p2/31e66598-526a-46d5-8abf-e7eae84bd186.json b/data/vals_ai/fireworks/deepseek-v3p2/31e66598-526a-46d5-8abf-e7eae84bd186.json new file mode 100644 index 0000000000000000000000000000000000000000..8d0af440021f64b928a67dc5fbad0b62ca4768f9 --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-v3p2/31e66598-526a-46d5-8abf-e7eae84bd186.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/fireworks_deepseek-v3p2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v3p2", + "id": "fireworks/deepseek-v3p2", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-v3p2", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 96.273, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "96.273", + "raw_stderr": "1.056", + "latency": "7.002", + "cost_per_test": "0.0002", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.056, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + }, + { + "evaluation_result_id": "lcb:hard:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 39.143, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "39.143", + "raw_stderr": "2.609", + "latency": "76.711", + "cost_per_test": "0.001234", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.609, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + }, + { + "evaluation_result_id": "lcb:medium:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 74.151, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "74.151", + "raw_stderr": "2.237", + "latency": "34.315", + "cost_per_test": "0.000597", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.237, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + }, + { + "evaluation_result_id": "lcb:overall:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 69.856, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "69.856", + "raw_stderr": "1.198", + "latency": "39.343", + "cost_per_test": "0.000677", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.198, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-v3p2/7055cd19-c909-4d51-b7db-081b03dfcef6.json b/data/vals_ai/fireworks/deepseek-v3p2/7055cd19-c909-4d51-b7db-081b03dfcef6.json new file mode 100644 index 0000000000000000000000000000000000000000..134454f495efe86ab4433de22273eda5c2e1e083 --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-v3p2/7055cd19-c909-4d51-b7db-081b03dfcef6.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/fireworks_deepseek-v3p2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v3p2", + "id": "fireworks/deepseek-v3p2", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-v3p2", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 14.0, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "14", + "raw_stderr": "6.364", + "latency": "1615.658", + "cost_per_test": "0.052688", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 6.364, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 14.833, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "14.833", + "raw_stderr": "6.353", + "latency": "3153.373", + "cost_per_test": "0.051233", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 6.353, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + }, + { + "evaluation_result_id": "ioi:overall:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 14.416, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "14.416", + "raw_stderr": "4.496", + "latency": "2384.516", + "cost_per_test": "0.05196", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 4.496, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-v3p2/88aa54c9-6c61-4804-bf2b-3107583dfb76.json b/data/vals_ai/fireworks/deepseek-v3p2/88aa54c9-6c61-4804-bf2b-3107583dfb76.json new file mode 100644 index 0000000000000000000000000000000000000000..b90155b1e55fe4cafeaa3720e5b3d5b527e2b675 --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-v3p2/88aa54c9-6c61-4804-bf2b-3107583dfb76.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/fireworks_deepseek-v3p2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v3p2", + "id": "fireworks/deepseek-v3p2", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-v3p2", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 8.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "8", + "raw_stderr": "2.713", + "latency": "3981.084", + "cost_per_test": "0.11773", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.713, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-v3p2/9441ee75-6fef-410c-b715-62a8a81733ec.json b/data/vals_ai/fireworks/deepseek-v3p2/9441ee75-6fef-410c-b715-62a8a81733ec.json new file mode 100644 index 0000000000000000000000000000000000000000..5f41abde6a0d7f4ce64d8961d7b5efb990125228 --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-v3p2/9441ee75-6fef-410c-b715-62a8a81733ec.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/fireworks_deepseek-v3p2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v3p2", + "id": "fireworks/deepseek-v3p2", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-v3p2", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 76.263, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.263", + "raw_stderr": "3.088", + "latency": "41.601", + "cost_per_test": "0.000597", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.088, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 76.263, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.263", + "raw_stderr": "3.024", + "latency": "52.056", + "cost_per_test": "0.000703", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.024, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 76.263, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.263", + "raw_stderr": "2.161", + "latency": "46.828", + "cost_per_test": "0.00065", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.161, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-v3p2/d4b7bd5d-db01-4570-b068-3bc3eb7945cf.json b/data/vals_ai/fireworks/deepseek-v3p2/d4b7bd5d-db01-4570-b068-3bc3eb7945cf.json new file mode 100644 index 0000000000000000000000000000000000000000..36f0ffeb0dc21d87976f4e01a2d4e09b3810a732 --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-v3p2/d4b7bd5d-db01-4570-b068-3bc3eb7945cf.json @@ -0,0 +1,450 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/fireworks_deepseek-v3p2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v3p2", + "id": "fireworks/deepseek-v3p2", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-v3p2", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.856, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.856", + "raw_stderr": "0.883", + "latency": "1.492", + "cost_per_test": "3.8e-05", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.883, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 62.948, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "62.948", + "raw_stderr": "0.215", + "latency": "1.526", + "cost_per_test": "8.7e-05", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.215, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 54.728, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "54.728", + "raw_stderr": "0.396", + "latency": "1.507", + "cost_per_test": "0.000151", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.396, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 65.962, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "65.962", + "raw_stderr": "0.419", + "latency": "2.185", + "cost_per_test": "8.6e-05", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.419, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 61.532, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "61.532", + "raw_stderr": "1.26", + "latency": "1.51", + "cost_per_test": "8.2e-05", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.26, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 68.748, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "68.748", + "raw_stderr": "1.346", + "latency": "4.89", + "cost_per_test": "7e-05", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.346, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/deepseek-v3p2/e3566ac0-5fbb-4ad7-8efb-c4d62ef2cbc2.json b/data/vals_ai/fireworks/deepseek-v3p2/e3566ac0-5fbb-4ad7-8efb-c4d62ef2cbc2.json new file mode 100644 index 0000000000000000000000000000000000000000..63b589e71c210ec22385cc69f4fcb89629edbeed --- /dev/null +++ b/data/vals_ai/fireworks/deepseek-v3p2/e3566ac0-5fbb-4ad7-8efb-c4d62ef2cbc2.json @@ -0,0 +1,1071 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/fireworks_deepseek-v3p2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "deepseek-v3p2", + "id": "fireworks/deepseek-v3p2", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/deepseek-v3p2", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.213, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.213", + "raw_stderr": "1.087", + "latency": "19.793", + "cost_per_test": "0.00025", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.087, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.692, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.692", + "raw_stderr": "1.256", + "latency": "43.276", + "cost_per_test": "0.000278", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.256, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.191, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.191", + "raw_stderr": "1.005", + "latency": "42.326", + "cost_per_test": "0.000418", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.005, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.61, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.61", + "raw_stderr": "1.826", + "latency": "62.738", + "cost_per_test": "0.000259", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.826, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.73, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.73", + "raw_stderr": "1.288", + "latency": "19.105", + "cost_per_test": "0.00021", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.288, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.606, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.606", + "raw_stderr": "1.346", + "latency": "47.81", + "cost_per_test": "0.000647", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.346, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.851, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.851", + "raw_stderr": "1.431", + "latency": "41.573", + "cost_per_test": "0.000192", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.431, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.328, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.328", + "raw_stderr": "2.216", + "latency": "63.49", + "cost_per_test": "0.000199", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.216, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 68.211, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "68.211", + "raw_stderr": "1.415", + "latency": "38.848", + "cost_per_test": "0.000331", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.415, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.71, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.71", + "raw_stderr": "0.889", + "latency": "28.831", + "cost_per_test": "0.000321", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.889, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.602, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.602", + "raw_stderr": "1.357", + "latency": "38.008", + "cost_per_test": "0.000157", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.357, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.056, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.056", + "raw_stderr": "0.381", + "latency": "40.772", + "cost_per_test": "0.000283", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.381, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.158, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.158", + "raw_stderr": "1.831", + "latency": "51.886", + "cost_per_test": "0.000163", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.831, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.299, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.299", + "raw_stderr": "0.945", + "latency": "31.74", + "cost_per_test": "0.00037", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.945, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:fireworks/deepseek-v3p2:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.586, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.586", + "raw_stderr": "1.404", + "latency": "41.38", + "cost_per_test": "0.000171", + "temperature": "1", + "max_output_tokens": "20480", + "reasoning_effort": "none", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.404, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 20480, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "none" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/gpt-oss-120b/0f8a6520-955b-4e13-bf31-dc532665ee39.json b/data/vals_ai/fireworks/gpt-oss-120b/0f8a6520-955b-4e13-bf31-dc532665ee39.json new file mode 100644 index 0000000000000000000000000000000000000000..28063629ce885d7a7ed652ebf4d04dae2e76e69d --- /dev/null +++ b/data/vals_ai/fireworks/gpt-oss-120b/0f8a6520-955b-4e13-bf31-dc532665ee39.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/fireworks_gpt-oss-120b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-oss-120b", + "id": "fireworks/gpt-oss-120b", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/gpt-oss-120b", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 86.179, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.179", + "raw_stderr": "1.039", + "latency": "6.163", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.039, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.395, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.395", + "raw_stderr": "0.245", + "latency": "6.506", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.245, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 76.602, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "76.602", + "raw_stderr": "0.785", + "latency": "5.366", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.785, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 75.938, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "75.938", + "raw_stderr": "0.517", + "latency": "7.852", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.517, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 70.893, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "70.893", + "raw_stderr": "1.158", + "latency": "8.044", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.158, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 68.62, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "68.62", + "raw_stderr": "1.891", + "latency": "13.18", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.891, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/gpt-oss-120b/27559250-fbd2-4563-808a-5198cced4165.json b/data/vals_ai/fireworks/gpt-oss-120b/27559250-fbd2-4563-808a-5198cced4165.json new file mode 100644 index 0000000000000000000000000000000000000000..ef25de505c49dc999c8e91b23ef4fd69bf4d609a --- /dev/null +++ b/data/vals_ai/fireworks/gpt-oss-120b/27559250-fbd2-4563-808a-5198cced4165.json @@ -0,0 +1,351 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/fireworks_gpt-oss-120b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-oss-120b", + "id": "fireworks/gpt-oss-120b", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/gpt-oss-120b", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 14.286, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "14.286", + "raw_stderr": "5.465", + "latency": "284.878", + "cost_per_test": "0.208965", + "max_output_tokens": "32768", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 5.465, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 27.586, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "27.586", + "raw_stderr": "2.772", + "latency": "266.875", + "cost_per_test": "0.173577", + "max_output_tokens": "32768", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.772, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 46.392, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "46.392", + "raw_stderr": "3.59", + "latency": "235.427", + "cost_per_test": "0.160298", + "max_output_tokens": "32768", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.59, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "313.243", + "cost_per_test": "0.204494", + "max_output_tokens": "32768", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.6, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.6", + "raw_stderr": "2.114", + "latency": "256.464", + "cost_per_test": "0.171583", + "max_output_tokens": "32768", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.114, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/gpt-oss-120b/28d33b71-5cb6-4b21-910c-dd60678aed53.json b/data/vals_ai/fireworks/gpt-oss-120b/28d33b71-5cb6-4b21-910c-dd60678aed53.json new file mode 100644 index 0000000000000000000000000000000000000000..1cb0ededf2ce08c5b06a15de1df08ea239915690 --- /dev/null +++ b/data/vals_ai/fireworks/gpt-oss-120b/28d33b71-5cb6-4b21-910c-dd60678aed53.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/fireworks_gpt-oss-120b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-oss-120b", + "id": "fireworks/gpt-oss-120b", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/gpt-oss-120b", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.447, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.447", + "raw_stderr": "0.689", + "latency": "39.979", + "cost_per_test": "0.001366", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.689, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 64.571, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "64.571", + "raw_stderr": "2.557", + "latency": "150.782", + "cost_per_test": "0.017407", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.557, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 86.684, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "86.684", + "raw_stderr": "1.736", + "latency": "54.348", + "cost_per_test": "0.006685", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.736, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 83.234, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "83.234", + "raw_stderr": "1.055", + "latency": "81.703", + "cost_per_test": "0.008486", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.055, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/gpt-oss-120b/6dd2a4f8-9512-4f7f-8fe0-f62bbe0e0f91.json b/data/vals_ai/fireworks/gpt-oss-120b/6dd2a4f8-9512-4f7f-8fe0-f62bbe0e0f91.json new file mode 100644 index 0000000000000000000000000000000000000000..aa8bcee5382327b6058bf78455015a138dc37b83 --- /dev/null +++ b/data/vals_ai/fireworks/gpt-oss-120b/6dd2a4f8-9512-4f7f-8fe0-f62bbe0e0f91.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/fireworks_gpt-oss-120b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-oss-120b", + "id": "fireworks/gpt-oss-120b", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/gpt-oss-120b", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.098, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.098", + "raw_stderr": "1.115", + "latency": "18.209", + "cost_per_test": "0.001015", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.115, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.452, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.452", + "raw_stderr": "1.199", + "latency": "60.567", + "cost_per_test": "0.002055", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.199, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.809, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.809", + "raw_stderr": "0.978", + "latency": "27.523", + "cost_per_test": "0.001869", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.978, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.122, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.122", + "raw_stderr": "1.758", + "latency": "20.726", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.758, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.242, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.242", + "raw_stderr": "1.266", + "latency": "42.971", + "cost_per_test": "0.001145", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.266, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.517, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.517", + "raw_stderr": "1.455", + "latency": "49.348", + "cost_per_test": "0.004726", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.455, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.795, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.795", + "raw_stderr": "1.498", + "latency": "40.606", + "cost_per_test": "0.001217", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.498, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 68.241, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "68.241", + "raw_stderr": "2.385", + "latency": "29.697", + "cost_per_test": "0.001333", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.385, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 59.037, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "59.037", + "raw_stderr": "1.483", + "latency": "69.15", + "cost_per_test": "0.002942", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.483, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.338, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.338", + "raw_stderr": "0.685", + "latency": "17.095", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.685, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.81, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.81", + "raw_stderr": "1.448", + "latency": "38.694", + "cost_per_test": "0.001439", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.448, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.166, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.166", + "raw_stderr": "0.395", + "latency": "35.41", + "cost_per_test": "0.001438", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.395, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 67.134, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "67.134", + "raw_stderr": "2.103", + "latency": "10.104", + "cost_per_test": "0.00128", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.103, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.913, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.913", + "raw_stderr": "0.945", + "latency": "18.674", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.945, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.82, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.82", + "raw_stderr": "1.477", + "latency": "52.376", + "cost_per_test": "0.001118", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.477, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/gpt-oss-120b/7b948422-5b76-47a7-972b-dec749f40052.json b/data/vals_ai/fireworks/gpt-oss-120b/7b948422-5b76-47a7-972b-dec749f40052.json new file mode 100644 index 0000000000000000000000000000000000000000..d05967a99832d3c2337cc2d49d01e4fec14eb9d6 --- /dev/null +++ b/data/vals_ai/fireworks/gpt-oss-120b/7b948422-5b76-47a7-972b-dec749f40052.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/fireworks_gpt-oss-120b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-oss-120b", + "id": "fireworks/gpt-oss-120b", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/gpt-oss-120b", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 75.758, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.758", + "raw_stderr": "3.487", + "latency": "46.429", + "cost_per_test": "0.005601", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.487, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 81.313, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.313", + "raw_stderr": "2.853", + "latency": "65.923", + "cost_per_test": "0.006824", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.853, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 78.536, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.536", + "raw_stderr": "2.253", + "latency": "56.176", + "cost_per_test": "0.006212", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.253, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/gpt-oss-120b/8f49e835-dc92-4876-a5c6-8efec0d94f27.json b/data/vals_ai/fireworks/gpt-oss-120b/8f49e835-dc92-4876-a5c6-8efec0d94f27.json new file mode 100644 index 0000000000000000000000000000000000000000..2f17215cc43194de9decb71293358823b1ad7a73 --- /dev/null +++ b/data/vals_ai/fireworks/gpt-oss-120b/8f49e835-dc92-4876-a5c6-8efec0d94f27.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/fireworks_gpt-oss-120b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-oss-120b", + "id": "fireworks/gpt-oss-120b", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/gpt-oss-120b", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.888, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.888", + "raw_stderr": "1.658", + "latency": "46.794", + "cost_per_test": "0.000868", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.658, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 51.981, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "51.981", + "raw_stderr": "1.706", + "latency": "41.001", + "cost_per_test": "0.017607", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.706, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 58.236, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.236", + "raw_stderr": "0.968", + "latency": "42.722", + "cost_per_test": "0.009601", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.968, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.839, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.839", + "raw_stderr": "1.666", + "latency": "40.371", + "cost_per_test": "0.010328", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.666, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/gpt-oss-120b/fcee7f1a-bd5b-4f27-9f55-4e21a5ca6bbf.json b/data/vals_ai/fireworks/gpt-oss-120b/fcee7f1a-bd5b-4f27-9f55-4e21a5ca6bbf.json new file mode 100644 index 0000000000000000000000000000000000000000..b84fdd5ca4bc7790f0af50b872cacb242220daea --- /dev/null +++ b/data/vals_ai/fireworks/gpt-oss-120b/fcee7f1a-bd5b-4f27-9f55-4e21a5ca6bbf.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/fireworks_gpt-oss-120b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-oss-120b", + "id": "fireworks/gpt-oss-120b", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/gpt-oss-120b", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 59.444, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.444", + "raw_stderr": "1.404", + "latency": "47.661", + "cost_per_test": "0.002958", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.404, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 71.586, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.586", + "raw_stderr": "0.878", + "latency": "47.661", + "cost_per_test": "0.002958", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.878, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:fireworks/gpt-oss-120b:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 83.729, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "83.729", + "raw_stderr": "1.055", + "latency": "47.661", + "cost_per_test": "0.002958", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.055, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/gpt-oss-20b/9a15bbed-7beb-4796-bda8-9f81ad943a31.json b/data/vals_ai/fireworks/gpt-oss-20b/9a15bbed-7beb-4796-bda8-9f81ad943a31.json new file mode 100644 index 0000000000000000000000000000000000000000..3423c304b1ddf99c9aabd56b2cd549d2afb4544f --- /dev/null +++ b/data/vals_ai/fireworks/gpt-oss-20b/9a15bbed-7beb-4796-bda8-9f81ad943a31.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/fireworks_gpt-oss-20b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-oss-20b", + "id": "fireworks/gpt-oss-20b", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/gpt-oss-20b", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 97.826, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "97.826", + "raw_stderr": "0.813", + "latency": "15.91", + "cost_per_test": "0.000953", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.813, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 56.389, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "56.389", + "raw_stderr": "2.614", + "latency": "205.668", + "cost_per_test": "0.013366", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.614, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 86.945, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "86.945", + "raw_stderr": "1.722", + "latency": "106.782", + "cost_per_test": "0.005799", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.722, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 80.387, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "80.387", + "raw_stderr": "1.078", + "latency": "109.453", + "cost_per_test": "0.006706", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.078, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/gpt-oss-20b/b8c644a5-4695-4920-83a6-68af34e57897.json b/data/vals_ai/fireworks/gpt-oss-20b/b8c644a5-4695-4920-83a6-68af34e57897.json new file mode 100644 index 0000000000000000000000000000000000000000..92493038ec92ffc4c0824357467008b2b7eacf67 --- /dev/null +++ b/data/vals_ai/fireworks/gpt-oss-20b/b8c644a5-4695-4920-83a6-68af34e57897.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/fireworks_gpt-oss-20b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-oss-20b", + "id": "fireworks/gpt-oss-20b", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/gpt-oss-20b", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.379, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.379", + "raw_stderr": "1.53", + "latency": "19.51", + "cost_per_test": "0.000667", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.53, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.355, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.355", + "raw_stderr": "1.538", + "latency": "20.238", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.538, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.389, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.389", + "raw_stderr": "1.306", + "latency": "31.536", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.306, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.22, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.22", + "raw_stderr": "2.036", + "latency": "23.628", + "cost_per_test": "0.001057", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.036, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.436, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.436", + "raw_stderr": "1.535", + "latency": "14.323", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 56.45, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "56.45", + "raw_stderr": "1.603", + "latency": "50.126", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.603, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 68.46, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "68.46", + "raw_stderr": "1.673", + "latency": "45.25", + "cost_per_test": "0.000695", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.673, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 61.68, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "61.68", + "raw_stderr": "2.5", + "latency": "18.353", + "cost_per_test": "0.000833", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.5, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 44.596, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "44.596", + "raw_stderr": "1.484", + "latency": "73.861", + "cost_per_test": "0.001907", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.484, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.935, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.935", + "raw_stderr": "1.014", + "latency": "20.684", + "cost_per_test": "0.000939", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.014, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 66.126, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.126", + "raw_stderr": "1.576", + "latency": "16.455", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.576, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.636, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.636", + "raw_stderr": "0.447", + "latency": "30.207", + "cost_per_test": "0.00062", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.447, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 58.717, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "58.717", + "raw_stderr": "2.214", + "latency": "24.252", + "cost_per_test": "0.000851", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.214, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.602, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.602", + "raw_stderr": "1.154", + "latency": "22.214", + "cost_per_test": "0.00112", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.154, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.554, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.554", + "raw_stderr": "1.661", + "latency": "42.466", + "cost_per_test": "0.000615", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.661, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/gpt-oss-20b/bf4b33e1-a53a-483c-946c-914ef390bfa5.json b/data/vals_ai/fireworks/gpt-oss-20b/bf4b33e1-a53a-483c-946c-914ef390bfa5.json new file mode 100644 index 0000000000000000000000000000000000000000..a1cb8d71888e012c5876bbfcd35606a180bd4d9d --- /dev/null +++ b/data/vals_ai/fireworks/gpt-oss-20b/bf4b33e1-a53a-483c-946c-914ef390bfa5.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/fireworks_gpt-oss-20b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-oss-20b", + "id": "fireworks/gpt-oss-20b", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/gpt-oss-20b", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.18, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.18", + "raw_stderr": "1.214", + "latency": "8.088", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.214, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 69.903, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "69.903", + "raw_stderr": "0.28", + "latency": "6.443", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.28, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 74.168, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "74.168", + "raw_stderr": "0.84", + "latency": "6.529", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.84, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 70.849, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "70.849", + "raw_stderr": "0.61", + "latency": "7.187", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.61, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 69.296, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "69.296", + "raw_stderr": "1.379", + "latency": "4.381", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.379, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 58.697, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "58.697", + "raw_stderr": "2.266", + "latency": "10.493", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.266, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/gpt-oss-20b/cd041855-3118-4e66-b08b-b599dad38025.json b/data/vals_ai/fireworks/gpt-oss-20b/cd041855-3118-4e66-b08b-b599dad38025.json new file mode 100644 index 0000000000000000000000000000000000000000..6dc8f04a4684eef2d33a82b4c2358f3d519fff2f --- /dev/null +++ b/data/vals_ai/fireworks/gpt-oss-20b/cd041855-3118-4e66-b08b-b599dad38025.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/fireworks_gpt-oss-20b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-oss-20b", + "id": "fireworks/gpt-oss-20b", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/gpt-oss-20b", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 50.695, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "50.695", + "raw_stderr": "1.43", + "latency": "49.004", + "cost_per_test": "0.002254", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.43, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 63.696, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.696", + "raw_stderr": "0.936", + "latency": "49.004", + "cost_per_test": "0.002254", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.936, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 76.697, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "76.697", + "raw_stderr": "1.209", + "latency": "49.004", + "cost_per_test": "0.002254", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.209, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/gpt-oss-20b/e625464a-b47b-4178-b73d-2a746ac01aad.json b/data/vals_ai/fireworks/gpt-oss-20b/e625464a-b47b-4178-b73d-2a746ac01aad.json new file mode 100644 index 0000000000000000000000000000000000000000..94c8ecba84610528125509ccf2ca20df4b0b85b1 --- /dev/null +++ b/data/vals_ai/fireworks/gpt-oss-20b/e625464a-b47b-4178-b73d-2a746ac01aad.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/fireworks_gpt-oss-20b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-oss-20b", + "id": "fireworks/gpt-oss-20b", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/gpt-oss-20b", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 65.152, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "65.152", + "raw_stderr": "3.553", + "latency": "120.357", + "cost_per_test": "0.00604", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.553, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 72.727, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.727", + "raw_stderr": "3.419", + "latency": "169.828", + "cost_per_test": "0.007436", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.419, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 68.94, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "68.94", + "raw_stderr": "2.465", + "latency": "145.092", + "cost_per_test": "0.006738", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.465, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/gpt-oss-20b/fed41dc3-d684-454d-96b5-2eb00de89ad6.json b/data/vals_ai/fireworks/gpt-oss-20b/fed41dc3-d684-454d-96b5-2eb00de89ad6.json new file mode 100644 index 0000000000000000000000000000000000000000..6ef0440f8cd98b3dce47b0ec6cc1fe8c2c1b5b7a --- /dev/null +++ b/data/vals_ai/fireworks/gpt-oss-20b/fed41dc3-d684-454d-96b5-2eb00de89ad6.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/fireworks_gpt-oss-20b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-oss-20b", + "id": "fireworks/gpt-oss-20b", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/gpt-oss-20b", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 58.974, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.974", + "raw_stderr": "1.679", + "latency": "32.257", + "cost_per_test": "0.000678", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.679, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 46.97, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "46.97", + "raw_stderr": "1.704", + "latency": "26.239", + "cost_per_test": "0.007679", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.704, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 53.147, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "53.147", + "raw_stderr": "0.979", + "latency": "33.968", + "cost_per_test": "0.004549", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.979, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:fireworks/gpt-oss-20b:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 53.497, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "53.497", + "raw_stderr": "1.703", + "latency": "43.407", + "cost_per_test": "0.005289", + "temperature": "1", + "max_output_tokens": "128000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.703, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/llama4-maverick-instruct-basic/177feeb9-5020-475c-ba80-0892aa44f1b4.json b/data/vals_ai/fireworks/llama4-maverick-instruct-basic/177feeb9-5020-475c-ba80-0892aa44f1b4.json new file mode 100644 index 0000000000000000000000000000000000000000..78c78d56f148f9cfef2c2c91a273a49ec3566322 --- /dev/null +++ b/data/vals_ai/fireworks/llama4-maverick-instruct-basic/177feeb9-5020-475c-ba80-0892aa44f1b4.json @@ -0,0 +1,99 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/fireworks_llama4-maverick-instruct-basic/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "llama4-maverick-instruct-basic", + "id": "fireworks/llama4-maverick-instruct-basic", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/llama4-maverick-instruct-basic", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 71.693, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.693", + "raw_stderr": "1.083", + "latency": "7.908", + "cost_per_test": "0.001078", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.083, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/llama4-maverick-instruct-basic/69518fd8-df7c-4642-a5ae-96edaefd316b.json b/data/vals_ai/fireworks/llama4-maverick-instruct-basic/69518fd8-df7c-4642-a5ae-96edaefd316b.json new file mode 100644 index 0000000000000000000000000000000000000000..2172e0747b00fc7807fc5af685292fff8e4bbbf1 --- /dev/null +++ b/data/vals_ai/fireworks/llama4-maverick-instruct-basic/69518fd8-df7c-4642-a5ae-96edaefd316b.json @@ -0,0 +1,288 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/fireworks_llama4-maverick-instruct-basic/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "llama4-maverick-instruct-basic", + "id": "fireworks/llama4-maverick-instruct-basic", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/llama4-maverick-instruct-basic", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 87.888, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "87.888", + "raw_stderr": "1.818", + "latency": "6.489", + "cost_per_test": "0.000672", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.818, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 10.0, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "10", + "raw_stderr": "1.604", + "latency": "10.527", + "cost_per_test": "0.000991", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.604, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 43.864, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "43.864", + "raw_stderr": "2.536", + "latency": "11.31", + "cost_per_test": "0.000832", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.536, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 47.251, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "47.251", + "raw_stderr": "1.169", + "latency": "9.442", + "cost_per_test": "0.000832", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.169, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/llama4-maverick-instruct-basic/87303892-af54-476f-b511-5768d6dcb808.json b/data/vals_ai/fireworks/llama4-maverick-instruct-basic/87303892-af54-476f-b511-5768d6dcb808.json new file mode 100644 index 0000000000000000000000000000000000000000..b37632b69c0ce48eb5888dc6ee63fdea72258238 --- /dev/null +++ b/data/vals_ai/fireworks/llama4-maverick-instruct-basic/87303892-af54-476f-b511-5768d6dcb808.json @@ -0,0 +1,216 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/fireworks_llama4-maverick-instruct-basic/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "llama4-maverick-instruct-basic", + "id": "fireworks/llama4-maverick-instruct-basic", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/llama4-maverick-instruct-basic", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 67.17, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.17", + "raw_stderr": "1.324", + "latency": "2.435", + "cost_per_test": "0.00073", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.324, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 58.506, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.506", + "raw_stderr": "0.967", + "latency": "2.716", + "cost_per_test": "0.000752", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.967, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 49.841, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "49.841", + "raw_stderr": "1.41", + "latency": "2.998", + "cost_per_test": "0.000774", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.41, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/llama4-maverick-instruct-basic/99f88210-ee1b-4914-ab82-6683b33f989a.json b/data/vals_ai/fireworks/llama4-maverick-instruct-basic/99f88210-ee1b-4914-ab82-6683b33f989a.json new file mode 100644 index 0000000000000000000000000000000000000000..537a8b3fa906dfc2e3f8850315d58bce6ebe5142 --- /dev/null +++ b/data/vals_ai/fireworks/llama4-maverick-instruct-basic/99f88210-ee1b-4914-ab82-6683b33f989a.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/fireworks_llama4-maverick-instruct-basic/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "llama4-maverick-instruct-basic", + "id": "fireworks/llama4-maverick-instruct-basic", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/llama4-maverick-instruct-basic", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 54.219, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "54.219", + "raw_stderr": "1.871", + "latency": "25.048", + "cost_per_test": "0.00246", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.871, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/llama4-maverick-instruct-basic/b002e6e1-42fe-4271-8d89-dae19b17444c.json b/data/vals_ai/fireworks/llama4-maverick-instruct-basic/b002e6e1-42fe-4271-8d89-dae19b17444c.json new file mode 100644 index 0000000000000000000000000000000000000000..53a47ea6c3c2cdd12d0028222cd35ff785435dde --- /dev/null +++ b/data/vals_ai/fireworks/llama4-maverick-instruct-basic/b002e6e1-42fe-4271-8d89-dae19b17444c.json @@ -0,0 +1,346 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/fireworks_llama4-maverick-instruct-basic/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "llama4-maverick-instruct-basic", + "id": "fireworks/llama4-maverick-instruct-basic", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/llama4-maverick-instruct-basic", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 34.921, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "34.921", + "raw_stderr": "5.982", + "latency": "43.826", + "cost_per_test": "0.003103", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 5.982, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 8.333, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "8.333", + "raw_stderr": "3.535", + "latency": "36.462", + "cost_per_test": "0.004148", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 2.778, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "2.778", + "raw_stderr": "2.66", + "latency": "75.714", + "cost_per_test": "0.018248", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.66, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 63.852, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "63.852", + "raw_stderr": "5.396", + "latency": "50.49", + "cost_per_test": "0.006765", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 5.396, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 35.702, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "35.702", + "raw_stderr": "2.933", + "latency": "43.593", + "cost_per_test": "0.004672", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.933, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/llama4-maverick-instruct-basic/babb959c-6118-436c-8709-81674d65616c.json b/data/vals_ai/fireworks/llama4-maverick-instruct-basic/babb959c-6118-436c-8709-81674d65616c.json new file mode 100644 index 0000000000000000000000000000000000000000..02adb1e282f922d0d50c1a151f1adbdf1c026160 --- /dev/null +++ b/data/vals_ai/fireworks/llama4-maverick-instruct-basic/babb959c-6118-436c-8709-81674d65616c.json @@ -0,0 +1,408 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/fireworks_llama4-maverick-instruct-basic/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "llama4-maverick-instruct-basic", + "id": "fireworks/llama4-maverick-instruct-basic", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/llama4-maverick-instruct-basic", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.427, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.427", + "raw_stderr": "1.022", + "latency": "0.165", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.022, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 76.755, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "76.755", + "raw_stderr": "0.196", + "latency": "0.169", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.196, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 76.518, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "76.518", + "raw_stderr": "0.635", + "latency": "1.06", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.635, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.812, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.812", + "raw_stderr": "0.42", + "latency": "0.472", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.42, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.766, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.766", + "raw_stderr": "0.819", + "latency": "0.162", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.819, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 73.592, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "73.592", + "raw_stderr": "1.499", + "latency": "0.806", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.499, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/llama4-maverick-instruct-basic/c7b15bda-3b49-4d0e-bd5b-32cd7b749d00.json b/data/vals_ai/fireworks/llama4-maverick-instruct-basic/c7b15bda-3b49-4d0e-bd5b-32cd7b749d00.json new file mode 100644 index 0000000000000000000000000000000000000000..1684aba8575de368ddfc6f5fe72f8531a8c1c0b1 --- /dev/null +++ b/data/vals_ai/fireworks/llama4-maverick-instruct-basic/c7b15bda-3b49-4d0e-bd5b-32cd7b749d00.json @@ -0,0 +1,225 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/fireworks_llama4-maverick-instruct-basic/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "llama4-maverick-instruct-basic", + "id": "fireworks/llama4-maverick-instruct-basic", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/llama4-maverick-instruct-basic", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 68.182, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "68.182", + "raw_stderr": "3.31", + "latency": "9.54", + "cost_per_test": "0.001096", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.31, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 70.707, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.707", + "raw_stderr": "3.337", + "latency": "10.602", + "cost_per_test": "0.000831", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.337, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 69.444, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.444", + "raw_stderr": "2.35", + "latency": "10.071", + "cost_per_test": "0.000964", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.35, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/llama4-maverick-instruct-basic/e523f9af-c621-4539-8dc3-54e432907f2b.json b/data/vals_ai/fireworks/llama4-maverick-instruct-basic/e523f9af-c621-4539-8dc3-54e432907f2b.json new file mode 100644 index 0000000000000000000000000000000000000000..79ff0e66eb67c94caf8f4cc8f1e2e6ea90491539 --- /dev/null +++ b/data/vals_ai/fireworks/llama4-maverick-instruct-basic/e523f9af-c621-4539-8dc3-54e432907f2b.json @@ -0,0 +1,216 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/fireworks_llama4-maverick-instruct-basic/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "llama4-maverick-instruct-basic", + "id": "fireworks/llama4-maverick-instruct-basic", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/llama4-maverick-instruct-basic", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 53.966, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "53.966", + "raw_stderr": "1.425", + "latency": "25.143", + "cost_per_test": "0.000527", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.425, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 66.558, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.558", + "raw_stderr": "0.919", + "latency": "25.143", + "cost_per_test": "0.000527", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.919, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 79.15, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "79.15", + "raw_stderr": "1.162", + "latency": "25.143", + "cost_per_test": "0.000527", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.162, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/llama4-maverick-instruct-basic/f7128cc4-e508-494f-a144-c263733e1b1e.json b/data/vals_ai/fireworks/llama4-maverick-instruct-basic/f7128cc4-e508-494f-a144-c263733e1b1e.json new file mode 100644 index 0000000000000000000000000000000000000000..f646ea81e8f66cb6feafd3c6696e77f8db74425b --- /dev/null +++ b/data/vals_ai/fireworks/llama4-maverick-instruct-basic/f7128cc4-e508-494f-a144-c263733e1b1e.json @@ -0,0 +1,981 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/fireworks_llama4-maverick-instruct-basic/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "llama4-maverick-instruct-basic", + "id": "fireworks/llama4-maverick-instruct-basic", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/llama4-maverick-instruct-basic", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.982, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.982", + "raw_stderr": "1.169", + "latency": "4.832", + "cost_per_test": "0.000689", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.169, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.016, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.016", + "raw_stderr": "1.337", + "latency": "4.606", + "cost_per_test": "0.000542", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.337, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.866, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.866", + "raw_stderr": "1.035", + "latency": "6.509", + "cost_per_test": "0.00085", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.035, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.463, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.463", + "raw_stderr": "1.919", + "latency": "4.686", + "cost_per_test": "0.000735", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.919, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.019, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.019", + "raw_stderr": "1.194", + "latency": "5.183", + "cost_per_test": "0.000626", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.194, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 70.898, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.898", + "raw_stderr": "1.459", + "latency": "7.294", + "cost_per_test": "0.000875", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.459, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.284, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.284", + "raw_stderr": "1.487", + "latency": "5.289", + "cost_per_test": "0.000695", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.487, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 69.816, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.816", + "raw_stderr": "2.352", + "latency": "28.434", + "cost_per_test": "0.000734", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.352, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 60.763, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "60.763", + "raw_stderr": "1.472", + "latency": "3.779", + "cost_per_test": "0.000633", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.472, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.563, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.563", + "raw_stderr": "0.832", + "latency": "5.065", + "cost_per_test": "0.000803", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.832, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.706, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.706", + "raw_stderr": "1.369", + "latency": "5.163", + "cost_per_test": "0.000541", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.369, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.415, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.415", + "raw_stderr": "0.395", + "latency": "6.803", + "cost_per_test": "0.000686", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.395, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.747, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.747", + "raw_stderr": "1.97", + "latency": "4.691", + "cost_per_test": "0.000595", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.97, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.605, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.605", + "raw_stderr": "0.945", + "latency": "5.019", + "cost_per_test": "0.000682", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.945, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.078, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.078", + "raw_stderr": "1.387", + "latency": "4.688", + "cost_per_test": "0.000597", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.387, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/llama4-maverick-instruct-basic/febfc3bb-d744-40c5-b268-afef8c8b18dc.json b/data/vals_ai/fireworks/llama4-maverick-instruct-basic/febfc3bb-d744-40c5-b268-afef8c8b18dc.json new file mode 100644 index 0000000000000000000000000000000000000000..af1fc9e7d00803fc29911fb649c29c1945348b46 --- /dev/null +++ b/data/vals_ai/fireworks/llama4-maverick-instruct-basic/febfc3bb-d744-40c5-b268-afef8c8b18dc.json @@ -0,0 +1,276 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/fireworks_llama4-maverick-instruct-basic/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "llama4-maverick-instruct-basic", + "id": "fireworks/llama4-maverick-instruct-basic", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/llama4-maverick-instruct-basic", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 55.828, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "55.828", + "raw_stderr": "1.695", + "latency": "3.167", + "cost_per_test": "0.000535", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.695, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 45.221, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "45.221", + "raw_stderr": "1.699", + "latency": "6.636", + "cost_per_test": "0.028134", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.699, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 49.728, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "49.728", + "raw_stderr": "0.981", + "latency": "4.724", + "cost_per_test": "0.014178", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.981, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:fireworks/llama4-maverick-instruct-basic:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 48.135, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "48.135", + "raw_stderr": "1.706", + "latency": "4.369", + "cost_per_test": "0.013864", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.706, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/qwen3-235b-a22b/6480b380-e272-4c80-bc0b-8447ee19cb17.json b/data/vals_ai/fireworks/qwen3-235b-a22b/6480b380-e272-4c80-bc0b-8447ee19cb17.json new file mode 100644 index 0000000000000000000000000000000000000000..30bb89a0409690cf222783ea3e1c7f6be744f7bc --- /dev/null +++ b/data/vals_ai/fireworks/qwen3-235b-a22b/6480b380-e272-4c80-bc0b-8447ee19cb17.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/fireworks_qwen3-235b-a22b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3-235b-a22b", + "id": "fireworks/qwen3-235b-a22b", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/qwen3-235b-a22b", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.795, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.795", + "raw_stderr": "1.15", + "latency": "42.234", + "cost_per_test": "0.001768", + "temperature": "0.6", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.15, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.692, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.692", + "raw_stderr": "1.228", + "latency": "58.89", + "cost_per_test": "0.002274", + "temperature": "0.6", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.228, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.163, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.163", + "raw_stderr": "0.978", + "latency": "59.35", + "cost_per_test": "0.003506", + "temperature": "0.6", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.978, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.366, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.366", + "raw_stderr": "1.781", + "latency": "45.551", + "cost_per_test": "0.002233", + "temperature": "0.6", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.781, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.256, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.256", + "raw_stderr": "1.223", + "latency": "47.446", + "cost_per_test": "0.001695", + "temperature": "0.6", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.223, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.154, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.154", + "raw_stderr": "1.31", + "latency": "110.064", + "cost_per_test": "0.005318", + "temperature": "0.6", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.31, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.061, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.061", + "raw_stderr": "1.513", + "latency": "31.091", + "cost_per_test": "0.001536", + "temperature": "0.6", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.513, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.654, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.654", + "raw_stderr": "2.322", + "latency": "37.78", + "cost_per_test": "0.001624", + "temperature": "0.6", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.322, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 59.401, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "59.401", + "raw_stderr": "1.49", + "latency": "71.288", + "cost_per_test": "0.002407", + "temperature": "0.6", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.49, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.562, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.562", + "raw_stderr": "0.852", + "latency": "48.296", + "cost_per_test": "0.002828", + "temperature": "0.6", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.852, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.03, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.03", + "raw_stderr": "1.364", + "latency": "40.542", + "cost_per_test": "0.001379", + "temperature": "0.6", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.364, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.246, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.246", + "raw_stderr": "0.387", + "latency": "52.354", + "cost_per_test": "0.002316", + "temperature": "0.6", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.387, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 72.345, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.345", + "raw_stderr": "2.028", + "latency": "40.377", + "cost_per_test": "0.001476", + "temperature": "0.6", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.028, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.762, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.762", + "raw_stderr": "0.818", + "latency": "57.666", + "cost_per_test": "0.003131", + "temperature": "0.6", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.818, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.206, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.206", + "raw_stderr": "1.369", + "latency": "42.376", + "cost_per_test": "0.001254", + "temperature": "0.6", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.369, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/qwen3-235b-a22b/931b49b5-750d-4ee3-bcff-6a4157a69b52.json b/data/vals_ai/fireworks/qwen3-235b-a22b/931b49b5-750d-4ee3-bcff-6a4157a69b52.json new file mode 100644 index 0000000000000000000000000000000000000000..dbb76a2b981a977cc4a37911f58ab33231e2a53d --- /dev/null +++ b/data/vals_ai/fireworks/qwen3-235b-a22b/931b49b5-750d-4ee3-bcff-6a4157a69b52.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/fireworks_qwen3-235b-a22b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3-235b-a22b", + "id": "fireworks/qwen3-235b-a22b", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/qwen3-235b-a22b", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "303.027", + "cost_per_test": "0.066363", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "402.103", + "cost_per_test": "0.152785", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "352.565", + "cost_per_test": "0.109574", + "temperature": "1", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/qwen3-235b-a22b/9ea0d6f7-2616-4226-b26b-04cb07878786.json b/data/vals_ai/fireworks/qwen3-235b-a22b/9ea0d6f7-2616-4226-b26b-04cb07878786.json new file mode 100644 index 0000000000000000000000000000000000000000..afe047c007f8461ec0a43f2f9fed2d668e2bfe7f --- /dev/null +++ b/data/vals_ai/fireworks/qwen3-235b-a22b/9ea0d6f7-2616-4226-b26b-04cb07878786.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/fireworks_qwen3-235b-a22b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3-235b-a22b", + "id": "fireworks/qwen3-235b-a22b", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/qwen3-235b-a22b", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 96.273, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "96.273", + "raw_stderr": "1.056", + "latency": "142.152", + "cost_per_test": "0.006518", + "temperature": "1", + "max_output_tokens": "38912", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.056, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 38912, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 35.429, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "35.429", + "raw_stderr": "2.557", + "latency": "556.283", + "cost_per_test": "0.01294", + "temperature": "1", + "max_output_tokens": "38912", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.557, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 38912, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 80.157, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "80.157", + "raw_stderr": "2.038", + "latency": "590.013", + "cost_per_test": "0.014575", + "temperature": "1", + "max_output_tokens": "38912", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.038, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 38912, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 70.62, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "70.62", + "raw_stderr": "1.145", + "latency": "429.483", + "cost_per_test": "0.011344", + "temperature": "1", + "max_output_tokens": "38912", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.145, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 38912, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/qwen3-235b-a22b/a5d958d4-4ee7-4b4d-8a5f-78202a9bdc17.json b/data/vals_ai/fireworks/qwen3-235b-a22b/a5d958d4-4ee7-4b4d-8a5f-78202a9bdc17.json new file mode 100644 index 0000000000000000000000000000000000000000..8001eaee428e771ae10e1259873b4f5b74216377 --- /dev/null +++ b/data/vals_ai/fireworks/qwen3-235b-a22b/a5d958d4-4ee7-4b4d-8a5f-78202a9bdc17.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/fireworks_qwen3-235b-a22b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3-235b-a22b", + "id": "fireworks/qwen3-235b-a22b", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/qwen3-235b-a22b", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.164, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.164", + "raw_stderr": "0.76", + "latency": "16.146", + "temperature": "0.6", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.76, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.681, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.681", + "raw_stderr": "0.225", + "latency": "11.547", + "temperature": "0.6", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.225, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 86.016, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.016", + "raw_stderr": "0.645", + "latency": "8.841", + "temperature": "0.6", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.645, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.179, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.179", + "raw_stderr": "0.441", + "latency": "12.366", + "temperature": "0.6", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.441, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.389, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.389", + "raw_stderr": "1.003", + "latency": "8.378", + "temperature": "0.6", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.003, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 70.646, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "70.646", + "raw_stderr": "1.674", + "latency": "16.92", + "temperature": "0.6", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.674, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/qwen3-235b-a22b/d6bf6a2b-d569-41ed-afcc-9ebb42a9cf50.json b/data/vals_ai/fireworks/qwen3-235b-a22b/d6bf6a2b-d569-41ed-afcc-9ebb42a9cf50.json new file mode 100644 index 0000000000000000000000000000000000000000..bd3eefc23867fd2802692364540a79ce54de9f88 --- /dev/null +++ b/data/vals_ai/fireworks/qwen3-235b-a22b/d6bf6a2b-d569-41ed-afcc-9ebb42a9cf50.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/fireworks_qwen3-235b-a22b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3-235b-a22b", + "id": "fireworks/qwen3-235b-a22b", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/qwen3-235b-a22b", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 59.035, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.035", + "raw_stderr": "1.406", + "latency": "92.742", + "cost_per_test": "0.004644", + "temperature": "0.6", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.406, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 70.646, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "70.646", + "raw_stderr": "0.89", + "latency": "92.742", + "cost_per_test": "0.004644", + "temperature": "0.6", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.89, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 82.257, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "82.257", + "raw_stderr": "1.092", + "latency": "92.742", + "cost_per_test": "0.004644", + "temperature": "0.6", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.092, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/fireworks/qwen3-235b-a22b/f3c37a95-cb0d-439c-8d0b-b831b64a43c5.json b/data/vals_ai/fireworks/qwen3-235b-a22b/f3c37a95-cb0d-439c-8d0b-b831b64a43c5.json new file mode 100644 index 0000000000000000000000000000000000000000..44b08dc3b1ba4a105d7ab17581517f1318e08a4c --- /dev/null +++ b/data/vals_ai/fireworks/qwen3-235b-a22b/f3c37a95-cb0d-439c-8d0b-b831b64a43c5.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/fireworks_qwen3-235b-a22b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "qwen3-235b-a22b", + "id": "fireworks/qwen3-235b-a22b", + "developer": "fireworks", + "additional_details": { + "vals_model_id": "fireworks/qwen3-235b-a22b", + "vals_provider": "Fireworks AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 68.182, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "68.182", + "raw_stderr": "3.375", + "latency": "279.853", + "cost_per_test": "0.005974", + "temperature": "0.6", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.375, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 72.222, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.222", + "raw_stderr": "3.337", + "latency": "332.583", + "cost_per_test": "0.006971", + "temperature": "0.6", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.337, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:fireworks/qwen3-235b-a22b:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 70.202, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.202", + "raw_stderr": "2.373", + "latency": "306.218", + "cost_per_test": "0.006472", + "temperature": "0.6", + "max_output_tokens": "16384", + "provider": "Fireworks AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.373, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-1.0-pro-002/4f6595ba-a757-4e29-b117-c634dfb2f48f.json b/data/vals_ai/google/gemini-1.0-pro-002/4f6595ba-a757-4e29-b117-c634dfb2f48f.json new file mode 100644 index 0000000000000000000000000000000000000000..fb95395fb9c66b4da14a3d1f2ba6118bd4a91ae7 --- /dev/null +++ b/data/vals_ai/google/gemini-1.0-pro-002/4f6595ba-a757-4e29-b117-c634dfb2f48f.json @@ -0,0 +1,408 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/google_gemini-1.0-pro-002/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-1.0-pro-002", + "id": "google/gemini-1.0-pro-002", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-1.0-pro-002", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:google/gemini-1.0-pro-002:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 74.505, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "74.505", + "raw_stderr": "1.106", + "latency": "0.386", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.106, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:google/gemini-1.0-pro-002:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 73.86, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "73.86", + "raw_stderr": "0.196", + "latency": "0.42", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.196, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:google/gemini-1.0-pro-002:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 72.707, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "72.707", + "raw_stderr": "0.677", + "latency": "0.622", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.677, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:google/gemini-1.0-pro-002:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 70.247, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "70.247", + "raw_stderr": "0.765", + "latency": "0.465", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.765, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:google/gemini-1.0-pro-002:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.78, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.78", + "raw_stderr": "0.848", + "latency": "0.436", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.848, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:google/gemini-1.0-pro-002:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 52.382, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "52.382", + "raw_stderr": "3.493", + "latency": "0.461", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.493, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-1.5-flash-001/d9398b3c-3473-4e98-befb-19fb0ca87212.json b/data/vals_ai/google/gemini-1.5-flash-001/d9398b3c-3473-4e98-befb-19fb0ca87212.json new file mode 100644 index 0000000000000000000000000000000000000000..2b0a66be835e94edaf0d0bd4ee3be8c517ab0246 --- /dev/null +++ b/data/vals_ai/google/gemini-1.5-flash-001/d9398b3c-3473-4e98-befb-19fb0ca87212.json @@ -0,0 +1,276 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/google_gemini-1.5-flash-001/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-1.5-flash-001", + "id": "google/gemini-1.5-flash-001", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-1.5-flash-001", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:google/gemini-1.5-flash-001:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 38.462, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "38.462", + "raw_stderr": "1.661", + "latency": "3.353", + "cost_per_test": "0.00014", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.661, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:google/gemini-1.5-flash-001:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 25.991, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "25.991", + "raw_stderr": "1.497", + "latency": "0.055", + "cost_per_test": "0.014092", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.497, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:google/gemini-1.5-flash-001:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 28.633, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "28.633", + "raw_stderr": "0.88", + "latency": "1.156", + "cost_per_test": "0.006413", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.88, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:google/gemini-1.5-flash-001:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 21.445, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "21.445", + "raw_stderr": "1.401", + "latency": "0.059", + "cost_per_test": "0.005006", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.401, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-1.5-flash-002/11214e11-546c-4943-8f38-d5f732fe6ded.json b/data/vals_ai/google/gemini-1.5-flash-002/11214e11-546c-4943-8f38-d5f732fe6ded.json new file mode 100644 index 0000000000000000000000000000000000000000..5c00d2cf678ec38486a9489bb9219300ce099c8d --- /dev/null +++ b/data/vals_ai/google/gemini-1.5-flash-002/11214e11-546c-4943-8f38-d5f732fe6ded.json @@ -0,0 +1,981 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/google_gemini-1.5-flash-002/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-1.5-flash-002", + "id": "google/gemini-1.5-flash-002", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-1.5-flash-002", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.706, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.706", + "raw_stderr": "1.421", + "latency": "1.504", + "cost_per_test": "0.000164", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.421, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.61, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.61", + "raw_stderr": "1.607", + "latency": "1.793", + "cost_per_test": "0.000151", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.607, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.555, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.555", + "raw_stderr": "1.341", + "latency": "2.534", + "cost_per_test": "0.00023", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.341, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 66.098, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.098", + "raw_stderr": "2.346", + "latency": "1.769", + "cost_per_test": "0.000191", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.346, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.934, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.934", + "raw_stderr": "1.513", + "latency": "1.376", + "cost_per_test": "0.000149", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.513, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 44.788, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "44.788", + "raw_stderr": "1.597", + "latency": "2.879", + "cost_per_test": "0.000227", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.597, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 62.714, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "62.714", + "raw_stderr": "1.691", + "latency": "1.396", + "cost_per_test": "0.000135", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.691, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 58.53, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "58.53", + "raw_stderr": "2.524", + "latency": "1.045", + "cost_per_test": "0.000199", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.524, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 42.598, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "42.598", + "raw_stderr": "1.49", + "latency": "1.323", + "cost_per_test": "0.000161", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.49, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.201, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.201", + "raw_stderr": "1.113", + "latency": "2.126", + "cost_per_test": "0.000223", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.113, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 62.338, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "62.338", + "raw_stderr": "1.594", + "latency": "1.045", + "cost_per_test": "0.000108", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.594, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 65.606, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "65.606", + "raw_stderr": "0.456", + "latency": "1.682", + "cost_per_test": "0.00017", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.456, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 56.513, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "56.513", + "raw_stderr": "2.219", + "latency": "1.282", + "cost_per_test": "0.000123", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.219, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.209, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.209", + "raw_stderr": "1.256", + "latency": "2.297", + "cost_per_test": "0.000188", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.256, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.687, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.687", + "raw_stderr": "1.539", + "latency": "1.181", + "cost_per_test": "0.000126", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.539, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-1.5-flash-002/2a49cefc-20b8-4944-b2aa-890eedd98aee.json b/data/vals_ai/google/gemini-1.5-flash-002/2a49cefc-20b8-4944-b2aa-890eedd98aee.json new file mode 100644 index 0000000000000000000000000000000000000000..4ffd48790b5ffd2d7166455376c4ec7fbf2cc901 --- /dev/null +++ b/data/vals_ai/google/gemini-1.5-flash-002/2a49cefc-20b8-4944-b2aa-890eedd98aee.json @@ -0,0 +1,408 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/google_gemini-1.5-flash-002/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-1.5-flash-002", + "id": "google/gemini-1.5-flash-002", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-1.5-flash-002", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.095, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.095", + "raw_stderr": "1.028", + "latency": "0.36", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.028, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 61.847, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "61.847", + "raw_stderr": "0.208", + "latency": "0.404", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.208, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 51.973, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "51.973", + "raw_stderr": "0.263", + "latency": "0.445", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.263, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 63.238, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "63.238", + "raw_stderr": "0.736", + "latency": "0.476", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.736, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 61.657, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "61.657", + "raw_stderr": "0.667", + "latency": "0.397", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.667, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 62.618, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "62.618", + "raw_stderr": "3.456", + "latency": "0.772", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.456, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-1.5-flash-002/364d18c0-90ca-4190-8cc0-7c2ac2675ac2.json b/data/vals_ai/google/gemini-1.5-flash-002/364d18c0-90ca-4190-8cc0-7c2ac2675ac2.json new file mode 100644 index 0000000000000000000000000000000000000000..66a199734fd771b6dcae59976baeead3d22acd3b --- /dev/null +++ b/data/vals_ai/google/gemini-1.5-flash-002/364d18c0-90ca-4190-8cc0-7c2ac2675ac2.json @@ -0,0 +1,288 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/google_gemini-1.5-flash-002/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-1.5-flash-002", + "id": "google/gemini-1.5-flash-002", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-1.5-flash-002", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 82.609, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "82.609", + "raw_stderr": "2.112", + "latency": "1.061", + "cost_per_test": "7.3e-05", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.112, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 5.143, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "5.143", + "raw_stderr": "1.181", + "latency": "3.478", + "cost_per_test": "0.0002", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.181, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 22.977, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "22.977", + "raw_stderr": "2.15", + "latency": "2.997", + "cost_per_test": "0.000168", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.15, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 36.91, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "36.91", + "raw_stderr": "1.079", + "latency": "2.512", + "cost_per_test": "0.000147", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.079, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-1.5-flash-002/37018c95-c68e-4092-8d16-c720889533c0.json b/data/vals_ai/google/gemini-1.5-flash-002/37018c95-c68e-4092-8d16-c720889533c0.json new file mode 100644 index 0000000000000000000000000000000000000000..706b5487698d0cf30ebaa851624c7a609e3d396f --- /dev/null +++ b/data/vals_ai/google/gemini-1.5-flash-002/37018c95-c68e-4092-8d16-c720889533c0.json @@ -0,0 +1,276 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/google_gemini-1.5-flash-002/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-1.5-flash-002", + "id": "google/gemini-1.5-flash-002", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-1.5-flash-002", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 37.762, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "37.762", + "raw_stderr": "1.655", + "latency": "13.387", + "cost_per_test": "0.004915", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.655, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 38.928, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "38.928", + "raw_stderr": "1.665", + "latency": "58.446", + "cost_per_test": "0.013757", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.665, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 38.19, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "38.19", + "raw_stderr": "0.958", + "latency": "28.407", + "cost_per_test": "0.007862", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.958, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 37.879, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "37.879", + "raw_stderr": "1.656", + "latency": "13.387", + "cost_per_test": "0.004915", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.656, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-1.5-flash-002/3a8b5faa-7633-4128-8631-6cfba64dca88.json b/data/vals_ai/google/gemini-1.5-flash-002/3a8b5faa-7633-4128-8631-6cfba64dca88.json new file mode 100644 index 0000000000000000000000000000000000000000..bcc62ab27bd4caec4b4d0513cb8a36eb3e120034 --- /dev/null +++ b/data/vals_ai/google/gemini-1.5-flash-002/3a8b5faa-7633-4128-8631-6cfba64dca88.json @@ -0,0 +1,216 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/google_gemini-1.5-flash-002/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-1.5-flash-002", + "id": "google/gemini-1.5-flash-002", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-1.5-flash-002", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 58.108, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.108", + "raw_stderr": "2.722", + "latency": "2.002", + "cost_per_test": "0.00011", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.722, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 42.766, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "42.766", + "raw_stderr": "1.835", + "latency": "2.1", + "cost_per_test": "0.00012", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.835, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 27.424, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "27.424", + "raw_stderr": "2.463", + "latency": "2.197", + "cost_per_test": "0.000129", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.463, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-1.5-flash-002/5bb551fe-b071-4175-92ea-c4bce9407ca0.json b/data/vals_ai/google/gemini-1.5-flash-002/5bb551fe-b071-4175-92ea-c4bce9407ca0.json new file mode 100644 index 0000000000000000000000000000000000000000..eaf94ce09828c369ce5f35ea25d286073ce61b9f --- /dev/null +++ b/data/vals_ai/google/gemini-1.5-flash-002/5bb551fe-b071-4175-92ea-c4bce9407ca0.json @@ -0,0 +1,216 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/google_gemini-1.5-flash-002/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-1.5-flash-002", + "id": "google/gemini-1.5-flash-002", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-1.5-flash-002", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 36.141, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "36.141", + "raw_stderr": "1.374", + "latency": "3.701", + "cost_per_test": "0.000155", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.374, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 48.202, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "48.202", + "raw_stderr": "0.98", + "latency": "3.701", + "cost_per_test": "0.000155", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.98, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 60.262, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.262", + "raw_stderr": "1.399", + "latency": "3.701", + "cost_per_test": "0.000155", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.399, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-1.5-flash-002/b2c3dd01-693f-4e62-afca-5f342da87ec3.json b/data/vals_ai/google/gemini-1.5-flash-002/b2c3dd01-693f-4e62-afca-5f342da87ec3.json new file mode 100644 index 0000000000000000000000000000000000000000..f551e8307f0e71ce4f9cea033ebc07183b2f5a5a --- /dev/null +++ b/data/vals_ai/google/gemini-1.5-flash-002/b2c3dd01-693f-4e62-afca-5f342da87ec3.json @@ -0,0 +1,225 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/google_gemini-1.5-flash-002/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-1.5-flash-002", + "id": "google/gemini-1.5-flash-002", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-1.5-flash-002", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 44.949, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "44.949", + "raw_stderr": "3.535", + "latency": "3.47", + "cost_per_test": "0.000299", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 46.97, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "46.97", + "raw_stderr": "3.547", + "latency": "3.311", + "cost_per_test": "0.000167", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.547, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 45.96, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "45.96", + "raw_stderr": "2.504", + "latency": "3.39", + "cost_per_test": "0.000233", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.504, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-1.5-flash-002/beeb68ef-e735-460f-aa40-2f893e03f5f3.json b/data/vals_ai/google/gemini-1.5-flash-002/beeb68ef-e735-460f-aa40-2f893e03f5f3.json new file mode 100644 index 0000000000000000000000000000000000000000..da066fbe65fd7ad17eadbf6f2f5bd90e8bc8849d --- /dev/null +++ b/data/vals_ai/google/gemini-1.5-flash-002/beeb68ef-e735-460f-aa40-2f893e03f5f3.json @@ -0,0 +1,99 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/google_gemini-1.5-flash-002/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-1.5-flash-002", + "id": "google/gemini-1.5-flash-002", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-1.5-flash-002", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:google/gemini-1.5-flash-002:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 57.192, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "57.192", + "raw_stderr": "1.189", + "latency": "41.314", + "cost_per_test": "0.000114", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.189, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-1.5-pro-002/2642887a-f7e7-4ec4-b678-fa52b90ac50a.json b/data/vals_ai/google/gemini-1.5-pro-002/2642887a-f7e7-4ec4-b678-fa52b90ac50a.json new file mode 100644 index 0000000000000000000000000000000000000000..d9a3c6fae77160b3a747dbb546b03ef995340443 --- /dev/null +++ b/data/vals_ai/google/gemini-1.5-pro-002/2642887a-f7e7-4ec4-b678-fa52b90ac50a.json @@ -0,0 +1,981 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/google_gemini-1.5-pro-002/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-1.5-pro-002", + "id": "google/gemini-1.5-pro-002", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-1.5-pro-002", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.145, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.145", + "raw_stderr": "1.207", + "latency": "2.688", + "cost_per_test": "0.0024", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.207, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.355, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.355", + "raw_stderr": "1.414", + "latency": "3.559", + "cost_per_test": "0.002419", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.414, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.975, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.975", + "raw_stderr": "1.211", + "latency": "5.361", + "cost_per_test": "0.003754", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.211, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.049, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.049", + "raw_stderr": "2.044", + "latency": "3.692", + "cost_per_test": "0.003045", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.044, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.687, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.687", + "raw_stderr": "1.359", + "latency": "2.351", + "cost_per_test": "0.002155", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.359, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 59.752, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "59.752", + "raw_stderr": "1.575", + "latency": "7.355", + "cost_per_test": "0.004022", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.575, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.594, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.594", + "raw_stderr": "1.544", + "latency": "2.18", + "cost_per_test": "0.001843", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.544, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 67.192, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "67.192", + "raw_stderr": "2.405", + "latency": "1.967", + "cost_per_test": "0.003125", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.405, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 54.223, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "54.223", + "raw_stderr": "1.501", + "latency": "2.483", + "cost_per_test": "0.002383", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.501, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.936, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.936", + "raw_stderr": "0.946", + "latency": "4.791", + "cost_per_test": "0.003813", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.946, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.0, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75", + "raw_stderr": "1.425", + "latency": "1.857", + "cost_per_test": "0.0016", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.425, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.294, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.294", + "raw_stderr": "0.417", + "latency": "3.343", + "cost_per_test": "0.002649", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.417, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 70.341, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.341", + "raw_stderr": "2.045", + "latency": "2.096", + "cost_per_test": "0.001742", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.045, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.293, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.293", + "raw_stderr": "1.084", + "latency": "4.62", + "cost_per_test": "0.003023", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.084, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.576, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.576", + "raw_stderr": "1.4", + "latency": "1.803", + "cost_per_test": "0.001761", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.4, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-1.5-pro-002/28a7a1b4-b2dd-4432-880c-f34c405afaee.json b/data/vals_ai/google/gemini-1.5-pro-002/28a7a1b4-b2dd-4432-880c-f34c405afaee.json new file mode 100644 index 0000000000000000000000000000000000000000..53449421445b0d9a36e2e3eac3ec6331b8b06d8d --- /dev/null +++ b/data/vals_ai/google/gemini-1.5-pro-002/28a7a1b4-b2dd-4432-880c-f34c405afaee.json @@ -0,0 +1,225 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/google_gemini-1.5-pro-002/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-1.5-pro-002", + "id": "google/gemini-1.5-pro-002", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-1.5-pro-002", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 56.061, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "56.061", + "raw_stderr": "3.527", + "latency": "7.712", + "cost_per_test": "0.004704", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.527, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 60.606, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "60.606", + "raw_stderr": "3.472", + "latency": "7.706", + "cost_per_test": "0.002606", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.472, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 58.334, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "58.334", + "raw_stderr": "2.475", + "latency": "7.709", + "cost_per_test": "0.003655", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.475, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-1.5-pro-002/2d8a42d5-5258-406e-b0cf-5c19b1a16825.json b/data/vals_ai/google/gemini-1.5-pro-002/2d8a42d5-5258-406e-b0cf-5c19b1a16825.json new file mode 100644 index 0000000000000000000000000000000000000000..ee9dd4018e1cdffbca834ee2d110cacc44085ad2 --- /dev/null +++ b/data/vals_ai/google/gemini-1.5-pro-002/2d8a42d5-5258-406e-b0cf-5c19b1a16825.json @@ -0,0 +1,288 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/google_gemini-1.5-pro-002/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-1.5-pro-002", + "id": "google/gemini-1.5-pro-002", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-1.5-pro-002", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 88.199, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "88.199", + "raw_stderr": "1.798", + "latency": "2.197", + "cost_per_test": "0.001189", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.798, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 7.714, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "7.714", + "raw_stderr": "1.426", + "latency": "5.667", + "cost_per_test": "0.002659", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.426, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 29.243, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "29.243", + "raw_stderr": "2.324", + "latency": "3.998", + "cost_per_test": "0.002077", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.324, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 41.719, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "41.719", + "raw_stderr": "1.089", + "latency": "3.954", + "cost_per_test": "0.001975", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.089, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-1.5-pro-002/378bf142-e7f2-4d3c-9084-edc4ed469412.json b/data/vals_ai/google/gemini-1.5-pro-002/378bf142-e7f2-4d3c-9084-edc4ed469412.json new file mode 100644 index 0000000000000000000000000000000000000000..31c5521f8f6843601dea1ff7e0a2f1bbb1eba79e --- /dev/null +++ b/data/vals_ai/google/gemini-1.5-pro-002/378bf142-e7f2-4d3c-9084-edc4ed469412.json @@ -0,0 +1,408 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/google_gemini-1.5-pro-002/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-1.5-pro-002", + "id": "google/gemini-1.5-pro-002", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-1.5-pro-002", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.32, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.32", + "raw_stderr": "0.88", + "latency": "0.552", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.88, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 72.377, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "72.377", + "raw_stderr": "0.217", + "latency": "0.57", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.217, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 64.126, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "64.126", + "raw_stderr": "0.589", + "latency": "0.673", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.589, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 69.08, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "69.08", + "raw_stderr": "0.751", + "latency": "0.898", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.751, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 65.108, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "65.108", + "raw_stderr": "0.666", + "latency": "0.599", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.666, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 61.468, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "61.468", + "raw_stderr": "3.533", + "latency": "2.096", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.533, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-1.5-pro-002/450208f3-1363-49a0-9b50-c9f7ced29a16.json b/data/vals_ai/google/gemini-1.5-pro-002/450208f3-1363-49a0-9b50-c9f7ced29a16.json new file mode 100644 index 0000000000000000000000000000000000000000..6968beff369bcec68f14cbb47ebe304cd8325a50 --- /dev/null +++ b/data/vals_ai/google/gemini-1.5-pro-002/450208f3-1363-49a0-9b50-c9f7ced29a16.json @@ -0,0 +1,99 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/google_gemini-1.5-pro-002/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-1.5-pro-002", + "id": "google/gemini-1.5-pro-002", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-1.5-pro-002", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 65.511, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "65.511", + "raw_stderr": "1.142", + "latency": "43.736", + "cost_per_test": "0.001846", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.142, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-1.5-pro-002/6c532619-1257-45c9-b733-d24cb70e9096.json b/data/vals_ai/google/gemini-1.5-pro-002/6c532619-1257-45c9-b733-d24cb70e9096.json new file mode 100644 index 0000000000000000000000000000000000000000..2cc1dfc00dc3bc4bbec97a3e7a30951b32525ad3 --- /dev/null +++ b/data/vals_ai/google/gemini-1.5-pro-002/6c532619-1257-45c9-b733-d24cb70e9096.json @@ -0,0 +1,216 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/google_gemini-1.5-pro-002/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-1.5-pro-002", + "id": "google/gemini-1.5-pro-002", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-1.5-pro-002", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 47.343, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "47.343", + "raw_stderr": "1.428", + "latency": "9.86", + "cost_per_test": "0.002545", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.428, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 59.485, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.485", + "raw_stderr": "0.962", + "latency": "9.86", + "cost_per_test": "0.002545", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.962, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 71.627, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.627", + "raw_stderr": "1.289", + "latency": "9.86", + "cost_per_test": "0.002545", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.289, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-1.5-pro-002/9bb5e7b0-f262-4f0b-ab98-57660740d349.json b/data/vals_ai/google/gemini-1.5-pro-002/9bb5e7b0-f262-4f0b-ab98-57660740d349.json new file mode 100644 index 0000000000000000000000000000000000000000..446f5f1076f34f5aab287239889f4cfbeb3f0b83 --- /dev/null +++ b/data/vals_ai/google/gemini-1.5-pro-002/9bb5e7b0-f262-4f0b-ab98-57660740d349.json @@ -0,0 +1,276 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/google_gemini-1.5-pro-002/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-1.5-pro-002", + "id": "google/gemini-1.5-pro-002", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-1.5-pro-002", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 40.676, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "40.676", + "raw_stderr": "1.677", + "latency": "20.551", + "cost_per_test": "0.082015", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.677, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 40.443, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "40.443", + "raw_stderr": "1.675", + "latency": "73.047", + "cost_per_test": "0.229417", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.675, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 40.521, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "40.521", + "raw_stderr": "0.967", + "latency": "38.05", + "cost_per_test": "0.131149", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.967, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 40.443, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "40.443", + "raw_stderr": "1.675", + "latency": "20.551", + "cost_per_test": "0.082015", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.675, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-1.5-pro-002/d232c50c-de4e-46d5-9277-e7a8c01fe04c.json b/data/vals_ai/google/gemini-1.5-pro-002/d232c50c-de4e-46d5-9277-e7a8c01fe04c.json new file mode 100644 index 0000000000000000000000000000000000000000..4081df356fcdae2bc0f22fe3f3a47dabc7f867f0 --- /dev/null +++ b/data/vals_ai/google/gemini-1.5-pro-002/d232c50c-de4e-46d5-9277-e7a8c01fe04c.json @@ -0,0 +1,216 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/google_gemini-1.5-pro-002/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-1.5-pro-002", + "id": "google/gemini-1.5-pro-002", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-1.5-pro-002", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 63.434, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.434", + "raw_stderr": "2.658", + "latency": "3.989", + "cost_per_test": "0.001918", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.658, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 56.756, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.756", + "raw_stderr": "1.916", + "latency": "4.205", + "cost_per_test": "0.002028", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.916, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:google/gemini-1.5-pro-002:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 50.079, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "50.079", + "raw_stderr": "2.759", + "latency": "4.421", + "cost_per_test": "0.002137", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.759, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.0-flash-001/189e5cb3-5090-477f-9a88-b262b01701b0.json b/data/vals_ai/google/gemini-2.0-flash-001/189e5cb3-5090-477f-9a88-b262b01701b0.json new file mode 100644 index 0000000000000000000000000000000000000000..a85be11220edd297933903f721c5500097ad51b4 --- /dev/null +++ b/data/vals_ai/google/gemini-2.0-flash-001/189e5cb3-5090-477f-9a88-b262b01701b0.json @@ -0,0 +1,216 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/google_gemini-2.0-flash-001/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.0-flash-001", + "id": "google/gemini-2.0-flash-001", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.0-flash-001", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 62.957, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.957", + "raw_stderr": "2.665", + "latency": "2.918", + "cost_per_test": "0.000265", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.665, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 59.658, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.658", + "raw_stderr": "1.91", + "latency": "3.794", + "cost_per_test": "0.00027", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.91, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 56.359, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.359", + "raw_stderr": "2.736", + "latency": "4.671", + "cost_per_test": "0.000274", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.736, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.0-flash-001/4e953d0d-1909-4fb2-92c5-970ffd362b11.json b/data/vals_ai/google/gemini-2.0-flash-001/4e953d0d-1909-4fb2-92c5-970ffd362b11.json new file mode 100644 index 0000000000000000000000000000000000000000..62605ca3cbc36fa09bcc8cdf4e35071c917d1cfb --- /dev/null +++ b/data/vals_ai/google/gemini-2.0-flash-001/4e953d0d-1909-4fb2-92c5-970ffd362b11.json @@ -0,0 +1,408 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/google_gemini-2.0-flash-001/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.0-flash-001", + "id": "google/gemini-2.0-flash-001", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.0-flash-001", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.388, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.388", + "raw_stderr": "0.956", + "latency": "0.316", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.956, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.938, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.938", + "raw_stderr": "0.203", + "latency": "0.322", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.203, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.905, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.905", + "raw_stderr": "0.585", + "latency": "0.334", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.585, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.364, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.364", + "raw_stderr": "0.782", + "latency": "0.447", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.782, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.211, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.211", + "raw_stderr": "1.04", + "latency": "0.324", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.04, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 64.379, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "64.379", + "raw_stderr": "3.591", + "latency": "0.939", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.591, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.0-flash-001/80b3891a-1cfa-4573-a989-d247bc1cbd84.json b/data/vals_ai/google/gemini-2.0-flash-001/80b3891a-1cfa-4573-a989-d247bc1cbd84.json new file mode 100644 index 0000000000000000000000000000000000000000..8b1f71468d309f0a5db74c92650a88d4bd797b10 --- /dev/null +++ b/data/vals_ai/google/gemini-2.0-flash-001/80b3891a-1cfa-4573-a989-d247bc1cbd84.json @@ -0,0 +1,276 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/google_gemini-2.0-flash-001/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.0-flash-001", + "id": "google/gemini-2.0-flash-001", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.0-flash-001", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 35.431, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "35.431", + "raw_stderr": "1.633", + "latency": "18.434", + "cost_per_test": "0.006552", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.633, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 30.07, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "30.07", + "raw_stderr": "1.566", + "latency": "53.247", + "cost_per_test": "0.018339", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.566, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 33.722, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "33.722", + "raw_stderr": "0.93", + "latency": "30.038", + "cost_per_test": "0.010481", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.93, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 35.664, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "35.664", + "raw_stderr": "1.635", + "latency": "18.434", + "cost_per_test": "0.006552", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.635, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.0-flash-001/8d3f3547-69dd-41f8-a365-f9dce519b4af.json b/data/vals_ai/google/gemini-2.0-flash-001/8d3f3547-69dd-41f8-a365-f9dce519b4af.json new file mode 100644 index 0000000000000000000000000000000000000000..eae5b2d7770cc340516e29198c6d8598ae70e646 --- /dev/null +++ b/data/vals_ai/google/gemini-2.0-flash-001/8d3f3547-69dd-41f8-a365-f9dce519b4af.json @@ -0,0 +1,99 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/google_gemini-2.0-flash-001/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.0-flash-001", + "id": "google/gemini-2.0-flash-001", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.0-flash-001", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 69.786, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.786", + "raw_stderr": "1.104", + "latency": "45.203", + "cost_per_test": "0.000365", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.104, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.0-flash-001/aeecf792-f28c-44c9-ba3d-88afccda6e4d.json b/data/vals_ai/google/gemini-2.0-flash-001/aeecf792-f28c-44c9-ba3d-88afccda6e4d.json new file mode 100644 index 0000000000000000000000000000000000000000..e21dc9351cbecd4ae64690cfa62af334e0e4c87e --- /dev/null +++ b/data/vals_ai/google/gemini-2.0-flash-001/aeecf792-f28c-44c9-ba3d-88afccda6e4d.json @@ -0,0 +1,216 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/google_gemini-2.0-flash-001/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.0-flash-001", + "id": "google/gemini-2.0-flash-001", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.0-flash-001", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 53.475, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "53.475", + "raw_stderr": "1.426", + "latency": "5.702", + "cost_per_test": "0.000296", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.426, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 65.25, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.25", + "raw_stderr": "0.933", + "latency": "5.702", + "cost_per_test": "0.000296", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.933, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 77.024, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "77.024", + "raw_stderr": "1.203", + "latency": "5.702", + "cost_per_test": "0.000296", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.203, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.0-flash-001/b2a26e06-295a-41f7-9fd5-d63b15174479.json b/data/vals_ai/google/gemini-2.0-flash-001/b2a26e06-295a-41f7-9fd5-d63b15174479.json new file mode 100644 index 0000000000000000000000000000000000000000..94ee5a70686fa0a3aa0403de00ce04ece6ffbe45 --- /dev/null +++ b/data/vals_ai/google/gemini-2.0-flash-001/b2a26e06-295a-41f7-9fd5-d63b15174479.json @@ -0,0 +1,981 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/google_gemini-2.0-flash-001/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.0-flash-001", + "id": "google/gemini-2.0-flash-001", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.0-flash-001", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.703, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.703", + "raw_stderr": "1.182", + "latency": "2.946", + "cost_per_test": "0.000309", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.182, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.876, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.876", + "raw_stderr": "1.371", + "latency": "4.674", + "cost_per_test": "0.000434", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.371, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.127, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.127", + "raw_stderr": "1.113", + "latency": "6.964", + "cost_per_test": "0.000652", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.113, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.488, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.488", + "raw_stderr": "1.957", + "latency": "6.226", + "cost_per_test": "0.000593", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.957, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.649, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.649", + "raw_stderr": "1.273", + "latency": "2.017", + "cost_per_test": "0.000231", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.273, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 67.699, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "67.699", + "raw_stderr": "1.502", + "latency": "9.302", + "cost_per_test": "0.000771", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.502, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.917, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.917", + "raw_stderr": "1.495", + "latency": "2.117", + "cost_per_test": "0.000209", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.495, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 69.029, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.029", + "raw_stderr": "2.369", + "latency": "1.28", + "cost_per_test": "0.000268", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.369, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 60.672, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "60.672", + "raw_stderr": "1.472", + "latency": "5.07", + "cost_per_test": "0.000399", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.472, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.755, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.755", + "raw_stderr": "1.073", + "latency": "8.408", + "cost_per_test": "0.000825", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.073, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.325, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.325", + "raw_stderr": "1.418", + "latency": "1.702", + "cost_per_test": "0.000172", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.418, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.376, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.376", + "raw_stderr": "0.407", + "latency": "4.318", + "cost_per_test": "0.000414", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.407, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.543, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.543", + "raw_stderr": "2.02", + "latency": "2.256", + "cost_per_test": "0.00022", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.02, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.911, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.911", + "raw_stderr": "1.019", + "latency": "5.669", + "cost_per_test": "0.000517", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.019, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.576, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.576", + "raw_stderr": "1.404", + "latency": "1.826", + "cost_per_test": "0.000194", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.404, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.0-flash-001/cde01f1c-96e5-41e4-a1ab-e3c4f17e458a.json b/data/vals_ai/google/gemini-2.0-flash-001/cde01f1c-96e5-41e4-a1ab-e3c4f17e458a.json new file mode 100644 index 0000000000000000000000000000000000000000..a9973702773bef755e63e672929ec30483e6c5e7 --- /dev/null +++ b/data/vals_ai/google/gemini-2.0-flash-001/cde01f1c-96e5-41e4-a1ab-e3c4f17e458a.json @@ -0,0 +1,288 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/google_gemini-2.0-flash-001/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.0-flash-001", + "id": "google/gemini-2.0-flash-001", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.0-flash-001", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 90.683, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "90.683", + "raw_stderr": "1.62", + "latency": "1.313", + "cost_per_test": "0.000123", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.62, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 8.286, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "8.286", + "raw_stderr": "1.473", + "latency": "5.372", + "cost_per_test": "0.000482", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.473, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 31.854, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "31.854", + "raw_stderr": "2.381", + "latency": "3.388", + "cost_per_test": "0.000308", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.381, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 43.608, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "43.608", + "raw_stderr": "1.078", + "latency": "3.358", + "cost_per_test": "0.000304", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.078, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.0-flash-001/efa56bdb-4b9e-4773-bded-2a631c5b7a62.json b/data/vals_ai/google/gemini-2.0-flash-001/efa56bdb-4b9e-4773-bded-2a631c5b7a62.json new file mode 100644 index 0000000000000000000000000000000000000000..39fadd76f833be63954c820f380044298054544a --- /dev/null +++ b/data/vals_ai/google/gemini-2.0-flash-001/efa56bdb-4b9e-4773-bded-2a631c5b7a62.json @@ -0,0 +1,225 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/google_gemini-2.0-flash-001/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.0-flash-001", + "id": "google/gemini-2.0-flash-001", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.0-flash-001", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 66.162, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.162", + "raw_stderr": "3.363", + "latency": "4.976", + "cost_per_test": "0.000512", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.363, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 64.141, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "64.141", + "raw_stderr": "3.408", + "latency": "6.63", + "cost_per_test": "0.000469", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.408, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:google/gemini-2.0-flash-001:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 65.152, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "65.152", + "raw_stderr": "2.394", + "latency": "5.803", + "cost_per_test": "0.00049", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.394, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.0-flash-exp/aa5a942d-30ad-49c1-9a36-58457ca46726.json b/data/vals_ai/google/gemini-2.0-flash-exp/aa5a942d-30ad-49c1-9a36-58457ca46726.json new file mode 100644 index 0000000000000000000000000000000000000000..aea44fd115b48c56f659fc323f5255b298cd4c95 --- /dev/null +++ b/data/vals_ai/google/gemini-2.0-flash-exp/aa5a942d-30ad-49c1-9a36-58457ca46726.json @@ -0,0 +1,216 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/google_gemini-2.0-flash-exp/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.0-flash-exp", + "id": "google/gemini-2.0-flash-exp", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.0-flash-exp", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:google/gemini-2.0-flash-exp:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 56.092, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.092", + "raw_stderr": "1.419", + "latency": "7.486", + "cost_per_test": "0.000302", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.419, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:google/gemini-2.0-flash-exp:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 67.744, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.744", + "raw_stderr": "0.915", + "latency": "7.486", + "cost_per_test": "0.000302", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.915, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:google/gemini-2.0-flash-exp:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 79.395, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "79.395", + "raw_stderr": "1.157", + "latency": "7.486", + "cost_per_test": "0.000302", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.157, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.0-flash-exp/d0bc672a-6be4-48bb-9ed8-b0383cbbd1cf.json b/data/vals_ai/google/gemini-2.0-flash-exp/d0bc672a-6be4-48bb-9ed8-b0383cbbd1cf.json new file mode 100644 index 0000000000000000000000000000000000000000..5faead53b72413687d02cfafd94e63dab5aa9e22 --- /dev/null +++ b/data/vals_ai/google/gemini-2.0-flash-exp/d0bc672a-6be4-48bb-9ed8-b0383cbbd1cf.json @@ -0,0 +1,216 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/google_gemini-2.0-flash-exp/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.0-flash-exp", + "id": "google/gemini-2.0-flash-exp", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.0-flash-exp", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:google/gemini-2.0-flash-exp:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 64.229, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.229", + "raw_stderr": "2.645", + "latency": "2.911", + "cost_per_test": "0.000135", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.645, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:google/gemini-2.0-flash-exp:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 60.374, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.374", + "raw_stderr": "1.902", + "latency": "2.796", + "cost_per_test": "0.000132", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.902, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:google/gemini-2.0-flash-exp:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 56.518, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.518", + "raw_stderr": "2.735", + "latency": "2.681", + "cost_per_test": "0.00013", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.735, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.0-flash-thinking-exp-01-21/dd1ebf5a-ba52-4ac5-8e30-9a30570b7cfd.json b/data/vals_ai/google/gemini-2.0-flash-thinking-exp-01-21/dd1ebf5a-ba52-4ac5-8e30-9a30570b7cfd.json new file mode 100644 index 0000000000000000000000000000000000000000..c6f82891fbb81cd65a8beb103758f904dcdc8cd5 --- /dev/null +++ b/data/vals_ai/google/gemini-2.0-flash-thinking-exp-01-21/dd1ebf5a-ba52-4ac5-8e30-9a30570b7cfd.json @@ -0,0 +1,216 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/google_gemini-2.0-flash-thinking-exp-01-21/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.0-flash-thinking-exp-01-21", + "id": "google/gemini-2.0-flash-thinking-exp-01-21", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.0-flash-thinking-exp-01-21", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:google/gemini-2.0-flash-thinking-exp-01-21:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 58.299, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.299", + "raw_stderr": "1.41", + "latency": "12.807", + "cost_per_test": "0.000827", + "max_output_tokens": "16384", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.41, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:google/gemini-2.0-flash-thinking-exp-01-21:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 69.788, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "69.788", + "raw_stderr": "0.899", + "latency": "12.807", + "cost_per_test": "0.000827", + "max_output_tokens": "16384", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.899, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:google/gemini-2.0-flash-thinking-exp-01-21:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 81.276, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "81.276", + "raw_stderr": "1.116", + "latency": "12.807", + "cost_per_test": "0.000827", + "max_output_tokens": "16384", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.116, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.0-pro-exp-02-05/a83bd4f0-64db-45f5-89fe-ad045a701609.json b/data/vals_ai/google/gemini-2.0-pro-exp-02-05/a83bd4f0-64db-45f5-89fe-ad045a701609.json new file mode 100644 index 0000000000000000000000000000000000000000..88e6ecba20e8e02bca00acd30034cdc4d84dbde5 --- /dev/null +++ b/data/vals_ai/google/gemini-2.0-pro-exp-02-05/a83bd4f0-64db-45f5-89fe-ad045a701609.json @@ -0,0 +1,216 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/google_gemini-2.0-pro-exp-02-05/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.0-pro-exp-02-05", + "id": "google/gemini-2.0-pro-exp-02-05", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.0-pro-exp-02-05", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:google/gemini-2.0-pro-exp-02-05:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 57.318, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.318", + "raw_stderr": "1.414", + "latency": "8.989", + "cost_per_test": "0.004012", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.414, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:google/gemini-2.0-pro-exp-02-05:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 68.152, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.152", + "raw_stderr": "0.916", + "latency": "8.989", + "cost_per_test": "0.004012", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.916, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:google/gemini-2.0-pro-exp-02-05:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 78.986, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "78.986", + "raw_stderr": "1.165", + "latency": "8.989", + "cost_per_test": "0.004012", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.165, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.0-pro-exp-02-05/c58adcb9-9d9b-4142-8890-3a78d4f9c6a2.json b/data/vals_ai/google/gemini-2.0-pro-exp-02-05/c58adcb9-9d9b-4142-8890-3a78d4f9c6a2.json new file mode 100644 index 0000000000000000000000000000000000000000..e12f294a67c2f4798a36b464d63808bdc3a7b285 --- /dev/null +++ b/data/vals_ai/google/gemini-2.0-pro-exp-02-05/c58adcb9-9d9b-4142-8890-3a78d4f9c6a2.json @@ -0,0 +1,276 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/google_gemini-2.0-pro-exp-02-05/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.0-pro-exp-02-05", + "id": "google/gemini-2.0-pro-exp-02-05", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.0-pro-exp-02-05", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:google/gemini-2.0-pro-exp-02-05:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 51.166, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "51.166", + "raw_stderr": "1.707", + "latency": "5.275", + "cost_per_test": "0.002429", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.707, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:google/gemini-2.0-pro-exp-02-05:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 38.695, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "38.695", + "raw_stderr": "1.663", + "latency": "35.856", + "cost_per_test": "0.229335", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.663, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:google/gemini-2.0-pro-exp-02-05:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 43.435, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "43.435", + "raw_stderr": "0.971", + "latency": "18.879", + "cost_per_test": "0.104585", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.971, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:google/gemini-2.0-pro-exp-02-05:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 40.443, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "40.443", + "raw_stderr": "1.675", + "latency": "15.506", + "cost_per_test": "0.081991", + "max_output_tokens": "8192", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.675, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025-thinking/1160ce09-f3f6-4d0a-af9a-4bdcdd73e2e1.json b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025-thinking/1160ce09-f3f6-4d0a-af9a-4bdcdd73e2e1.json new file mode 100644 index 0000000000000000000000000000000000000000..3280e94035abd8b35b95b08620b51e79a6d7671b --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025-thinking/1160ce09-f3f6-4d0a-af9a-4bdcdd73e2e1.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/google_gemini-2.5-flash-lite-preview-09-2025-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-lite-preview-09-2025-thinking", + "id": "google/gemini-2.5-flash-lite-preview-09-2025-thinking", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-lite-preview-09-2025-thinking", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 52.003, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.003", + "raw_stderr": "1.429", + "latency": "3.94", + "cost_per_test": "0.000509", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.429, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 64.718, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.718", + "raw_stderr": "0.931", + "latency": "3.94", + "cost_per_test": "0.000509", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.931, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 77.433, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "77.433", + "raw_stderr": "1.195", + "latency": "3.94", + "cost_per_test": "0.000509", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.195, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025-thinking/3e7e98b6-bbb5-41bb-9a7f-70f2e826f08a.json b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025-thinking/3e7e98b6-bbb5-41bb-9a7f-70f2e826f08a.json new file mode 100644 index 0000000000000000000000000000000000000000..9f6454c98582183d89e0df51a103b4ae18df471e --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025-thinking/3e7e98b6-bbb5-41bb-9a7f-70f2e826f08a.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/google_gemini-2.5-flash-lite-preview-09-2025-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-lite-preview-09-2025-thinking", + "id": "google/gemini-2.5-flash-lite-preview-09-2025-thinking", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-lite-preview-09-2025-thinking", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 58.506, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.506", + "raw_stderr": "1.389", + "latency": "31.147", + "cost_per_test": "0.000285", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.389, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 57.552, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.552", + "raw_stderr": "0.985", + "latency": "28.206", + "cost_per_test": "0.000286", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.985, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 56.598, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.598", + "raw_stderr": "1.397", + "latency": "25.266", + "cost_per_test": "0.000286", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.397, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025-thinking/40a4b6a7-7cf6-4cf0-813f-a3911a8493aa.json b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025-thinking/40a4b6a7-7cf6-4cf0-813f-a3911a8493aa.json new file mode 100644 index 0000000000000000000000000000000000000000..a3317dc2dc0f648d1379932877725231a343afac --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025-thinking/40a4b6a7-7cf6-4cf0-813f-a3911a8493aa.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/google_gemini-2.5-flash-lite-preview-09-2025-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-lite-preview-09-2025-thinking", + "id": "google/gemini-2.5-flash-lite-preview-09-2025-thinking", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-lite-preview-09-2025-thinking", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 66.877, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "66.877", + "raw_stderr": "1.923", + "latency": "11.518", + "cost_per_test": "0.002567", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.923, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025-thinking/4fe921c1-8e2a-4f0c-8bdb-c26bfaedcef7.json b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025-thinking/4fe921c1-8e2a-4f0c-8bdb-c26bfaedcef7.json new file mode 100644 index 0000000000000000000000000000000000000000..88e8329629b8493b446a26949ad821dc8bb45de3 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025-thinking/4fe921c1-8e2a-4f0c-8bdb-c26bfaedcef7.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/google_gemini-2.5-flash-lite-preview-09-2025-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-lite-preview-09-2025-thinking", + "id": "google/gemini-2.5-flash-lite-preview-09-2025-thinking", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-lite-preview-09-2025-thinking", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 53.846, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "53.846", + "raw_stderr": "1.702", + "latency": "57.443", + "cost_per_test": "0.000212", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.702, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.023, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.023", + "raw_stderr": "1.672", + "latency": "70.079", + "cost_per_test": "0.01883", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.672, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 57.576, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.576", + "raw_stderr": "0.973", + "latency": "54.225", + "cost_per_test": "0.008596", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.973, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 58.858, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.858", + "raw_stderr": "1.68", + "latency": "35.154", + "cost_per_test": "0.006747", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.68, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025-thinking/7a2f5de9-6d69-432a-b2c5-005068df1f84.json b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025-thinking/7a2f5de9-6d69-432a-b2c5-005068df1f84.json new file mode 100644 index 0000000000000000000000000000000000000000..d3820da668c7de63af00efc18338b68e995894eb --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025-thinking/7a2f5de9-6d69-432a-b2c5-005068df1f84.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/google_gemini-2.5-flash-lite-preview-09-2025-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-lite-preview-09-2025-thinking", + "id": "google/gemini-2.5-flash-lite-preview-09-2025-thinking", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-lite-preview-09-2025-thinking", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 86.297, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.297", + "raw_stderr": "0.905", + "latency": "3.707", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.905, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.552, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.552", + "raw_stderr": "0.209", + "latency": "3.05", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.209, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 85.416, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.416", + "raw_stderr": "0.613", + "latency": "2.956", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.613, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.044, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.044", + "raw_stderr": "0.438", + "latency": "3.139", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.438, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.104, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.104", + "raw_stderr": "0.84", + "latency": "2.768", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.84, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 76.852, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "76.852", + "raw_stderr": "1.686", + "latency": "3.212", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.686, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025-thinking/952011d6-c005-49aa-8708-4cec6f61b882.json b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025-thinking/952011d6-c005-49aa-8708-4cec6f61b882.json new file mode 100644 index 0000000000000000000000000000000000000000..4a663cea2bd874250c650882a7afa514b5295c9a --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025-thinking/952011d6-c005-49aa-8708-4cec6f61b882.json @@ -0,0 +1,101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/google_gemini-2.5-flash-lite-preview-09-2025-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-lite-preview-09-2025-thinking", + "id": "google/gemini-2.5-flash-lite-preview-09-2025-thinking", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-lite-preview-09-2025-thinking", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 75.434, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.434", + "raw_stderr": "1.035", + "latency": "25.728", + "cost_per_test": "0.000356", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.035, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025-thinking/98157a8d-b33c-4583-9583-de4282261bf7.json b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025-thinking/98157a8d-b33c-4583-9583-de4282261bf7.json new file mode 100644 index 0000000000000000000000000000000000000000..de2e3a75dfaae61253b83f4990d0ed30c450b3a9 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025-thinking/98157a8d-b33c-4583-9583-de4282261bf7.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/google_gemini-2.5-flash-lite-preview-09-2025-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-lite-preview-09-2025-thinking", + "id": "google/gemini-2.5-flash-lite-preview-09-2025-thinking", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-lite-preview-09-2025-thinking", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 69.697, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.697", + "raw_stderr": "3.266", + "latency": "21.398", + "cost_per_test": "0.000537", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.266, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 70.707, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.707", + "raw_stderr": "3.234", + "latency": "19.822", + "cost_per_test": "0.000425", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.234, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 70.202, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.202", + "raw_stderr": "2.298", + "latency": "20.61", + "cost_per_test": "0.000481", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.298, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025-thinking/b39613ec-da8f-46ba-bbf8-4dea6fa62eab.json b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025-thinking/b39613ec-da8f-46ba-bbf8-4dea6fa62eab.json new file mode 100644 index 0000000000000000000000000000000000000000..47a08ffc3ba9ed4a961637517b0f8c92f7f8f5d6 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025-thinking/b39613ec-da8f-46ba-bbf8-4dea6fa62eab.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/google_gemini-2.5-flash-lite-preview-09-2025-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-lite-preview-09-2025-thinking", + "id": "google/gemini-2.5-flash-lite-preview-09-2025-thinking", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-lite-preview-09-2025-thinking", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.137, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.137", + "raw_stderr": "0.754", + "latency": "10.915", + "cost_per_test": "0.000188", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.754, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 37.429, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "37.429", + "raw_stderr": "2.587", + "latency": "72.684", + "cost_per_test": "0.000652", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.587, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 78.59, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "78.59", + "raw_stderr": "2.096", + "latency": "32.492", + "cost_per_test": "0.000383", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.096, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 71.385, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "71.385", + "raw_stderr": "1.138", + "latency": "38.697", + "cost_per_test": "0.000408", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.138, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025-thinking/d198953c-7920-48f7-aaf0-4a74a8f57876.json b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025-thinking/d198953c-7920-48f7-aaf0-4a74a8f57876.json new file mode 100644 index 0000000000000000000000000000000000000000..09e65b3febde0c9dd0644069b61c2790f9b1524e --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025-thinking/d198953c-7920-48f7-aaf0-4a74a8f57876.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/google_gemini-2.5-flash-lite-preview-09-2025-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-lite-preview-09-2025-thinking", + "id": "google/gemini-2.5-flash-lite-preview-09-2025-thinking", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-lite-preview-09-2025-thinking", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.516, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.516", + "raw_stderr": "1.094", + "latency": "7.739", + "cost_per_test": "0.000294", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.094, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.411, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.411", + "raw_stderr": "1.291", + "latency": "8.627", + "cost_per_test": "0.000273", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.291, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.102, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.102", + "raw_stderr": "0.996", + "latency": "11.67", + "cost_per_test": "0.000436", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.996, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.951, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.951", + "raw_stderr": "1.899", + "latency": "9.315", + "cost_per_test": "0.000345", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.899, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.716, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.716", + "raw_stderr": "1.239", + "latency": "8.223", + "cost_per_test": "0.000285", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.239, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.414, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.414", + "raw_stderr": "1.451", + "latency": "20.77", + "cost_per_test": "0.00048", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.451, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.061, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.061", + "raw_stderr": "1.513", + "latency": "7.426", + "cost_per_test": "0.000241", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.513, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 66.404, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.404", + "raw_stderr": "2.42", + "latency": "8.115", + "cost_per_test": "0.000345", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.42, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 59.037, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "59.037", + "raw_stderr": "1.482", + "latency": "10.155", + "cost_per_test": "0.000305", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.482, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.118, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.118", + "raw_stderr": "0.774", + "latency": "9.519", + "cost_per_test": "0.000398", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.774, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.84, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.84", + "raw_stderr": "1.388", + "latency": "8.282", + "cost_per_test": "0.00019", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.388, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.118, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.118", + "raw_stderr": "0.397", + "latency": "9.618", + "cost_per_test": "0.000314", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.397, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.944, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.944", + "raw_stderr": "2.011", + "latency": "7.43", + "cost_per_test": "0.000237", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.011, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.066, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.066", + "raw_stderr": "0.961", + "latency": "10.632", + "cost_per_test": "0.000356", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.961, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.078, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.078", + "raw_stderr": "1.387", + "latency": "6.742", + "cost_per_test": "0.000205", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.387, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025-thinking/e919587b-b2d3-4883-8f5a-f37559ef8d44.json b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025-thinking/e919587b-b2d3-4883-8f5a-f37559ef8d44.json new file mode 100644 index 0000000000000000000000000000000000000000..8ea5a45a1fa99a1b521c43ef3d07a561534c0dff --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025-thinking/e919587b-b2d3-4883-8f5a-f37559ef8d44.json @@ -0,0 +1,346 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/google_gemini-2.5-flash-lite-preview-09-2025-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-lite-preview-09-2025-thinking", + "id": "google/gemini-2.5-flash-lite-preview-09-2025-thinking", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-lite-preview-09-2025-thinking", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 16.667, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "16.667", + "raw_stderr": "4.677", + "latency": "56.975", + "cost_per_test": "0.000959", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 4.677, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 30.952, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "30.952", + "raw_stderr": "5.896", + "latency": "22.4", + "cost_per_test": "0.000634", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.896, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 36.42, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "36.42", + "raw_stderr": "6.34", + "latency": "45.785", + "cost_per_test": "0.001696", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.34, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 44.708, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "44.708", + "raw_stderr": "5.585", + "latency": "19.434", + "cost_per_test": "0.001204", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.585, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:google/gemini-2.5-flash-lite-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 30.776, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "30.776", + "raw_stderr": "3.124", + "latency": "32.936", + "cost_per_test": "0.000932", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.124, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025/09d4a446-23db-475a-bf6e-70625e23104a.json b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025/09d4a446-23db-475a-bf6e-70625e23104a.json new file mode 100644 index 0000000000000000000000000000000000000000..299a195af501410d0223aa758dde32b24cdbe74a --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025/09d4a446-23db-475a-bf6e-70625e23104a.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/google_gemini-2.5-flash-lite-preview-09-2025/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-lite-preview-09-2025", + "id": "google/gemini-2.5-flash-lite-preview-09-2025", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-lite-preview-09-2025", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 53.73, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "53.73", + "raw_stderr": "1.702", + "latency": "59.381", + "cost_per_test": "0.000222", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.702, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 59.907, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.907", + "raw_stderr": "1.673", + "latency": "70.729", + "cost_per_test": "0.018892", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.673, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 56.294, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.294", + "raw_stderr": "0.976", + "latency": "54.75", + "cost_per_test": "0.008646", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.976, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 55.245, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "55.245", + "raw_stderr": "1.698", + "latency": "34.139", + "cost_per_test": "0.006824", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.698, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025/20307645-63fc-4f52-9aef-7f33b022df59.json b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025/20307645-63fc-4f52-9aef-7f33b022df59.json new file mode 100644 index 0000000000000000000000000000000000000000..ae206569bcd024135f39d50f101004e4cd4d96bd --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025/20307645-63fc-4f52-9aef-7f33b022df59.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/google_gemini-2.5-flash-lite-preview-09-2025/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-lite-preview-09-2025", + "id": "google/gemini-2.5-flash-lite-preview-09-2025", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-lite-preview-09-2025", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 75.824, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "75.824", + "raw_stderr": "1.851", + "latency": "4.491", + "cost_per_test": "0.001332", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.851, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025/4fff6250-1382-4894-8608-94ebc7de54c4.json b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025/4fff6250-1382-4894-8608-94ebc7de54c4.json new file mode 100644 index 0000000000000000000000000000000000000000..727a33d9548128221d546c9a6f8b70fc3ab0cc40 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025/4fff6250-1382-4894-8608-94ebc7de54c4.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/google_gemini-2.5-flash-lite-preview-09-2025/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-lite-preview-09-2025", + "id": "google/gemini-2.5-flash-lite-preview-09-2025", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-lite-preview-09-2025", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 97.516, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "97.516", + "raw_stderr": "0.867", + "latency": "3.293", + "cost_per_test": "0.000512", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.867, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 33.429, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.429", + "raw_stderr": "2.522", + "latency": "23.363", + "cost_per_test": "0.003753", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.522, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 72.063, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "72.063", + "raw_stderr": "2.293", + "latency": "25.669", + "cost_per_test": "0.001743", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.293, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 67.669, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "67.669", + "raw_stderr": "1.172", + "latency": "17.442", + "cost_per_test": "0.002003", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.172, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025/815ce2ed-6957-4b5c-b221-055025d25cff.json b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025/815ce2ed-6957-4b5c-b221-055025d25cff.json new file mode 100644 index 0000000000000000000000000000000000000000..62a1ea2d244884cd4e9150a4eecf62702e11fa5f --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025/815ce2ed-6957-4b5c-b221-055025d25cff.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/google_gemini-2.5-flash-lite-preview-09-2025/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-lite-preview-09-2025", + "id": "google/gemini-2.5-flash-lite-preview-09-2025", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-lite-preview-09-2025", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 62.121, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "62.121", + "raw_stderr": "3.447", + "latency": "7.17", + "cost_per_test": "0.001336", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.447, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 66.162, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.162", + "raw_stderr": "3.375", + "latency": "7.377", + "cost_per_test": "0.001135", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.375, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 64.142, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "64.142", + "raw_stderr": "2.412", + "latency": "7.274", + "cost_per_test": "0.001236", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.412, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025/93dcdf93-de2d-4cb5-b209-b89c4769a782.json b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025/93dcdf93-de2d-4cb5-b209-b89c4769a782.json new file mode 100644 index 0000000000000000000000000000000000000000..502be6d8a5c7fdfbf83201e42c6f7b51a7c5ae36 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025/93dcdf93-de2d-4cb5-b209-b89c4769a782.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/google_gemini-2.5-flash-lite-preview-09-2025/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-lite-preview-09-2025", + "id": "google/gemini-2.5-flash-lite-preview-09-2025", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-lite-preview-09-2025", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 53.966, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "53.966", + "raw_stderr": "1.425", + "latency": "3.597", + "cost_per_test": "0.000477", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.425, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 66.231, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.231", + "raw_stderr": "0.923", + "latency": "3.597", + "cost_per_test": "0.000477", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.923, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 78.496, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "78.496", + "raw_stderr": "1.175", + "latency": "3.597", + "cost_per_test": "0.000477", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.175, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025/b64c0d76-ba16-4e0a-8edc-95085cfcf79c.json b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025/b64c0d76-ba16-4e0a-8edc-95085cfcf79c.json new file mode 100644 index 0000000000000000000000000000000000000000..1ac07a49c6690b9b2020ae6c7a4e7578c922e52f --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025/b64c0d76-ba16-4e0a-8edc-95085cfcf79c.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/google_gemini-2.5-flash-lite-preview-09-2025/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-lite-preview-09-2025", + "id": "google/gemini-2.5-flash-lite-preview-09-2025", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-lite-preview-09-2025", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 55.167, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "55.167", + "raw_stderr": "1.402", + "latency": "13.637", + "cost_per_test": "0.000622", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.402, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 55.803, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "55.803", + "raw_stderr": "0.99", + "latency": "8.042", + "cost_per_test": "0.000479", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.99, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 56.439, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.439", + "raw_stderr": "1.398", + "latency": "2.448", + "cost_per_test": "0.000336", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.398, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025/bf786dee-a96f-4a76-9815-f62f4899cc6e.json b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025/bf786dee-a96f-4a76-9815-f62f4899cc6e.json new file mode 100644 index 0000000000000000000000000000000000000000..b3dc685468d88e521ada2371934447e80bd41587 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025/bf786dee-a96f-4a76-9815-f62f4899cc6e.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/google_gemini-2.5-flash-lite-preview-09-2025/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-lite-preview-09-2025", + "id": "google/gemini-2.5-flash-lite-preview-09-2025", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-lite-preview-09-2025", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.878, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.878", + "raw_stderr": "1.051", + "latency": "0.359", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.051, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 76.624, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "76.624", + "raw_stderr": "0.192", + "latency": "0.369", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.192, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 86.447, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.447", + "raw_stderr": "0.619", + "latency": "0.402", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.619, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.008, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.008", + "raw_stderr": "0.443", + "latency": "0.456", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.443, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.108, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.108", + "raw_stderr": "0.907", + "latency": "0.366", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.907, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 73.981, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "73.981", + "raw_stderr": "1.598", + "latency": "0.783", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.598, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025/c634277c-1637-450b-8d17-7c616deccc57.json b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025/c634277c-1637-450b-8d17-7c616deccc57.json new file mode 100644 index 0000000000000000000000000000000000000000..2bb3be0cc5bdb7bc40218799306fc09588529364 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025/c634277c-1637-450b-8d17-7c616deccc57.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/google_gemini-2.5-flash-lite-preview-09-2025/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-lite-preview-09-2025", + "id": "google/gemini-2.5-flash-lite-preview-09-2025", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-lite-preview-09-2025", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.563, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.563", + "raw_stderr": "1.189", + "latency": "2.685", + "cost_per_test": "0.000451", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.189, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.397, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.397", + "raw_stderr": "1.325", + "latency": "2.586", + "cost_per_test": "0.000453", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.325, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.336, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.336", + "raw_stderr": "1.051", + "latency": "4.106", + "cost_per_test": "0.000798", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.051, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.951, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.951", + "raw_stderr": "1.899", + "latency": "3.595", + "cost_per_test": "0.00063", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.899, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.479, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.479", + "raw_stderr": "1.246", + "latency": "2.493", + "cost_per_test": "0.000428", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.246, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 68.524, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "68.524", + "raw_stderr": "1.492", + "latency": "6.686", + "cost_per_test": "0.001175", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.492, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.061, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.061", + "raw_stderr": "1.513", + "latency": "2.435", + "cost_per_test": "0.000374", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.513, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 67.979, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "67.979", + "raw_stderr": "2.39", + "latency": "2.792", + "cost_per_test": "0.000524", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.39, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 60.581, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "60.581", + "raw_stderr": "1.473", + "latency": "3.583", + "cost_per_test": "0.000563", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.473, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.044, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.044", + "raw_stderr": "0.777", + "latency": "3.386", + "cost_per_test": "0.000703", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.777, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.407, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.407", + "raw_stderr": "1.397", + "latency": "2.052", + "cost_per_test": "0.0003", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.397, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.639, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.639", + "raw_stderr": "0.4", + "latency": "3.19", + "cost_per_test": "0.000556", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.4, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.347, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.347", + "raw_stderr": "1.979", + "latency": "2.348", + "cost_per_test": "0.000367", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.979, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.45, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.45", + "raw_stderr": "1.005", + "latency": "3.727", + "cost_per_test": "0.000677", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.005, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.825, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.825", + "raw_stderr": "1.421", + "latency": "2.191", + "cost_per_test": "0.000338", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.421, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025/e20921da-e0bd-4dd5-afab-0b7b8f9c97f5.json b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025/e20921da-e0bd-4dd5-afab-0b7b8f9c97f5.json new file mode 100644 index 0000000000000000000000000000000000000000..a7994f28e1b498ff4aa47ccb0c3b7c22661d1546 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025/e20921da-e0bd-4dd5-afab-0b7b8f9c97f5.json @@ -0,0 +1,101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/google_gemini-2.5-flash-lite-preview-09-2025/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-lite-preview-09-2025", + "id": "google/gemini-2.5-flash-lite-preview-09-2025", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-lite-preview-09-2025", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 72.543, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.543", + "raw_stderr": "1.073", + "latency": "20.069", + "cost_per_test": "0.000899", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.073, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025/ffaeb09b-3c2d-4182-8c18-9e69726d9d22.json b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025/ffaeb09b-3c2d-4182-8c18-9e69726d9d22.json new file mode 100644 index 0000000000000000000000000000000000000000..7feab098a1f39c765daca514adae447a3066514e --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-lite-preview-09-2025/ffaeb09b-3c2d-4182-8c18-9e69726d9d22.json @@ -0,0 +1,346 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/google_gemini-2.5-flash-lite-preview-09-2025/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-lite-preview-09-2025", + "id": "google/gemini-2.5-flash-lite-preview-09-2025", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-lite-preview-09-2025", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 13.492, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "13.492", + "raw_stderr": "4.287", + "latency": "11.273", + "cost_per_test": "0.001268", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 4.287, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 26.588, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "26.588", + "raw_stderr": "5.653", + "latency": "17.054", + "cost_per_test": "0.00094", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.653, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 28.086, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "28.086", + "raw_stderr": "5.876", + "latency": "30.662", + "cost_per_test": "0.0019", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.876, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 43.84, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "43.84", + "raw_stderr": "5.573", + "latency": "15.696", + "cost_per_test": "0.001188", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.573, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:google/gemini-2.5-flash-lite-preview-09-2025:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 27.973, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "27.973", + "raw_stderr": "3.007", + "latency": "14.674", + "cost_per_test": "0.001132", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.007, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-lite/4d660fa4-ae20-4976-85f8-b0e8d98c5694.json b/data/vals_ai/google/gemini-2.5-flash-lite/4d660fa4-ae20-4976-85f8-b0e8d98c5694.json new file mode 100644 index 0000000000000000000000000000000000000000..66b80d6f8c49129432e6ae83ec516172c21719df --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-lite/4d660fa4-ae20-4976-85f8-b0e8d98c5694.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/google_gemini-2.5-flash-lite/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-lite", + "id": "google/gemini-2.5-flash-lite", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-lite", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:google/gemini-2.5-flash-lite:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 72.832, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "72.832", + "raw_stderr": "1.982", + "latency": "4.847", + "cost_per_test": "0.001211", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.982, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-lite/809b8cfb-967e-4ffb-9221-269c40ee88d0.json b/data/vals_ai/google/gemini-2.5-flash-lite/809b8cfb-967e-4ffb-9221-269c40ee88d0.json new file mode 100644 index 0000000000000000000000000000000000000000..6b7a722c935b5f0deb06455f5e20cd5aac977721 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-lite/809b8cfb-967e-4ffb-9221-269c40ee88d0.json @@ -0,0 +1,346 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/google_gemini-2.5-flash-lite/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-lite", + "id": "google/gemini-2.5-flash-lite", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-lite", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:google/gemini-2.5-flash-lite:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 20.635, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "20.635", + "raw_stderr": "5.078", + "latency": "12.776", + "cost_per_test": "0.001121", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.078, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:google/gemini-2.5-flash-lite:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 33.334, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "33.334", + "raw_stderr": "6.023", + "latency": "10.448", + "cost_per_test": "0.000876", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.023, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:google/gemini-2.5-flash-lite:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 13.889, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "13.889", + "raw_stderr": "4.298", + "latency": "13.337", + "cost_per_test": "0.001699", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 4.298, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:google/gemini-2.5-flash-lite:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 28.629, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "28.629", + "raw_stderr": "5.077", + "latency": "18.12", + "cost_per_test": "0.000791", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.077, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:google/gemini-2.5-flash-lite:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 27.533, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "27.533", + "raw_stderr": "3.124", + "latency": "13.781", + "cost_per_test": "0.000929", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.124, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-preview-04-17-thinking/1483ded8-66e5-482e-9e2a-ace743fe6ef4.json b/data/vals_ai/google/gemini-2.5-flash-preview-04-17-thinking/1483ded8-66e5-482e-9e2a-ace743fe6ef4.json new file mode 100644 index 0000000000000000000000000000000000000000..5868647bd203ba447cc99b620f9ba0a4c1427898 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-preview-04-17-thinking/1483ded8-66e5-482e-9e2a-ace743fe6ef4.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/google_gemini-2.5-flash-preview-04-17-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-preview-04-17-thinking", + "id": "google/gemini-2.5-flash-preview-04-17-thinking", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-preview-04-17-thinking", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:google/gemini-2.5-flash-preview-04-17-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 57.313, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.313", + "raw_stderr": "1.395", + "latency": "10.651", + "cost_per_test": "0.005675", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.395, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:google/gemini-2.5-flash-preview-04-17-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 57.512, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.512", + "raw_stderr": "0.986", + "latency": "8.4", + "cost_per_test": "0.0041", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.986, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:google/gemini-2.5-flash-preview-04-17-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 57.711, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.711", + "raw_stderr": "1.393", + "latency": "6.149", + "cost_per_test": "0.002525", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.393, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-preview-04-17-thinking/699b973d-91d6-4d9b-8e42-2003294bf8b4.json b/data/vals_ai/google/gemini-2.5-flash-preview-04-17-thinking/699b973d-91d6-4d9b-8e42-2003294bf8b4.json new file mode 100644 index 0000000000000000000000000000000000000000..12709d8d18f11386a8611cb26e119663e80ffb15 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-preview-04-17-thinking/699b973d-91d6-4d9b-8e42-2003294bf8b4.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/google_gemini-2.5-flash-preview-04-17-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-preview-04-17-thinking", + "id": "google/gemini-2.5-flash-preview-04-17-thinking", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-preview-04-17-thinking", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:google/gemini-2.5-flash-preview-04-17-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 58.872, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.872", + "raw_stderr": "1.407", + "latency": "15.355", + "cost_per_test": "0.007461", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.407, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:google/gemini-2.5-flash-preview-04-17-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 70.524, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "70.524", + "raw_stderr": "0.891", + "latency": "15.355", + "cost_per_test": "0.007461", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.891, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:google/gemini-2.5-flash-preview-04-17-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 82.175, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "82.175", + "raw_stderr": "1.094", + "latency": "15.355", + "cost_per_test": "0.007461", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.094, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-preview-04-17-thinking/887bac25-1c47-4c3f-ba9e-6d56430dc238.json b/data/vals_ai/google/gemini-2.5-flash-preview-04-17-thinking/887bac25-1c47-4c3f-ba9e-6d56430dc238.json new file mode 100644 index 0000000000000000000000000000000000000000..c9896f703b3c8c75cb3539b885da2753d0c16e8d --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-preview-04-17-thinking/887bac25-1c47-4c3f-ba9e-6d56430dc238.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/google_gemini-2.5-flash-preview-04-17-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-preview-04-17-thinking", + "id": "google/gemini-2.5-flash-preview-04-17-thinking", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-preview-04-17-thinking", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:google/gemini-2.5-flash-preview-04-17-thinking:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 68.634, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "68.634", + "raw_stderr": "2.586", + "latency": "82.664", + "cost_per_test": "0.000715", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.586, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:google/gemini-2.5-flash-preview-04-17-thinking:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 26.286, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "26.286", + "raw_stderr": "2.353", + "latency": "159.017", + "cost_per_test": "0.00387", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.353, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:google/gemini-2.5-flash-preview-04-17-thinking:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 45.692, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "45.692", + "raw_stderr": "2.545", + "latency": "133.159", + "cost_per_test": "0.001376", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.545, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:google/gemini-2.5-flash-preview-04-17-thinking:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 46.871, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "46.871", + "raw_stderr": "1.441", + "latency": "124.947", + "cost_per_test": "0.001987", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.441, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-preview-04-17-thinking/bebe3121-071e-4193-afd7-329e032aee68.json b/data/vals_ai/google/gemini-2.5-flash-preview-04-17-thinking/bebe3121-071e-4193-afd7-329e032aee68.json new file mode 100644 index 0000000000000000000000000000000000000000..441b947f7cb7b6320767f3f19251f899d65c9171 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-preview-04-17-thinking/bebe3121-071e-4193-afd7-329e032aee68.json @@ -0,0 +1,101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/google_gemini-2.5-flash-preview-04-17-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-preview-04-17-thinking", + "id": "google/gemini-2.5-flash-preview-04-17-thinking", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-preview-04-17-thinking", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:google/gemini-2.5-flash-preview-04-17-thinking:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 71.924, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.924", + "raw_stderr": "1.08", + "latency": "471.866", + "cost_per_test": "0.002641", + "temperature": "0.1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.08, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.1, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-preview-04-17/03669816-7d04-4722-a335-1522a93db6ac.json b/data/vals_ai/google/gemini-2.5-flash-preview-04-17/03669816-7d04-4722-a335-1522a93db6ac.json new file mode 100644 index 0000000000000000000000000000000000000000..eb3c71575861915d62b52585c549492806c6aa92 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-preview-04-17/03669816-7d04-4722-a335-1522a93db6ac.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/google_gemini-2.5-flash-preview-04-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-preview-04-17", + "id": "google/gemini-2.5-flash-preview-04-17", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-preview-04-17", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:google/gemini-2.5-flash-preview-04-17:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 54.779, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "54.779", + "raw_stderr": "1.699", + "latency": "1.657", + "cost_per_test": "0.000864", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.699, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:google/gemini-2.5-flash-preview-04-17:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 52.331, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.331", + "raw_stderr": "1.705", + "latency": "20.232", + "cost_per_test": "0.040537", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.705, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:google/gemini-2.5-flash-preview-04-17:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 54.157, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "54.157", + "raw_stderr": "0.982", + "latency": "8.992", + "cost_per_test": "0.020528", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.982, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:google/gemini-2.5-flash-preview-04-17:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 55.361, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "55.361", + "raw_stderr": "1.697", + "latency": "5.088", + "cost_per_test": "0.020182", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.697, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-preview-04-17/4b3bd002-275b-4f06-a4cf-e23d683aa43a.json b/data/vals_ai/google/gemini-2.5-flash-preview-04-17/4b3bd002-275b-4f06-a4cf-e23d683aa43a.json new file mode 100644 index 0000000000000000000000000000000000000000..af6c0e1a73cd12572ed88f040187222d45abface --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-preview-04-17/4b3bd002-275b-4f06-a4cf-e23d683aa43a.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/google_gemini-2.5-flash-preview-04-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-preview-04-17", + "id": "google/gemini-2.5-flash-preview-04-17", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-preview-04-17", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:google/gemini-2.5-flash-preview-04-17:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 64.646, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "64.646", + "raw_stderr": "3.456", + "latency": "48.056", + "cost_per_test": "0.03391", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.456, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:google/gemini-2.5-flash-preview-04-17:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 50.505, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "50.505", + "raw_stderr": "3.363", + "latency": "38.848", + "cost_per_test": "0.026324", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.363, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:google/gemini-2.5-flash-preview-04-17:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 57.576, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "57.576", + "raw_stderr": "2.411", + "latency": "43.452", + "cost_per_test": "0.030117", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.411, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-preview-04-17/67bf4b76-39eb-42e6-a3ce-571cf4353e5c.json b/data/vals_ai/google/gemini-2.5-flash-preview-04-17/67bf4b76-39eb-42e6-a3ce-571cf4353e5c.json new file mode 100644 index 0000000000000000000000000000000000000000..8571d31cf3041a1735707c4ab8d3b32750db5c1b --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-preview-04-17/67bf4b76-39eb-42e6-a3ce-571cf4353e5c.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/google_gemini-2.5-flash-preview-04-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-preview-04-17", + "id": "google/gemini-2.5-flash-preview-04-17", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-preview-04-17", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:google/gemini-2.5-flash-preview-04-17:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 94.72, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "94.72", + "raw_stderr": "1.246", + "latency": "1.696", + "cost_per_test": "0.001012", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.246, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:google/gemini-2.5-flash-preview-04-17:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 19.429, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "19.429", + "raw_stderr": "2.115", + "latency": "35.389", + "cost_per_test": "0.025301", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.115, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:google/gemini-2.5-flash-preview-04-17:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 56.658, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "56.658", + "raw_stderr": "2.532", + "latency": "12.137", + "cost_per_test": "0.008849", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.532, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:google/gemini-2.5-flash-preview-04-17:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 56.936, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "56.936", + "raw_stderr": "1.176", + "latency": "16.407", + "cost_per_test": "0.011721", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.176, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-preview-04-17/8c5f73ca-f944-4814-97df-249f792a287f.json b/data/vals_ai/google/gemini-2.5-flash-preview-04-17/8c5f73ca-f944-4814-97df-249f792a287f.json new file mode 100644 index 0000000000000000000000000000000000000000..b1183147c57574d0a219468eb86e3118a9f7d426 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-preview-04-17/8c5f73ca-f944-4814-97df-249f792a287f.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/google_gemini-2.5-flash-preview-04-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-preview-04-17", + "id": "google/gemini-2.5-flash-preview-04-17", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-preview-04-17", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:google/gemini-2.5-flash-preview-04-17:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 59.608, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.608", + "raw_stderr": "1.403", + "latency": "5.388", + "cost_per_test": "0.00261", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.403, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:google/gemini-2.5-flash-preview-04-17:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 71.178, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.178", + "raw_stderr": "0.885", + "latency": "5.388", + "cost_per_test": "0.00261", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.885, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:google/gemini-2.5-flash-preview-04-17:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 82.747, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "82.747", + "raw_stderr": "1.08", + "latency": "5.388", + "cost_per_test": "0.00261", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.08, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-preview-04-17/9f9d00c0-0c7c-42a5-ac5e-89e6065398a4.json b/data/vals_ai/google/gemini-2.5-flash-preview-04-17/9f9d00c0-0c7c-42a5-ac5e-89e6065398a4.json new file mode 100644 index 0000000000000000000000000000000000000000..193cd659896920b5c81227d24c17cf80064b006d --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-preview-04-17/9f9d00c0-0c7c-42a5-ac5e-89e6065398a4.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/google_gemini-2.5-flash-preview-04-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-preview-04-17", + "id": "google/gemini-2.5-flash-preview-04-17", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-preview-04-17", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:google/gemini-2.5-flash-preview-04-17:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 62.083, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.083", + "raw_stderr": "1.368", + "latency": "4.467", + "cost_per_test": "0.002617", + "temperature": "0.1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.368, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.1, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:google/gemini-2.5-flash-preview-04-17:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 58.426, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.426", + "raw_stderr": "0.98", + "latency": "3.556", + "cost_per_test": "0.001974", + "temperature": "0.1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.98, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.1, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:google/gemini-2.5-flash-preview-04-17:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 54.769, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "54.769", + "raw_stderr": "1.403", + "latency": "2.644", + "cost_per_test": "0.00133", + "temperature": "0.1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.403, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.1, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-preview-04-17/b6889005-f5ae-4ed2-a32f-30b2b02ac1be.json b/data/vals_ai/google/gemini-2.5-flash-preview-04-17/b6889005-f5ae-4ed2-a32f-30b2b02ac1be.json new file mode 100644 index 0000000000000000000000000000000000000000..68208d0065f548636fb7cc41b884b14789d08914 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-preview-04-17/b6889005-f5ae-4ed2-a32f-30b2b02ac1be.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/google_gemini-2.5-flash-preview-04-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-preview-04-17", + "id": "google/gemini-2.5-flash-preview-04-17", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-preview-04-17", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:google/gemini-2.5-flash-preview-04-17:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 86.958, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.958", + "raw_stderr": "0.881", + "latency": "0.447", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.881, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:google/gemini-2.5-flash-preview-04-17:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.564, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.564", + "raw_stderr": "0.19", + "latency": "0.416", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.19, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:google/gemini-2.5-flash-preview-04-17:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 89.456, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.456", + "raw_stderr": "0.558", + "latency": "0.425", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.558, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:google/gemini-2.5-flash-preview-04-17:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.796, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.796", + "raw_stderr": "0.399", + "latency": "0.545", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.399, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:google/gemini-2.5-flash-preview-04-17:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.207, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.207", + "raw_stderr": "0.824", + "latency": "0.423", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.824, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:google/gemini-2.5-flash-preview-04-17:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.796, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.796", + "raw_stderr": "1.478", + "latency": "1.015", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.478, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-preview-04-17/c756e74c-bca0-4f43-8ca6-6d0306e1030e.json b/data/vals_ai/google/gemini-2.5-flash-preview-04-17/c756e74c-bca0-4f43-8ca6-6d0306e1030e.json new file mode 100644 index 0000000000000000000000000000000000000000..f59baf2154d005f07840c8facc5a8a17340f11aa --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-preview-04-17/c756e74c-bca0-4f43-8ca6-6d0306e1030e.json @@ -0,0 +1,101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/google_gemini-2.5-flash-preview-04-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-preview-04-17", + "id": "google/gemini-2.5-flash-preview-04-17", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-preview-04-17", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:google/gemini-2.5-flash-preview-04-17:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 67.995, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "67.995", + "raw_stderr": "1.121", + "latency": "37.103", + "cost_per_test": "0.02536", + "temperature": "0.1", + "max_output_tokens": "16384", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.121, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.1, + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-preview-09-2025-thinking/28fc0650-0f72-4b7c-a4e0-e945b378f22b.json b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025-thinking/28fc0650-0f72-4b7c-a4e0-e945b378f22b.json new file mode 100644 index 0000000000000000000000000000000000000000..a1fab57304abb89c209aa40bff1175cc3ee8628b --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025-thinking/28fc0650-0f72-4b7c-a4e0-e945b378f22b.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/google_gemini-2.5-flash-preview-09-2025-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-preview-09-2025-thinking", + "id": "google/gemini-2.5-flash-preview-09-2025-thinking", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-preview-09-2025-thinking", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 56.76, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.76", + "raw_stderr": "1.691", + "latency": "57.491", + "cost_per_test": "0.00088", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.691, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.606, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.606", + "raw_stderr": "1.668", + "latency": "67.517", + "cost_per_test": "0.056863", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.668, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 59.751, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.751", + "raw_stderr": "0.966", + "latency": "51.757", + "cost_per_test": "0.026119", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.966, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.888, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.888", + "raw_stderr": "1.658", + "latency": "30.263", + "cost_per_test": "0.020614", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.658, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-preview-09-2025-thinking/5350733b-6803-4bf8-815e-fbcb901ce65a.json b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025-thinking/5350733b-6803-4bf8-815e-fbcb901ce65a.json new file mode 100644 index 0000000000000000000000000000000000000000..371f22ae1cccd9aabd73586c3fb5f7621aded3bc --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025-thinking/5350733b-6803-4bf8-815e-fbcb901ce65a.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/google_gemini-2.5-flash-preview-09-2025-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-preview-09-2025-thinking", + "id": "google/gemini-2.5-flash-preview-09-2025-thinking", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-preview-09-2025-thinking", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 68.919, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.919", + "raw_stderr": "1.305", + "latency": "28.199", + "cost_per_test": "0.001186", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.305, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 61.924, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.924", + "raw_stderr": "0.958", + "latency": "18.331", + "cost_per_test": "0.001154", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.958, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 54.928, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "54.928", + "raw_stderr": "1.403", + "latency": "8.463", + "cost_per_test": "0.001121", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.403, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-preview-09-2025-thinking/5bb50120-658e-480d-9b90-9b2ed744d422.json b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025-thinking/5bb50120-658e-480d-9b90-9b2ed744d422.json new file mode 100644 index 0000000000000000000000000000000000000000..c132868a516e5b1b5b41a5c6723c6ebce45589ff --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025-thinking/5bb50120-658e-480d-9b90-9b2ed744d422.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/google_gemini-2.5-flash-preview-09-2025-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-preview-09-2025-thinking", + "id": "google/gemini-2.5-flash-preview-09-2025-thinking", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-preview-09-2025-thinking", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.329, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.329", + "raw_stderr": "0.994", + "latency": "8.652", + "cost_per_test": "0.00127", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.994, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.861, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.861", + "raw_stderr": "1.075", + "latency": "8.597", + "cost_per_test": "0.001312", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.075, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.841, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.841", + "raw_stderr": "0.898", + "latency": "13.498", + "cost_per_test": "0.002146", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.898, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.073, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.073", + "raw_stderr": "1.657", + "latency": "10.423", + "cost_per_test": "0.001574", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.657, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.441, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.441", + "raw_stderr": "1.141", + "latency": "8.176", + "cost_per_test": "0.001203", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.141, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.193, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.193", + "raw_stderr": "1.348", + "latency": "20.916", + "cost_per_test": "0.002718", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.348, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.139, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.139", + "raw_stderr": "1.468", + "latency": "8.366", + "cost_per_test": "0.001022", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.468, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.378, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.378", + "raw_stderr": "2.176", + "latency": "9.309", + "cost_per_test": "0.001366", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.176, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 68.756, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "68.756", + "raw_stderr": "1.397", + "latency": "11.802", + "cost_per_test": "0.001468", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.397, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.708, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.708", + "raw_stderr": "0.661", + "latency": "10.711", + "cost_per_test": "0.001851", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.661, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.545, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.545", + "raw_stderr": "1.327", + "latency": "7.515", + "cost_per_test": "0.000798", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.327, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.664, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.664", + "raw_stderr": "0.361", + "latency": "10.445", + "cost_per_test": "0.00146", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.361, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.758, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.758", + "raw_stderr": "1.831", + "latency": "8.332", + "cost_per_test": "0.001001", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.831, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.684, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.684", + "raw_stderr": "0.879", + "latency": "11.403", + "cost_per_test": "0.001754", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.879, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.586, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.586", + "raw_stderr": "1.278", + "latency": "8.534", + "cost_per_test": "0.00095", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.278, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-preview-09-2025-thinking/9afd4474-3f28-4df0-b7a0-10fed8627eb5.json b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025-thinking/9afd4474-3f28-4df0-b7a0-10fed8627eb5.json new file mode 100644 index 0000000000000000000000000000000000000000..5571ae76c6063d01220f7cc83d2a0f05ae793447 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025-thinking/9afd4474-3f28-4df0-b7a0-10fed8627eb5.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/google_gemini-2.5-flash-preview-09-2025-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-preview-09-2025-thinking", + "id": "google/gemini-2.5-flash-preview-09-2025-thinking", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-preview-09-2025-thinking", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 75.758, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.758", + "raw_stderr": "3.046", + "latency": "23.633", + "cost_per_test": "0.002925", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.046, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 77.273, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.273", + "raw_stderr": "3.046", + "latency": "23.427", + "cost_per_test": "0.002741", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.046, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 76.516, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.516", + "raw_stderr": "2.154", + "latency": "23.53", + "cost_per_test": "0.002833", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.154, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-preview-09-2025-thinking/a307620c-2e45-40c9-87ed-e9f11f95134e.json b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025-thinking/a307620c-2e45-40c9-87ed-e9f11f95134e.json new file mode 100644 index 0000000000000000000000000000000000000000..3792a918e483752ed96b7e5ce0b496bf451b97ae --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025-thinking/a307620c-2e45-40c9-87ed-e9f11f95134e.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/google_gemini-2.5-flash-preview-09-2025-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-preview-09-2025-thinking", + "id": "google/gemini-2.5-flash-preview-09-2025-thinking", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-preview-09-2025-thinking", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 97.826, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "97.826", + "raw_stderr": "0.813", + "latency": "18.059", + "cost_per_test": "0.001072", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.813, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 48.571, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "48.571", + "raw_stderr": "2.672", + "latency": "83.508", + "cost_per_test": "0.003962", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.672, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 82.245, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "82.245", + "raw_stderr": "1.953", + "latency": "44.446", + "cost_per_test": "0.002115", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.953, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 76.214, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "76.214", + "raw_stderr": "1.136", + "latency": "48.671", + "cost_per_test": "0.002383", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.136, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-preview-09-2025-thinking/ac7ea992-9c4c-435c-903b-0558848bf37c.json b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025-thinking/ac7ea992-9c4c-435c-903b-0558848bf37c.json new file mode 100644 index 0000000000000000000000000000000000000000..efa0adb097a649c86250ecd9347ccf07ef4b813f --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025-thinking/ac7ea992-9c4c-435c-903b-0558848bf37c.json @@ -0,0 +1,346 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/google_gemini-2.5-flash-preview-09-2025-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-preview-09-2025-thinking", + "id": "google/gemini-2.5-flash-preview-09-2025-thinking", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-preview-09-2025-thinking", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 30.159, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "30.159", + "raw_stderr": "5.759", + "latency": "26.613", + "cost_per_test": "0.002299", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.759, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 34.524, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "34.524", + "raw_stderr": "5.983", + "latency": "32.116", + "cost_per_test": "0.00228", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.983, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 54.012, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "54.012", + "raw_stderr": "6.484", + "latency": "65.915", + "cost_per_test": "0.005383", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.484, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 53.441, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "53.441", + "raw_stderr": "5.603", + "latency": "23.095", + "cost_per_test": "0.002453", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.603, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 39.375, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "39.375", + "raw_stderr": "3.339", + "latency": "27.275", + "cost_per_test": "0.002344", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.339, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-preview-09-2025-thinking/cb626bed-2c7c-4854-86bb-b57a73b76620.json b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025-thinking/cb626bed-2c7c-4854-86bb-b57a73b76620.json new file mode 100644 index 0000000000000000000000000000000000000000..89c30861373f8338da05ccdbc1e3ea5ee1706944 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025-thinking/cb626bed-2c7c-4854-86bb-b57a73b76620.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/google_gemini-2.5-flash-preview-09-2025-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-preview-09-2025-thinking", + "id": "google/gemini-2.5-flash-preview-09-2025-thinking", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-preview-09-2025-thinking", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 85.782, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.782", + "raw_stderr": "0.925", + "latency": "1.464", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.925, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.483, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.483", + "raw_stderr": "0.194", + "latency": "2.31", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.194, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.565, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.565", + "raw_stderr": "0.569", + "latency": "1.881", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.569, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.625, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.625", + "raw_stderr": "0.414", + "latency": "1.965", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.414, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.406, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.406", + "raw_stderr": "0.719", + "latency": "1.43", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.719, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.888, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.888", + "raw_stderr": "1.596", + "latency": "2.739", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.596, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-preview-09-2025-thinking/cead9c37-c42c-4bc1-9cd5-a9c7bb23a499.json b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025-thinking/cead9c37-c42c-4bc1-9cd5-a9c7bb23a499.json new file mode 100644 index 0000000000000000000000000000000000000000..02965064b5c13bcc86a3989e643733fa2ae89ddd --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025-thinking/cead9c37-c42c-4bc1-9cd5-a9c7bb23a499.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/google_gemini-2.5-flash-preview-09-2025-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-preview-09-2025-thinking", + "id": "google/gemini-2.5-flash-preview-09-2025-thinking", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-preview-09-2025-thinking", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 60.67, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.67", + "raw_stderr": "1.397", + "latency": "10.635", + "cost_per_test": "0.003099", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.397, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 72.404, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "72.404", + "raw_stderr": "0.872", + "latency": "10.635", + "cost_per_test": "0.003099", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.872, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 84.137, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "84.137", + "raw_stderr": "1.045", + "latency": "10.635", + "cost_per_test": "0.003099", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.045, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-preview-09-2025-thinking/d6b2b454-6d72-42b3-be80-3f721541db80.json b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025-thinking/d6b2b454-6d72-42b3-be80-3f721541db80.json new file mode 100644 index 0000000000000000000000000000000000000000..fa209825bced54b6cfda3dd44c59382bcdc7c164 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025-thinking/d6b2b454-6d72-42b3-be80-3f721541db80.json @@ -0,0 +1,101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/google_gemini-2.5-flash-preview-09-2025-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-preview-09-2025-thinking", + "id": "google/gemini-2.5-flash-preview-09-2025-thinking", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-preview-09-2025-thinking", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 80.751, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.751", + "raw_stderr": "0.948", + "latency": "27.866", + "cost_per_test": "0.001858", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.948, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-preview-09-2025-thinking/ef6b6b6c-87fc-42d8-8474-ae6d7e9513a0.json b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025-thinking/ef6b6b6c-87fc-42d8-8474-ae6d7e9513a0.json new file mode 100644 index 0000000000000000000000000000000000000000..9e8affd22ecca48dfdd970b63f7834b5d234de55 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025-thinking/ef6b6b6c-87fc-42d8-8474-ae6d7e9513a0.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/google_gemini-2.5-flash-preview-09-2025-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-preview-09-2025-thinking", + "id": "google/gemini-2.5-flash-preview-09-2025-thinking", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-preview-09-2025-thinking", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:google/gemini-2.5-flash-preview-09-2025-thinking:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 78.497, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "78.497", + "raw_stderr": "1.993", + "latency": "31.385", + "cost_per_test": "0.014526", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.993, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-preview-09-2025/14f050d8-f194-4a3e-b7b2-b1e568253974.json b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025/14f050d8-f194-4a3e-b7b2-b1e568253974.json new file mode 100644 index 0000000000000000000000000000000000000000..3e08be83896c8d61bd236632bb026a3ef43db746 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025/14f050d8-f194-4a3e-b7b2-b1e568253974.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/google_gemini-2.5-flash-preview-09-2025/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-preview-09-2025", + "id": "google/gemini-2.5-flash-preview-09-2025", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-preview-09-2025", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 77.946, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "77.946", + "raw_stderr": "1.923", + "latency": "22.893", + "cost_per_test": "0.014385", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.923, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-preview-09-2025/4dc67aef-3f4e-4417-b937-da4cb701e6ea.json b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025/4dc67aef-3f4e-4417-b937-da4cb701e6ea.json new file mode 100644 index 0000000000000000000000000000000000000000..dac40ec2cf67501dcc77d08eb61c950feceadf31 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025/4dc67aef-3f4e-4417-b937-da4cb701e6ea.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/google_gemini-2.5-flash-preview-09-2025/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-preview-09-2025", + "id": "google/gemini-2.5-flash-preview-09-2025", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-preview-09-2025", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 97.826, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "97.826", + "raw_stderr": "0.813", + "latency": "14.3", + "cost_per_test": "0.001109", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.813, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 44.857, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "44.857", + "raw_stderr": "2.658", + "latency": "84.82", + "cost_per_test": "0.003876", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.658, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 82.507, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "82.507", + "raw_stderr": "1.941", + "latency": "41.234", + "cost_per_test": "0.002115", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.941, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 75.063, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "75.063", + "raw_stderr": "1.13", + "latency": "46.785", + "cost_per_test": "0.002367", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.13, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-preview-09-2025/76637a9c-8062-492f-9162-5eceb880d5d5.json b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025/76637a9c-8062-492f-9162-5eceb880d5d5.json new file mode 100644 index 0000000000000000000000000000000000000000..acaf6aa23f911b67e01fca1057cbfaee18e409ea --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025/76637a9c-8062-492f-9162-5eceb880d5d5.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/google_gemini-2.5-flash-preview-09-2025/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-preview-09-2025", + "id": "google/gemini-2.5-flash-preview-09-2025", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-preview-09-2025", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.05, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.05", + "raw_stderr": "1.01", + "latency": "5.274", + "cost_per_test": "0.001275", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.01, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.973, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.973", + "raw_stderr": "1.115", + "latency": "5.683", + "cost_per_test": "0.001306", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.115, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.046, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.046", + "raw_stderr": "0.928", + "latency": "9.256", + "cost_per_test": "0.002146", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.928, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.317, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.317", + "raw_stderr": "1.643", + "latency": "7.075", + "cost_per_test": "0.001549", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.643, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.441, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.441", + "raw_stderr": "1.141", + "latency": "5.184", + "cost_per_test": "0.001202", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.141, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.471, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.471", + "raw_stderr": "1.363", + "latency": "16.513", + "cost_per_test": "0.00273", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.363, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.628, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.628", + "raw_stderr": "1.457", + "latency": "4.928", + "cost_per_test": "0.001017", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.457, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.803, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.803", + "raw_stderr": "2.224", + "latency": "5.53", + "cost_per_test": "0.001343", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.224, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 69.846, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.846", + "raw_stderr": "1.383", + "latency": "29.965", + "cost_per_test": "0.001465", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.383, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.116, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.116", + "raw_stderr": "0.689", + "latency": "7.385", + "cost_per_test": "0.00183", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.689, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.303, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.303", + "raw_stderr": "1.308", + "latency": "4.209", + "cost_per_test": "0.000805", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.308, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.686, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.686", + "raw_stderr": "0.362", + "latency": "8.525", + "cost_per_test": "0.001454", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.362, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.359, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.359", + "raw_stderr": "1.812", + "latency": "5.223", + "cost_per_test": "0.000995", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.812, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.53, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.53", + "raw_stderr": "0.849", + "latency": "8.471", + "cost_per_test": "0.001752", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.849, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.714, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.714", + "raw_stderr": "1.239", + "latency": "4.655", + "cost_per_test": "0.000934", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.239, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-preview-09-2025/799f936f-da77-4f6f-9c61-7aed9903410e.json b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025/799f936f-da77-4f6f-9c61-7aed9903410e.json new file mode 100644 index 0000000000000000000000000000000000000000..9626337bfb402bea0e6fd757a80e1b12b8b9c5c3 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025/799f936f-da77-4f6f-9c61-7aed9903410e.json @@ -0,0 +1,346 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/google_gemini-2.5-flash-preview-09-2025/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-preview-09-2025", + "id": "google/gemini-2.5-flash-preview-09-2025", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-preview-09-2025", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 39.683, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "39.683", + "raw_stderr": "6.139", + "latency": "21.605", + "cost_per_test": "0.002255", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.139, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 35.318, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "35.318", + "raw_stderr": "6.118", + "latency": "34.456", + "cost_per_test": "0.002213", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.118, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 55.556, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "55.556", + "raw_stderr": "6.395", + "latency": "49.598", + "cost_per_test": "0.005121", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.395, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 57.316, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "57.316", + "raw_stderr": "5.556", + "latency": "38.478", + "cost_per_test": "0.002481", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.556, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 44.106, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "44.106", + "raw_stderr": "3.432", + "latency": "31.513", + "cost_per_test": "0.002316", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.432, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-preview-09-2025/865530e4-843d-4d04-9823-1b3d4314a599.json b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025/865530e4-843d-4d04-9823-1b3d4314a599.json new file mode 100644 index 0000000000000000000000000000000000000000..a981570f41d36512f262e12a977a7ed7cfca65a9 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025/865530e4-843d-4d04-9823-1b3d4314a599.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/google_gemini-2.5-flash-preview-09-2025/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-preview-09-2025", + "id": "google/gemini-2.5-flash-preview-09-2025", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-preview-09-2025", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 56.177, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.177", + "raw_stderr": "1.694", + "latency": "57.467", + "cost_per_test": "0.000911", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.694, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.49, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.49", + "raw_stderr": "1.669", + "latency": "69.3", + "cost_per_test": "0.056875", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.669, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 58.974, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.974", + "raw_stderr": "0.969", + "latency": "51.895", + "cost_per_test": "0.026123", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.969, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.256, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.256", + "raw_stderr": "1.671", + "latency": "28.917", + "cost_per_test": "0.020583", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.671, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-preview-09-2025/ad7e5f28-3f6b-454a-8b4e-c6419f2c1d34.json b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025/ad7e5f28-3f6b-454a-8b4e-c6419f2c1d34.json new file mode 100644 index 0000000000000000000000000000000000000000..a62d2d6f83374b0eaeedbbb41d2c0b440c960daa --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025/ad7e5f28-3f6b-454a-8b4e-c6419f2c1d34.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/google_gemini-2.5-flash-preview-09-2025/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-preview-09-2025", + "id": "google/gemini-2.5-flash-preview-09-2025", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-preview-09-2025", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 69.157, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "69.157", + "raw_stderr": "1.302", + "latency": "9.224", + "cost_per_test": "0.001183", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.302, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 62.599, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.599", + "raw_stderr": "0.956", + "latency": "7.274", + "cost_per_test": "0.001152", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.956, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 56.041, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.041", + "raw_stderr": "1.399", + "latency": "5.325", + "cost_per_test": "0.001122", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.399, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-preview-09-2025/b92d1d8d-0d18-42da-98cd-c5f63412c037.json b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025/b92d1d8d-0d18-42da-98cd-c5f63412c037.json new file mode 100644 index 0000000000000000000000000000000000000000..afc613e48064a0a45f1ce34fd305d289062404db --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025/b92d1d8d-0d18-42da-98cd-c5f63412c037.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/google_gemini-2.5-flash-preview-09-2025/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-preview-09-2025", + "id": "google/gemini-2.5-flash-preview-09-2025", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-preview-09-2025", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 86.519, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.519", + "raw_stderr": "0.91", + "latency": "0.873", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.91, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.385, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.385", + "raw_stderr": "0.193", + "latency": "1.309", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.193, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 86.31, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.31", + "raw_stderr": "0.609", + "latency": "1.015", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.609, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.486, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.486", + "raw_stderr": "0.431", + "latency": "1.154", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.431, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.031, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.031", + "raw_stderr": "0.795", + "latency": "0.841", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.795, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.183, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.183", + "raw_stderr": "1.664", + "latency": "1.732", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.664, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-preview-09-2025/d26f6fa4-be79-4c20-9eba-8af3d782e6c9.json b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025/d26f6fa4-be79-4c20-9eba-8af3d782e6c9.json new file mode 100644 index 0000000000000000000000000000000000000000..e8ae35ec0fbec023e6fe753cdbb8776c5fa479c9 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025/d26f6fa4-be79-4c20-9eba-8af3d782e6c9.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/google_gemini-2.5-flash-preview-09-2025/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-preview-09-2025", + "id": "google/gemini-2.5-flash-preview-09-2025", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-preview-09-2025", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 60.998, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.998", + "raw_stderr": "1.395", + "latency": "6.267", + "cost_per_test": "0.003098", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.395, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 72.731, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "72.731", + "raw_stderr": "0.869", + "latency": "6.267", + "cost_per_test": "0.003098", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.869, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 84.464, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "84.464", + "raw_stderr": "1.036", + "latency": "6.267", + "cost_per_test": "0.003098", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.036, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-preview-09-2025/e6c78181-8ce3-4317-954e-cdeec6652641.json b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025/e6c78181-8ce3-4317-954e-cdeec6652641.json new file mode 100644 index 0000000000000000000000000000000000000000..c4aafe146266be4fb8311612fbeb21d0645ac96f --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025/e6c78181-8ce3-4317-954e-cdeec6652641.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/google_gemini-2.5-flash-preview-09-2025/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-preview-09-2025", + "id": "google/gemini-2.5-flash-preview-09-2025", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-preview-09-2025", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 82.323, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.323", + "raw_stderr": "2.711", + "latency": "19.829", + "cost_per_test": "0.002954", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.711, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 80.808, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.808", + "raw_stderr": "2.93", + "latency": "19.137", + "cost_per_test": "0.002833", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.93, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 81.566, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.566", + "raw_stderr": "1.996", + "latency": "19.483", + "cost_per_test": "0.002894", + "temperature": "1", + "max_output_tokens": "65535", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.996, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-preview-09-2025/f4430314-31fb-435a-a75a-05f7ff5fb827.json b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025/f4430314-31fb-435a-a75a-05f7ff5fb827.json new file mode 100644 index 0000000000000000000000000000000000000000..9673a278d82467d58353cdca3416ea13be1ac495 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-preview-09-2025/f4430314-31fb-435a-a75a-05f7ff5fb827.json @@ -0,0 +1,101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/google_gemini-2.5-flash-preview-09-2025/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-preview-09-2025", + "id": "google/gemini-2.5-flash-preview-09-2025", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-preview-09-2025", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:google/gemini-2.5-flash-preview-09-2025:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 79.48, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.48", + "raw_stderr": "0.971", + "latency": "25.643", + "cost_per_test": "0.00187", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.971, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-thinking/038fbc16-f9e4-44f8-9d58-ca1d58598b90.json b/data/vals_ai/google/gemini-2.5-flash-thinking/038fbc16-f9e4-44f8-9d58-ca1d58598b90.json new file mode 100644 index 0000000000000000000000000000000000000000..271d7db714d54dcb58226fb48c8a07a7a86615b1 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-thinking/038fbc16-f9e4-44f8-9d58-ca1d58598b90.json @@ -0,0 +1,346 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/google_gemini-2.5-flash-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-thinking", + "id": "google/gemini-2.5-flash-thinking", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-thinking", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:google/gemini-2.5-flash-thinking:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 31.746, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "31.746", + "raw_stderr": "5.841", + "latency": "37.231", + "cost_per_test": "0.002985", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.841, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:google/gemini-2.5-flash-thinking:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 42.064, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "42.064", + "raw_stderr": "6.329", + "latency": "39.172", + "cost_per_test": "0.002966", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.329, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:google/gemini-2.5-flash-thinking:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 51.852, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "51.852", + "raw_stderr": "6.386", + "latency": "41.551", + "cost_per_test": "0.005404", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.386, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:google/gemini-2.5-flash-thinking:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 44.997, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "44.997", + "raw_stderr": "5.588", + "latency": "42.446", + "cost_per_test": "0.002638", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.588, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:google/gemini-2.5-flash-thinking:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 39.602, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "39.602", + "raw_stderr": "3.422", + "latency": "39.616", + "cost_per_test": "0.002863", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.422, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash-thinking/2821711d-967b-40c3-8c37-701d62b96817.json b/data/vals_ai/google/gemini-2.5-flash-thinking/2821711d-967b-40c3-8c37-701d62b96817.json new file mode 100644 index 0000000000000000000000000000000000000000..74d8575fa5a2d016e457e14f95b61d561e1b1e94 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash-thinking/2821711d-967b-40c3-8c37-701d62b96817.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/google_gemini-2.5-flash-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash-thinking", + "id": "google/gemini-2.5-flash-thinking", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash-thinking", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:google/gemini-2.5-flash-thinking:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 82.983, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "82.983", + "raw_stderr": "1.908", + "latency": "22.529", + "cost_per_test": "0.014824", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.908, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash/6157440d-4803-4ab0-8c8c-db5c191996fa.json b/data/vals_ai/google/gemini-2.5-flash/6157440d-4803-4ab0-8c8c-db5c191996fa.json new file mode 100644 index 0000000000000000000000000000000000000000..0b9af446f26f3fdbffbce67947dec8c9247ac0ea --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash/6157440d-4803-4ab0-8c8c-db5c191996fa.json @@ -0,0 +1,346 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/google_gemini-2.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash", + "id": "google/gemini-2.5-flash", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:google/gemini-2.5-flash:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 31.746, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "31.746", + "raw_stderr": "5.841", + "latency": "44.271", + "cost_per_test": "0.00286", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.841, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:google/gemini-2.5-flash:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 44.048, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "44.048", + "raw_stderr": "6.355", + "latency": "47.36", + "cost_per_test": "0.002682", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.355, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:google/gemini-2.5-flash:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 55.864, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "55.864", + "raw_stderr": "6.148", + "latency": "35.521", + "cost_per_test": "0.005276", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.148, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:google/gemini-2.5-flash:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 58.473, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "58.473", + "raw_stderr": "5.535", + "latency": "29.009", + "cost_per_test": "0.002669", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:google/gemini-2.5-flash:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 44.756, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "44.756", + "raw_stderr": "3.418", + "latency": "40.213", + "cost_per_test": "0.002737", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.418, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash/bf48cab9-1b02-4a06-8543-126305defd04.json b/data/vals_ai/google/gemini-2.5-flash/bf48cab9-1b02-4a06-8543-126305defd04.json new file mode 100644 index 0000000000000000000000000000000000000000..7c6ebeb277da29861726ca185be4501137fb6f74 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash/bf48cab9-1b02-4a06-8543-126305defd04.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/google_gemini-2.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash", + "id": "google/gemini-2.5-flash", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:google/gemini-2.5-flash:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 5.222, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "5.222", + "raw_stderr": "2.059", + "latency": "623.818", + "cost_per_test": "0.07068", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.059, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:google/gemini-2.5-flash:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "76.555", + "cost_per_test": "0.003315", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:google/gemini-2.5-flash:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 2.611, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "2.611", + "raw_stderr": "1.03", + "latency": "350.186", + "cost_per_test": "0.036998", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.03, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-flash/d6ac11f8-1faf-4dc4-8d49-f256b8b2e16a.json b/data/vals_ai/google/gemini-2.5-flash/d6ac11f8-1faf-4dc4-8d49-f256b8b2e16a.json new file mode 100644 index 0000000000000000000000000000000000000000..5901bae36d024535a0ec31993f9c1a185963fc8b --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-flash/d6ac11f8-1faf-4dc4-8d49-f256b8b2e16a.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/google_gemini-2.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-flash", + "id": "google/gemini-2.5-flash", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-flash", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:google/gemini-2.5-flash:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 82.869, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "82.869", + "raw_stderr": "1.909", + "latency": "22.786", + "cost_per_test": "0.014869", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.909, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-pro-exp-03-25/0bcb32b9-bbd7-4c7d-873c-a69cb9d6ff22.json b/data/vals_ai/google/gemini-2.5-pro-exp-03-25/0bcb32b9-bbd7-4c7d-873c-a69cb9d6ff22.json new file mode 100644 index 0000000000000000000000000000000000000000..f624806238b113c42876e19d6a52427af60d1827 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-pro-exp-03-25/0bcb32b9-bbd7-4c7d-873c-a69cb9d6ff22.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/google_gemini-2.5-pro-exp-03-25/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-pro-exp-03-25", + "id": "google/gemini-2.5-pro-exp-03-25", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-pro-exp-03-25", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 80.303, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.303", + "raw_stderr": "2.853", + "latency": "39.776", + "cost_per_test": "0.014206", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.853, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 81.313, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.313", + "raw_stderr": "2.799", + "latency": "42.429", + "cost_per_test": "0.011682", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.799, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 80.808, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.808", + "raw_stderr": "1.998", + "latency": "41.102", + "cost_per_test": "0.012944", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.998, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-pro-exp-03-25/2f41b9f6-f3b3-458f-a9c4-f9599c40b993.json b/data/vals_ai/google/gemini-2.5-pro-exp-03-25/2f41b9f6-f3b3-458f-a9c4-f9599c40b993.json new file mode 100644 index 0000000000000000000000000000000000000000..1d080aa1d7581044965c2a2e67722bd9be9f72ec --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-pro-exp-03-25/2f41b9f6-f3b3-458f-a9c4-f9599c40b993.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/google_gemini-2.5-pro-exp-03-25/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-pro-exp-03-25", + "id": "google/gemini-2.5-pro-exp-03-25", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-pro-exp-03-25", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 61.57, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.57", + "raw_stderr": "1.391", + "latency": "21.775", + "cost_per_test": "0.010628", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.391, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 72.894, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "72.894", + "raw_stderr": "0.869", + "latency": "21.775", + "cost_per_test": "0.010628", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.869, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 84.219, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "84.219", + "raw_stderr": "1.042", + "latency": "21.775", + "cost_per_test": "0.010628", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.042, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-pro-exp-03-25/6d43377a-2922-4c10-9bcb-d7d2038d289c.json b/data/vals_ai/google/gemini-2.5-pro-exp-03-25/6d43377a-2922-4c10-9bcb-d7d2038d289c.json new file mode 100644 index 0000000000000000000000000000000000000000..acda4226a21ff320c0eff30751f38282df823842 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-pro-exp-03-25/6d43377a-2922-4c10-9bcb-d7d2038d289c.json @@ -0,0 +1,101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/google_gemini-2.5-pro-exp-03-25/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-pro-exp-03-25", + "id": "google/gemini-2.5-pro-exp-03-25", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-pro-exp-03-25", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 81.34, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.34", + "raw_stderr": "0.936", + "latency": "24.145", + "cost_per_test": "0.007771", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.936, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-pro-exp-03-25/896fce3d-adb4-4b5c-b646-a256357455d6.json b/data/vals_ai/google/gemini-2.5-pro-exp-03-25/896fce3d-adb4-4b5c-b646-a256357455d6.json new file mode 100644 index 0000000000000000000000000000000000000000..81ad1e56fd76aae8d470cf29ed6479d50f759681 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-pro-exp-03-25/896fce3d-adb4-4b5c-b646-a256357455d6.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/google_gemini-2.5-pro-exp-03-25/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-pro-exp-03-25", + "id": "google/gemini-2.5-pro-exp-03-25", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-pro-exp-03-25", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 88.312, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.312", + "raw_stderr": "0.806", + "latency": "4.261", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.806, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.736, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.736", + "raw_stderr": "0.186", + "latency": "3.486", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.186, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.099, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.099", + "raw_stderr": "0.585", + "latency": "2.709", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.585, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 84.322, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.322", + "raw_stderr": "0.408", + "latency": "4.022", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.408, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.094, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.094", + "raw_stderr": "0.871", + "latency": "2.732", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.871, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.37, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.37", + "raw_stderr": "1.544", + "latency": "6.92", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.544, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-pro-exp-03-25/9d89ce1d-ef4d-471d-8b98-b7a99f1954ff.json b/data/vals_ai/google/gemini-2.5-pro-exp-03-25/9d89ce1d-ef4d-471d-8b98-b7a99f1954ff.json new file mode 100644 index 0000000000000000000000000000000000000000..97a6246f237c1c8ac2099aa1ae105752fecbc2fa --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-pro-exp-03-25/9d89ce1d-ef4d-471d-8b98-b7a99f1954ff.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/google_gemini-2.5-pro-exp-03-25/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-pro-exp-03-25", + "id": "google/gemini-2.5-pro-exp-03-25", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-pro-exp-03-25", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 59.907, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.907", + "raw_stderr": "1.673", + "latency": "16.774", + "cost_per_test": "0.003881", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.673, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 57.809, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.809", + "raw_stderr": "1.686", + "latency": "11.897", + "cost_per_test": "0.169274", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.686, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 59.829, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.829", + "raw_stderr": "0.966", + "latency": "16.302", + "cost_per_test": "0.08579", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.966, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.772, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.772", + "raw_stderr": "1.659", + "latency": "20.235", + "cost_per_test": "0.084215", + "temperature": "1", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.659, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-pro-exp-03-25/da3d15fe-8442-47fd-9c64-68811bbb61d7.json b/data/vals_ai/google/gemini-2.5-pro-exp-03-25/da3d15fe-8442-47fd-9c64-68811bbb61d7.json new file mode 100644 index 0000000000000000000000000000000000000000..bf6a979c75b4002292bbde525e264be0820af365 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-pro-exp-03-25/da3d15fe-8442-47fd-9c64-68811bbb61d7.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/google_gemini-2.5-pro-exp-03-25/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-pro-exp-03-25", + "id": "google/gemini-2.5-pro-exp-03-25", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-pro-exp-03-25", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 74.642, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "74.642", + "raw_stderr": "1.227", + "latency": "10.727", + "cost_per_test": "0.00346", + "temperature": "0.1", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.227, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.1, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 67.17, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.17", + "raw_stderr": "0.924", + "latency": "8.91", + "cost_per_test": "0.003638", + "temperature": "0.1", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.924, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.1, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 59.698, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.698", + "raw_stderr": "1.383", + "latency": "7.092", + "cost_per_test": "0.003816", + "temperature": "0.1", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.383, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.1, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-pro-exp-03-25/f4872075-dfdc-45d1-8f19-c0d96825ea1f.json b/data/vals_ai/google/gemini-2.5-pro-exp-03-25/f4872075-dfdc-45d1-8f19-c0d96825ea1f.json new file mode 100644 index 0000000000000000000000000000000000000000..0b01f25b112f9b4d63d5f80ae949ee1e4bedfaf7 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-pro-exp-03-25/f4872075-dfdc-45d1-8f19-c0d96825ea1f.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/google_gemini-2.5-pro-exp-03-25/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-pro-exp-03-25", + "id": "google/gemini-2.5-pro-exp-03-25", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-pro-exp-03-25", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.608, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.608", + "raw_stderr": "0.977", + "latency": "14.363", + "cost_per_test": "0.0082", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.977, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.1, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.1", + "raw_stderr": "1.109", + "latency": "14.635", + "cost_per_test": "0.007486", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.109, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.548, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.548", + "raw_stderr": "0.87", + "latency": "22.87", + "cost_per_test": "0.011033", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.87, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.854, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.854", + "raw_stderr": "1.721", + "latency": "19.727", + "cost_per_test": "0.009532", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.721, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.796, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.796", + "raw_stderr": "1.127", + "latency": "13.595", + "cost_per_test": "0.007895", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.127, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.773, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.773", + "raw_stderr": "1.29", + "latency": "38.297", + "cost_per_test": "0.013937", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.29, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.951, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.951", + "raw_stderr": "1.4", + "latency": "15.379", + "cost_per_test": "0.007625", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.4, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.378, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.378", + "raw_stderr": "2.176", + "latency": "15.675", + "cost_per_test": "0.009111", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.176, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.026, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.026", + "raw_stderr": "1.367", + "latency": "18.956", + "cost_per_test": "0.009806", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.367, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.79, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.79", + "raw_stderr": "1.003", + "latency": "22.856", + "cost_per_test": "0.011456", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.003, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.576, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.576", + "raw_stderr": "1.248", + "latency": "12.065", + "cost_per_test": "0.005927", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.248, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.058, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.058", + "raw_stderr": "0.358", + "latency": "18.46", + "cost_per_test": "0.009006", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.358, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.565, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.565", + "raw_stderr": "1.698", + "latency": "15.171", + "cost_per_test": "0.00737", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.698, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.761, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.761", + "raw_stderr": "0.841", + "latency": "21.681", + "cost_per_test": "0.009687", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.841, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:google/gemini-2.5-pro-exp-03-25:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.088, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.088", + "raw_stderr": "1.261", + "latency": "13.171", + "cost_per_test": "0.007016", + "temperature": "0.7", + "max_output_tokens": "64000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.261, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-pro-preview-03-25/f91cf506-5103-41ad-bf9c-738b0c286ea5.json b/data/vals_ai/google/gemini-2.5-pro-preview-03-25/f91cf506-5103-41ad-bf9c-738b0c286ea5.json new file mode 100644 index 0000000000000000000000000000000000000000..1e24d62e470a719669b7ed66a6fa7bb7f37d5947 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-pro-preview-03-25/f91cf506-5103-41ad-bf9c-738b0c286ea5.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/google_gemini-2.5-pro-preview-03-25/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-pro-preview-03-25", + "id": "google/gemini-2.5-pro-preview-03-25", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-pro-preview-03-25", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:google/gemini-2.5-pro-preview-03-25:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.758, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.758", + "raw_stderr": "0.617", + "latency": "96.699", + "cost_per_test": "0.012934", + "temperature": "1", + "max_output_tokens": "32512", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.617, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32512, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:google/gemini-2.5-pro-preview-03-25:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 52.571, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "52.571", + "raw_stderr": "2.669", + "latency": "228.861", + "cost_per_test": "0.02524", + "temperature": "1", + "max_output_tokens": "32512", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.669, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32512, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:google/gemini-2.5-pro-preview-03-25:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 86.162, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "86.162", + "raw_stderr": "1.764", + "latency": "168.41", + "cost_per_test": "0.021947", + "temperature": "1", + "max_output_tokens": "32512", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.764, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32512, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:google/gemini-2.5-pro-preview-03-25:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 79.164, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "79.164", + "raw_stderr": "1.086", + "latency": "164.657", + "cost_per_test": "0.02004", + "temperature": "1", + "max_output_tokens": "32512", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.086, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32512, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-pro/2d3a5143-59a9-46aa-8e6c-5466fb431bf5.json b/data/vals_ai/google/gemini-2.5-pro/2d3a5143-59a9-46aa-8e6c-5466fb431bf5.json new file mode 100644 index 0000000000000000000000000000000000000000..d2aa12f91b6dbd8e41f123a09f1640f593c9d5f4 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-pro/2d3a5143-59a9-46aa-8e6c-5466fb431bf5.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/google_gemini-2.5-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-pro", + "id": "google/gemini-2.5-pro", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-pro", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:google/gemini-2.5-pro:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 73.552, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "73.552", + "raw_stderr": "1.91", + "latency": "35.909", + "cost_per_test": "0.046379", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.91, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-pro/41ec4788-e9b0-4518-bef5-c9dc926a6b12.json b/data/vals_ai/google/gemini-2.5-pro/41ec4788-e9b0-4518-bef5-c9dc926a6b12.json new file mode 100644 index 0000000000000000000000000000000000000000..6f625a20a41a6e099d796bf82e6ef28c29d5f413 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-pro/41ec4788-e9b0-4518-bef5-c9dc926a6b12.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/google_gemini-2.5-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-pro", + "id": "google/gemini-2.5-pro", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-pro", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:google/gemini-2.5-pro:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 19.0, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "19", + "raw_stderr": "10.263", + "latency": "2159.535", + "cost_per_test": "0.842034", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 10.263, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:google/gemini-2.5-pro:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 15.167, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "15.167", + "raw_stderr": "6.853", + "latency": "3770.063", + "cost_per_test": "0.388888", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.853, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:google/gemini-2.5-pro:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 17.084, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "17.084", + "raw_stderr": "6.17", + "latency": "2964.799", + "cost_per_test": "0.615461", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.17, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-pro/4b4b7db4-d0c3-4297-8968-41686bbd9908.json b/data/vals_ai/google/gemini-2.5-pro/4b4b7db4-d0c3-4297-8968-41686bbd9908.json new file mode 100644 index 0000000000000000000000000000000000000000..a3eb4328249fef8ea0982f1cf075e21a4c52ed85 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-pro/4b4b7db4-d0c3-4297-8968-41686bbd9908.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/google_gemini-2.5-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-pro", + "id": "google/gemini-2.5-pro", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-pro", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:google/gemini-2.5-pro:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 62.937, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.937", + "raw_stderr": "1.649", + "latency": "32.815", + "cost_per_test": "0.201351", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.649, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:google/gemini-2.5-pro:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 56.876, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.876", + "raw_stderr": "1.691", + "latency": "20.714", + "cost_per_test": "0.003627", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.691, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:google/gemini-2.5-pro:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.8, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.8", + "raw_stderr": "0.961", + "latency": "27.425", + "cost_per_test": "0.09535", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.961, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:google/gemini-2.5-pro:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 62.587, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.587", + "raw_stderr": "1.652", + "latency": "28.746", + "cost_per_test": "0.081072", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.652, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-pro/7646b1c7-8fda-49a2-81b8-b18016ca631c.json b/data/vals_ai/google/gemini-2.5-pro/7646b1c7-8fda-49a2-81b8-b18016ca631c.json new file mode 100644 index 0000000000000000000000000000000000000000..c284e4f942244e5741eea5356b0eaca8e6de1d6e --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-pro/7646b1c7-8fda-49a2-81b8-b18016ca631c.json @@ -0,0 +1,361 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/google_gemini-2.5-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-pro", + "id": "google/gemini-2.5-pro", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-pro", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:google/gemini-2.5-pro:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 14.286, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "14.286", + "raw_stderr": "5.465", + "latency": "274.125", + "cost_per_test": "0.462942", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.465, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:google/gemini-2.5-pro:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 47.51, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "47.51", + "raw_stderr": "3.097", + "latency": "230.32", + "cost_per_test": "0.387754", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.097, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:google/gemini-2.5-pro:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 73.196, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "73.196", + "raw_stderr": "3.188", + "latency": "158.803", + "cost_per_test": "0.265263", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.188, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:google/gemini-2.5-pro:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "486.969", + "cost_per_test": "1.226948", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:google/gemini-2.5-pro:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 54.4, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "54.4", + "raw_stderr": "2.23", + "latency": "207.791", + "cost_per_test": "0.351578", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.23, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-pro/8242315e-e693-4537-9bb3-560299c0ef8f.json b/data/vals_ai/google/gemini-2.5-pro/8242315e-e693-4537-9bb3-560299c0ef8f.json new file mode 100644 index 0000000000000000000000000000000000000000..f373b1d3211ae16f7e8766ef68cd77f2f23b2587 --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-pro/8242315e-e693-4537-9bb3-560299c0ef8f.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/google_gemini-2.5-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-pro", + "id": "google/gemini-2.5-pro", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-pro", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:google/gemini-2.5-pro:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 77.583, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "77.583", + "raw_stderr": "1.176", + "latency": "28.249", + "cost_per_test": "0.004647", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.176, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:google/gemini-2.5-pro:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 68.918, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.918", + "raw_stderr": "0.907", + "latency": "24.008", + "cost_per_test": "0.004794", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.907, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:google/gemini-2.5-pro:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 60.254, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.254", + "raw_stderr": "1.38", + "latency": "19.767", + "cost_per_test": "0.004942", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.38, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-pro/c38d5ff1-9d97-48b3-a8b8-b9f729cefc79.json b/data/vals_ai/google/gemini-2.5-pro/c38d5ff1-9d97-48b3-a8b8-b9f729cefc79.json new file mode 100644 index 0000000000000000000000000000000000000000..0bdd0548494e968ec7f80c6c2d37a10f0ede407e --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-pro/c38d5ff1-9d97-48b3-a8b8-b9f729cefc79.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/google_gemini-2.5-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-pro", + "id": "google/gemini-2.5-pro", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-pro", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:google/gemini-2.5-pro:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 0.4, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "0.4", + "raw_stderr": "0.4", + "latency": "2097.144", + "cost_per_test": "1.223831", + "temperature": "1", + "max_output_tokens": "65536", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.4, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-2.5-pro/feb69821-3bc5-4aaf-9f4f-ec957d824fe3.json b/data/vals_ai/google/gemini-2.5-pro/feb69821-3bc5-4aaf-9f4f-ec957d824fe3.json new file mode 100644 index 0000000000000000000000000000000000000000..32b380382811e0e65190fda70573b41637f4508f --- /dev/null +++ b/data/vals_ai/google/gemini-2.5-pro/feb69821-3bc5-4aaf-9f4f-ec957d824fe3.json @@ -0,0 +1,346 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/google_gemini-2.5-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-2.5-pro", + "id": "google/gemini-2.5-pro", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-2.5-pro", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:google/gemini-2.5-pro:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 30.159, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "30.159", + "raw_stderr": "5.759", + "latency": "59.266", + "cost_per_test": "0.00787", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.759, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:google/gemini-2.5-pro:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 43.651, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "43.651", + "raw_stderr": "6.33", + "latency": "59.764", + "cost_per_test": "0.009334", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.33, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:google/gemini-2.5-pro:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 70.679, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "70.679", + "raw_stderr": "5.808", + "latency": "62.783", + "cost_per_test": "0.019416", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.808, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:google/gemini-2.5-pro:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 51.938, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "51.938", + "raw_stderr": "5.612", + "latency": "39.772", + "cost_per_test": "0.009215", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.612, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:google/gemini-2.5-pro:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 41.916, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "41.916", + "raw_stderr": "3.411", + "latency": "52.934", + "cost_per_test": "0.008806", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.411, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-flash-preview/081c42c9-f350-4960-a21d-a4584e8d750f.json b/data/vals_ai/google/gemini-3-flash-preview/081c42c9-f350-4960-a21d-a4584e8d750f.json new file mode 100644 index 0000000000000000000000000000000000000000..2dc2e7c46527d00c6ae66efd097901a9a79799ec --- /dev/null +++ b/data/vals_ai/google/gemini-3-flash-preview/081c42c9-f350-4960-a21d-a4584e8d750f.json @@ -0,0 +1,418 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/google_gemini-3-flash-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-flash-preview", + "id": "google/gemini-3-flash-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-flash-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 67.716, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "67.716", + "raw_stderr": "1.596", + "latency": "14.833", + "cost_per_test": "0.039017", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.596, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 42.476, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "42.476", + "raw_stderr": "0.666", + "latency": "241.608", + "cost_per_test": "0.543671", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.666, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:overall:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 49.546, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "49.546", + "raw_stderr": "1.545", + "latency": "353.286", + "cost_per_test": "0.397051", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.545, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 68.627, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "68.627", + "raw_stderr": "4.594", + "latency": "327.172", + "cost_per_test": "0.306186" + }, + "uncertainty": { + "standard_error": { + "value": 4.594, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 53.933, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "53.933", + "raw_stderr": "1.297", + "latency": "376.158", + "cost_per_test": "0.154312", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.297, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 21.954, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "21.954", + "raw_stderr": "6.673", + "latency": "806.66", + "cost_per_test": "0.942067", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.673, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-flash-preview/2eb14436-0660-4c73-a897-83bb6e01b5a4.json b/data/vals_ai/google/gemini-3-flash-preview/2eb14436-0660-4c73-a897-83bb6e01b5a4.json new file mode 100644 index 0000000000000000000000000000000000000000..263f8881d7896f697c58ef1f663062ec8fe88177 --- /dev/null +++ b/data/vals_ai/google/gemini-3-flash-preview/2eb14436-0660-4c73-a897-83bb6e01b5a4.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/google_gemini-3-flash-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-flash-preview", + "id": "google/gemini-3-flash-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-flash-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 99.068, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "99.068", + "raw_stderr": "0.535", + "latency": "47.032", + "cost_per_test": "0.02567", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:hard:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 69.714, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "69.714", + "raw_stderr": "2.456", + "latency": "136.003", + "cost_per_test": "0.074818", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.456, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:medium:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 87.99, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "87.99", + "raw_stderr": "1.661", + "latency": "92.758", + "cost_per_test": "0.051088", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.661, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:overall:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 85.591, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "85.591", + "raw_stderr": "1.004", + "latency": "91.931", + "cost_per_test": "0.050525", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.004, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-flash-preview/2eef6868-4e8c-4ab0-8d75-dfe2124db642.json b/data/vals_ai/google/gemini-3-flash-preview/2eef6868-4e8c-4ab0-8d75-dfe2124db642.json new file mode 100644 index 0000000000000000000000000000000000000000..ee3e30077815faf4ee1ebab6d6c1dc3d2737cf1e --- /dev/null +++ b/data/vals_ai/google/gemini-3-flash-preview/2eef6868-4e8c-4ab0-8d75-dfe2124db642.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/google_gemini-3-flash-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-flash-preview", + "id": "google/gemini-3-flash-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-flash-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 41.0, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "41", + "raw_stderr": "17.098", + "latency": "2307.779", + "cost_per_test": "0.401804", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 17.098, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 37.167, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "37.167", + "raw_stderr": "12.346", + "latency": "1385.274", + "cost_per_test": "0.3483", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 12.346, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "ioi:overall:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 39.084, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "39.084", + "raw_stderr": "10.545", + "latency": "1846.526", + "cost_per_test": "0.375052", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 10.545, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-flash-preview/3f5b9209-1b86-4236-a2ca-5b734434254b.json b/data/vals_ai/google/gemini-3-flash-preview/3f5b9209-1b86-4236-a2ca-5b734434254b.json new file mode 100644 index 0000000000000000000000000000000000000000..9b80c3927f9d333d9ac14e70e3d320bf6b699928 --- /dev/null +++ b/data/vals_ai/google/gemini-3-flash-preview/3f5b9209-1b86-4236-a2ca-5b734434254b.json @@ -0,0 +1,450 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/google_gemini-3-flash-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-flash-preview", + "id": "google/gemini-3-flash-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-flash-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 90.077, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.077", + "raw_stderr": "0.613", + "latency": "4.23", + "cost_per_test": "0.001736", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.613, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.908, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.908", + "raw_stderr": "0.302", + "latency": "5.669", + "cost_per_test": "0.002582", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.302, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 91.145, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "91.145", + "raw_stderr": "0.537", + "latency": "4.207", + "cost_per_test": "0.002102", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.537, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 86.858, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.858", + "raw_stderr": "0.356", + "latency": "4.673", + "cost_per_test": "0.002067", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.356, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.87, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.87", + "raw_stderr": "0.882", + "latency": "4.036", + "cost_per_test": "0.001805", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.882, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 85.291, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.291", + "raw_stderr": "1.281", + "latency": "5.224", + "cost_per_test": "0.002112", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.281, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-flash-preview/4355630c-a5d0-4681-a48c-1cc13cd723e4.json b/data/vals_ai/google/gemini-3-flash-preview/4355630c-a5d0-4681-a48c-1cc13cd723e4.json new file mode 100644 index 0000000000000000000000000000000000000000..c25f47838735b72adb0d844e5d57ea496973c37d --- /dev/null +++ b/data/vals_ai/google/gemini-3-flash-preview/4355630c-a5d0-4681-a48c-1cc13cd723e4.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/google_gemini-3-flash-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-flash-preview", + "id": "google/gemini-3-flash-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-flash-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 62.306, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.306", + "raw_stderr": "1.386", + "latency": "12.001", + "cost_per_test": "0.005764", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.386, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 73.876, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "73.876", + "raw_stderr": "0.857", + "latency": "12.001", + "cost_per_test": "0.005764", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.857, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 85.446, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "85.446", + "raw_stderr": "1.008", + "latency": "12.001", + "cost_per_test": "0.005764", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.008, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-flash-preview/56d61231-9d73-48e3-8e49-57edb2160d06.json b/data/vals_ai/google/gemini-3-flash-preview/56d61231-9d73-48e3-8e49-57edb2160d06.json new file mode 100644 index 0000000000000000000000000000000000000000..7a4db7ba7753f5043b09860e567f94b4904a09ab --- /dev/null +++ b/data/vals_ai/google/gemini-3-flash-preview/56d61231-9d73-48e3-8e49-57edb2160d06.json @@ -0,0 +1,366 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/google_gemini-3-flash-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-flash-preview", + "id": "google/gemini-3-flash-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-flash-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 41.27, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "41.27", + "raw_stderr": "6.178", + "latency": "47.126", + "cost_per_test": "0.020685", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.178, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:automata:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 43.254, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "43.254", + "raw_stderr": "6.336", + "latency": "48.336", + "cost_per_test": "0.019315", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.336, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:calculus:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 70.988, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "70.988", + "raw_stderr": "5.5", + "latency": "60.109", + "cost_per_test": "0.014869", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.5, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 71.024, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "71.024", + "raw_stderr": "5.096", + "latency": "32.629", + "cost_per_test": "0.014687", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.096, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:overall:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 51.849, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "51.849", + "raw_stderr": "3.404", + "latency": "42.697", + "cost_per_test": "0.018229", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.404, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-flash-preview/58cd8a5d-91ab-4cbe-93f4-355a62e24aef.json b/data/vals_ai/google/gemini-3-flash-preview/58cd8a5d-91ab-4cbe-93f4-355a62e24aef.json new file mode 100644 index 0000000000000000000000000000000000000000..b321c51e11dd1fcd125b720308a9bcdd5ae241ef --- /dev/null +++ b/data/vals_ai/google/gemini-3-flash-preview/58cd8a5d-91ab-4cbe-93f4-355a62e24aef.json @@ -0,0 +1,381 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/google_gemini-3-flash-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-flash-preview", + "id": "google/gemini-3-flash-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-flash-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 38.095, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "38.095", + "raw_stderr": "7.584", + "latency": "375.223", + "cost_per_test": "0.356393", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 7.584, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 72.031, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "72.031", + "raw_stderr": "2.784", + "latency": "288.103", + "cost_per_test": "0.287155", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.784, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 87.629, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "87.629", + "raw_stderr": "2.37", + "latency": "260.616", + "cost_per_test": "0.253349", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.37, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "365.995", + "cost_per_test": "0.349943", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:overall:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 75.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "75", + "raw_stderr": "1.938", + "latency": "285.223", + "cost_per_test": "0.280231", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.938, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-flash-preview/63a96472-79e3-4fb5-bd31-1875ace0ac65.json b/data/vals_ai/google/gemini-3-flash-preview/63a96472-79e3-4fb5-bd31-1875ace0ac65.json new file mode 100644 index 0000000000000000000000000000000000000000..e7c48164e4a981176927b18565e7a169f061826b --- /dev/null +++ b/data/vals_ai/google/gemini-3-flash-preview/63a96472-79e3-4fb5-bd31-1875ace0ac65.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/google_gemini-3-flash-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-flash-preview", + "id": "google/gemini-3-flash-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-flash-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 75.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "75", + "raw_stderr": "14.434", + "latency": "8369.51", + "cost_per_test": "3.433448", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 14.434, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 40.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "40", + "raw_stderr": "3.333", + "latency": "1115.935", + "cost_per_test": "0.457793", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 60.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "60", + "raw_stderr": "2.777", + "latency": "608.692", + "cost_per_test": "0.249705", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.777, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 53.933, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "53.933", + "raw_stderr": "1.297", + "latency": "376.158", + "cost_per_test": "0.154312", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.297, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-flash-preview/83dc109c-00c7-4c67-9e06-8ec0f84dca42.json b/data/vals_ai/google/gemini-3-flash-preview/83dc109c-00c7-4c67-9e06-8ec0f84dca42.json new file mode 100644 index 0000000000000000000000000000000000000000..307853054ba91af30adb9c7a6ac1998c266974c0 --- /dev/null +++ b/data/vals_ai/google/gemini-3-flash-preview/83dc109c-00c7-4c67-9e06-8ec0f84dca42.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/google_gemini-3-flash-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-flash-preview", + "id": "google/gemini-3-flash-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-flash-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 20.204, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "20.204", + "raw_stderr": "3.946", + "latency": "806.66", + "cost_per_test": "0.942067", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.946, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-flash-preview/a517529e-207b-4173-a7a8-75571972b8f4.json b/data/vals_ai/google/gemini-3-flash-preview/a517529e-207b-4173-a7a8-75571972b8f4.json new file mode 100644 index 0000000000000000000000000000000000000000..35ea3bd64f3113dc60677fe8b88d5f55bfd236f1 --- /dev/null +++ b/data/vals_ai/google/gemini-3-flash-preview/a517529e-207b-4173-a7a8-75571972b8f4.json @@ -0,0 +1,762 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/google_gemini-3-flash-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-flash-preview", + "id": "google/gemini-3-flash-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-flash-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 35.931, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "35.931", + "raw_stderr": "1.014", + "latency": "226.691", + "cost_per_test": "0.50541", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.014, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 30.204, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "30.204", + "raw_stderr": "0.675", + "latency": "246.543", + "cost_per_test": "0.570657", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.675, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 38.555, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "38.555", + "raw_stderr": "1.771", + "latency": "279.59", + "cost_per_test": "0.762386", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.771, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 45.757, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "45.757", + "raw_stderr": "3.161", + "latency": "255.838", + "cost_per_test": "0.667368", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.161, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 62.076, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.076", + "raw_stderr": "0.589", + "latency": "193.438", + "cost_per_test": "0.27137", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.589, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 14.434, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "14.434", + "raw_stderr": "2.066", + "latency": "250.719", + "cost_per_test": "0.486996", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.066, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 44.608, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "44.608", + "raw_stderr": "0.446", + "latency": "157.579", + "cost_per_test": "0.524815", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.446, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 64.457, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.457", + "raw_stderr": "2.015", + "latency": "262.693", + "cost_per_test": "0.680828", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.015, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 59.004, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.004", + "raw_stderr": "3.028", + "latency": "151.685", + "cost_per_test": "0.296649", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.028, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:overall:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 42.551, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "42.551", + "raw_stderr": "0.434", + "latency": "246.543", + "cost_per_test": "0.570657", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.434, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 18.138, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "18.138", + "raw_stderr": "0.789", + "latency": "453.499", + "cost_per_test": "0.955464", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.789, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-flash-preview/ac127c92-4e8a-4db6-8cbe-ee8321a388ea.json b/data/vals_ai/google/gemini-3-flash-preview/ac127c92-4e8a-4db6-8cbe-ee8321a388ea.json new file mode 100644 index 0000000000000000000000000000000000000000..6c957b6b065c3a3658d7915843cf2f9ec0fcf535 --- /dev/null +++ b/data/vals_ai/google/gemini-3-flash-preview/ac127c92-4e8a-4db6-8cbe-ee8321a388ea.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/code-migration/google_gemini-3-flash-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Code Migration", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-flash-preview", + "id": "google/gemini-3-flash-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-flash-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "code-migration:cli:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.code-migration.cli", + "source_data": { + "dataset_name": "Code Migration - CLI", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cli", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (CLI).", + "metric_id": "vals_ai.code-migration.cli.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 8.521, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cli", + "task_name": "CLI", + "dataset_type": "private", + "industry": "coding", + "raw_score": "8.521", + "raw_stderr": "1.458", + "latency": "294.005", + "cost_per_test": "0.353141", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.458, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "code-migration:cobol:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.code-migration.cobol", + "source_data": { + "dataset_name": "Code Migration - COBOL", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cobol", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (COBOL).", + "metric_id": "vals_ai.code-migration.cobol.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cobol", + "task_name": "COBOL", + "dataset_type": "private", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "254.923", + "cost_per_test": "0.300925", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "code-migration:code_quality:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.code-migration.code_quality", + "source_data": { + "dataset_name": "Code Migration - Code Quality", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "code_quality", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Code Quality).", + "metric_id": "vals_ai.code-migration.code_quality.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 2.662, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "code_quality", + "task_name": "Code Quality", + "dataset_type": "private", + "industry": "coding", + "raw_score": "2.662", + "raw_stderr": "0.047", + "latency": "284.234", + "cost_per_test": "0.340087", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.047, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "code-migration:overall:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.code-migration.overall", + "source_data": { + "dataset_name": "Code Migration - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Overall).", + "metric_id": "vals_ai.code-migration.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 6.391, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "6.391", + "raw_stderr": "1.093", + "latency": "284.234", + "cost_per_test": "0.340087", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.093, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-flash-preview/b866aad8-0959-4a9a-9ea1-9bb35cba2dcd.json b/data/vals_ai/google/gemini-3-flash-preview/b866aad8-0959-4a9a-9ea1-9bb35cba2dcd.json new file mode 100644 index 0000000000000000000000000000000000000000..03783419fdf629410d6e028c46cdac0334d6dfe5 --- /dev/null +++ b/data/vals_ai/google/gemini-3-flash-preview/b866aad8-0959-4a9a-9ea1-9bb35cba2dcd.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/google_gemini-3-flash-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-flash-preview", + "id": "google/gemini-3-flash-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-flash-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 69.917, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "69.917", + "raw_stderr": "1.899", + "latency": "23.491", + "cost_per_test": "0.014379", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.899, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-flash-preview/be081cee-f479-41e0-aab3-0fea1d38fdda.json b/data/vals_ai/google/gemini-3-flash-preview/be081cee-f479-41e0-aab3-0fea1d38fdda.json new file mode 100644 index 0000000000000000000000000000000000000000..c2e3464b34d665225bec934a3c19821577b6f10b --- /dev/null +++ b/data/vals_ai/google/gemini-3-flash-preview/be081cee-f479-41e0-aab3-0fea1d38fdda.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/google_gemini-3-flash-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-flash-preview", + "id": "google/gemini-3-flash-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-flash-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 75.755, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "75.755", + "raw_stderr": "1.208", + "latency": "17.537", + "cost_per_test": "0.009888", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.208, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 68.72, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.72", + "raw_stderr": "0.914", + "latency": "12.838", + "cost_per_test": "0.007477", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.914, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 61.685, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.685", + "raw_stderr": "1.371", + "latency": "8.14", + "cost_per_test": "0.005066", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.371, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-flash-preview/ca613a50-371a-4938-b7a4-c4c48bd41a85.json b/data/vals_ai/google/gemini-3-flash-preview/ca613a50-371a-4938-b7a4-c4c48bd41a85.json new file mode 100644 index 0000000000000000000000000000000000000000..15df146c81fbb3626f6bcbb5f2c875dc37b9a3cc --- /dev/null +++ b/data/vals_ai/google/gemini-3-flash-preview/ca613a50-371a-4938-b7a4-c4c48bd41a85.json @@ -0,0 +1,1071 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/google_gemini-3-flash-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-flash-preview", + "id": "google/gemini-3-flash-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-flash-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.7, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.7", + "raw_stderr": "0.837", + "latency": "25.614", + "cost_per_test": "0.013204", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.837, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.269, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.269", + "raw_stderr": "0.951", + "latency": "31.509", + "cost_per_test": "0.017383", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.951, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.873, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.873", + "raw_stderr": "0.812", + "latency": "50.721", + "cost_per_test": "0.02887", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.812, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.976, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.976", + "raw_stderr": "1.415", + "latency": "36.425", + "cost_per_test": "0.019599", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.415, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.995, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.995", + "raw_stderr": "0.985", + "latency": "25.017", + "cost_per_test": "0.012976", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.985, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.307, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.307", + "raw_stderr": "1.069", + "latency": "66.298", + "cost_per_test": "0.037611", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.069, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.641, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.641", + "raw_stderr": "1.324", + "latency": "37.102", + "cost_per_test": "0.018444", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.324, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.84, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.84", + "raw_stderr": "2.016", + "latency": "27.069", + "cost_per_test": "0.013377", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.016, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.474, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.474", + "raw_stderr": "1.239", + "latency": "54.017", + "cost_per_test": "0.027027", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.239, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 95.707, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "95.707", + "raw_stderr": "0.551", + "latency": "39.099", + "cost_per_test": "0.022316", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.551, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.905, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.905", + "raw_stderr": "1.11", + "latency": "24.514", + "cost_per_test": "0.012398", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.11, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.592, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.592", + "raw_stderr": "0.315", + "latency": "36.249", + "cost_per_test": "0.019364", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.315, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.371, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.371", + "raw_stderr": "1.582", + "latency": "27.179", + "cost_per_test": "0.013969", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.582, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.38, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.38", + "raw_stderr": "0.69", + "latency": "42.839", + "cost_per_test": "0.023862", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.69, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.847, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.847", + "raw_stderr": "1.114", + "latency": "20.086", + "cost_per_test": "0.010053", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.114, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-flash-preview/dda590de-b312-4203-86be-1b8b9d2abe81.json b/data/vals_ai/google/gemini-3-flash-preview/dda590de-b312-4203-86be-1b8b9d2abe81.json new file mode 100644 index 0000000000000000000000000000000000000000..e91ed58da8b1dfb7c17610cf9cd59babcb6f30b5 --- /dev/null +++ b/data/vals_ai/google/gemini-3-flash-preview/dda590de-b312-4203-86be-1b8b9d2abe81.json @@ -0,0 +1,550 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_multimodal_index/google_gemini-3-flash-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Multimodal Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-flash-preview", + "id": "google/gemini-3-flash-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-flash-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_multimodal_index:corp_fin_v2:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.vals_multimodal_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Multimodal Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (CorpFin v2).", + "metric_id": "vals_ai.vals_multimodal_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 67.716, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "67.716", + "raw_stderr": "1.596", + "latency": "14.833", + "cost_per_test": "0.039017", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.596, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:finance_agent:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.vals_multimodal_index.finance_agent", + "source_data": { + "dataset_name": "Vals Multimodal Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_multimodal_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 42.476, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "42.476", + "raw_stderr": "0.666", + "latency": "241.608", + "cost_per_test": "0.543671", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.666, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:mortgage_tax:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.vals_multimodal_index.mortgage_tax", + "source_data": { + "dataset_name": "Vals Multimodal Index - Mortgage Tax", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "mortgage_tax", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Mortgage Tax).", + "metric_id": "vals_ai.vals_multimodal_index.mortgage_tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 68.72, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "mortgage_tax", + "task_name": "Mortgage Tax", + "dataset_type": "private", + "industry": "index", + "raw_score": "68.72", + "raw_stderr": "0.914", + "latency": "12.838", + "cost_per_test": "0.007477", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.914, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:overall:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.vals_multimodal_index.overall", + "source_data": { + "dataset_name": "Vals Multimodal Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Overall).", + "metric_id": "vals_ai.vals_multimodal_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 52.187, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "52.187", + "raw_stderr": "1.414", + "latency": "260.281", + "cost_per_test": "0.28728", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.414, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:sage:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.vals_multimodal_index.sage", + "source_data": { + "dataset_name": "Vals Multimodal Index - SAGE", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "sage", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SAGE).", + "metric_id": "vals_ai.vals_multimodal_index.sage.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 51.849, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "sage", + "task_name": "SAGE", + "dataset_type": "private", + "industry": "index", + "raw_score": "51.849", + "raw_stderr": "3.404", + "latency": "42.697", + "cost_per_test": "0.018229", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.404, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:swebench:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.vals_multimodal_index.swebench", + "source_data": { + "dataset_name": "Vals Multimodal Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SWE-bench).", + "metric_id": "vals_ai.vals_multimodal_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 68.627, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "68.627", + "raw_stderr": "4.594", + "latency": "327.172", + "cost_per_test": "0.306186" + }, + "uncertainty": { + "standard_error": { + "value": 4.594, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:terminal_bench_2_1:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.vals_multimodal_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Multimodal Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_multimodal_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 53.933, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "53.933", + "raw_stderr": "1.297", + "latency": "376.158", + "cost_per_test": "0.154312", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.297, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:vibe_code_bench:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.vals_multimodal_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Multimodal Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_multimodal_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 21.954, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "21.954", + "raw_stderr": "6.673", + "latency": "806.66", + "cost_per_test": "0.942067", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.673, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-flash-preview/e667b850-1491-44c8-b82d-fb7091348e43.json b/data/vals_ai/google/gemini-3-flash-preview/e667b850-1491-44c8-b82d-fb7091348e43.json new file mode 100644 index 0000000000000000000000000000000000000000..41211fdd10ad80665d0cb69cc3a625bb79a73e37 --- /dev/null +++ b/data/vals_ai/google/gemini-3-flash-preview/e667b850-1491-44c8-b82d-fb7091348e43.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/google_gemini-3-flash-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-flash-preview", + "id": "google/gemini-3-flash-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-flash-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 86.364, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.364", + "raw_stderr": "2.439", + "latency": "85.167", + "cost_per_test": "0.04534", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.439, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 89.394, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.394", + "raw_stderr": "2.188", + "latency": "82.513", + "cost_per_test": "0.043352", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.188, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 87.879, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.879", + "raw_stderr": "1.638", + "latency": "83.84", + "cost_per_test": "0.044346", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.638, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-flash-preview/ed99c799-7183-4f5f-961d-7cc8407c2b1c.json b/data/vals_ai/google/gemini-3-flash-preview/ed99c799-7183-4f5f-961d-7cc8407c2b1c.json new file mode 100644 index 0000000000000000000000000000000000000000..fb4c52631c50e9f99547ee6ef65981f3549a8393 --- /dev/null +++ b/data/vals_ai/google/gemini-3-flash-preview/ed99c799-7183-4f5f-961d-7cc8407c2b1c.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/programbench/google_gemini-3-flash-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProgramBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-flash-preview", + "id": "google/gemini-3-flash-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-flash-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "programbench:almost:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.programbench.almost", + "source_data": { + "dataset_name": "ProgramBench - Almost Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "almost", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Almost Resolved).", + "metric_id": "vals_ai.programbench.almost.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "almost", + "task_name": "Almost Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "418.92", + "cost_per_test": "0.439965", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "programbench:overall:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.programbench.overall", + "source_data": { + "dataset_name": "ProgramBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Overall).", + "metric_id": "vals_ai.programbench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "418.92", + "cost_per_test": "0.439965", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "programbench:partial:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.programbench.partial", + "source_data": { + "dataset_name": "ProgramBench - Raw Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "partial", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Raw Pass Rate).", + "metric_id": "vals_ai.programbench.partial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 33.83, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "partial", + "task_name": "Raw Pass Rate", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.83", + "raw_stderr": "1.622", + "latency": "418.92", + "cost_per_test": "0.439965", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.622, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "programbench:strict:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.programbench.strict", + "source_data": { + "dataset_name": "ProgramBench - Fully Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "strict", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Fully Resolved).", + "metric_id": "vals_ai.programbench.strict.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "strict", + "task_name": "Fully Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "418.92", + "cost_per_test": "0.439965", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-flash-preview/ef462e4f-ffe8-4b8c-9417-8511426b954e.json b/data/vals_ai/google/gemini-3-flash-preview/ef462e4f-ffe8-4b8c-9417-8511426b954e.json new file mode 100644 index 0000000000000000000000000000000000000000..507b555db328d473d148d6844d0d301538b43397 --- /dev/null +++ b/data/vals_ai/google/gemini-3-flash-preview/ef462e4f-ffe8-4b8c-9417-8511426b954e.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/google_gemini-3-flash-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-flash-preview", + "id": "google/gemini-3-flash-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-flash-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 15.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "15", + "raw_stderr": "3.571", + "latency": "1141.548", + "cost_per_test": "0.327286", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.571, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-flash-preview/f0c1f69c-3b18-45b0-b73b-0425a267e2aa.json b/data/vals_ai/google/gemini-3-flash-preview/f0c1f69c-3b18-45b0-b73b-0425a267e2aa.json new file mode 100644 index 0000000000000000000000000000000000000000..04daa948bf8035c5619fa97c5089e0efd3f5d63a --- /dev/null +++ b/data/vals_ai/google/gemini-3-flash-preview/f0c1f69c-3b18-45b0-b73b-0425a267e2aa.json @@ -0,0 +1,105 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/google_gemini-3-flash-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-flash-preview", + "id": "google/gemini-3-flash-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-flash-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 87.63, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.63", + "raw_stderr": "0.792", + "latency": "27.856", + "cost_per_test": "0.014422", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.792, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-flash-preview/f29073f3-9675-4558-a18d-b17d2d0418cd.json b/data/vals_ai/google/gemini-3-flash-preview/f29073f3-9675-4558-a18d-b17d2d0418cd.json new file mode 100644 index 0000000000000000000000000000000000000000..712be3d89199b81505256c01b58205d6d7d37a58 --- /dev/null +++ b/data/vals_ai/google/gemini-3-flash-preview/f29073f3-9675-4558-a18d-b17d2d0418cd.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/google_gemini-3-flash-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-flash-preview", + "id": "google/gemini-3-flash-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-flash-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 68.065, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.065", + "raw_stderr": "1.592", + "latency": "11.684", + "cost_per_test": "0.098516", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.592, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 63.52, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.52", + "raw_stderr": "1.643", + "latency": "7.32", + "cost_per_test": "0.003928", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.643, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 66.434, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.434", + "raw_stderr": "0.93", + "latency": "11.279", + "cost_per_test": "0.047154", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.93, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:google/gemini-3-flash-preview:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 67.716, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.716", + "raw_stderr": "1.596", + "latency": "14.833", + "cost_per_test": "0.039017", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.596, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-pro-preview/01b5741a-1654-4e68-a4bd-6f50328fc4e7.json b/data/vals_ai/google/gemini-3-pro-preview/01b5741a-1654-4e68-a4bd-6f50328fc4e7.json new file mode 100644 index 0000000000000000000000000000000000000000..e6051475ff580922770c76be803c36ef318ddf09 --- /dev/null +++ b/data/vals_ai/google/gemini-3-pro-preview/01b5741a-1654-4e68-a4bd-6f50328fc4e7.json @@ -0,0 +1,1071 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/google_gemini-3-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-pro-preview", + "id": "google/gemini-3-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 95.397, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "95.397", + "raw_stderr": "0.783", + "latency": "17.19", + "cost_per_test": "0.026082", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.783, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.663, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.663", + "raw_stderr": "0.867", + "latency": "19.25", + "cost_per_test": "0.030708", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.867, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.845, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.845", + "raw_stderr": "0.766", + "latency": "33.183", + "cost_per_test": "0.053294", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.766, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.463, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.463", + "raw_stderr": "1.38", + "latency": "22.896", + "cost_per_test": "0.035823", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.38, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.943, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.943", + "raw_stderr": "0.937", + "latency": "17.613", + "cost_per_test": "0.026377", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.937, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.616, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.616", + "raw_stderr": "1.058", + "latency": "52.3", + "cost_per_test": "0.080638", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.058, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.619, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.619", + "raw_stderr": "1.294", + "latency": "22.817", + "cost_per_test": "0.031299", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.294, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.777, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.777", + "raw_stderr": "1.84", + "latency": "19.767", + "cost_per_test": "0.029725", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.84, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.47, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.47", + "raw_stderr": "1.146", + "latency": "30.789", + "cost_per_test": "0.044563", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.146, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 96.003, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "96.003", + "raw_stderr": "0.533", + "latency": "24.058", + "cost_per_test": "0.03927", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.533, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.987, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.987", + "raw_stderr": "1.07", + "latency": "16.616", + "cost_per_test": "0.024161", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.07, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.102, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.102", + "raw_stderr": "0.295", + "latency": "24.272", + "cost_per_test": "0.036926", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.295, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.978, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.978", + "raw_stderr": "1.402", + "latency": "16.751", + "cost_per_test": "0.024276", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.402, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.687, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.687", + "raw_stderr": "0.675", + "latency": "29.722", + "cost_per_test": "0.046396", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.675, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.977, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.977", + "raw_stderr": "1.014", + "latency": "16.858", + "cost_per_test": "0.024357", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.014, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-pro-preview/493d47fd-6f37-470b-9f84-824d0010fd6f.json b/data/vals_ai/google/gemini-3-pro-preview/493d47fd-6f37-470b-9f84-824d0010fd6f.json new file mode 100644 index 0000000000000000000000000000000000000000..e616b8105f9270dcb4e016327c5e54ae11781ae8 --- /dev/null +++ b/data/vals_ai/google/gemini-3-pro-preview/493d47fd-6f37-470b-9f84-824d0010fd6f.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/google_gemini-3-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-pro-preview", + "id": "google/gemini-3-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 61.079, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.079", + "raw_stderr": "1.394", + "latency": "24.404", + "cost_per_test": "0.034377", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.394, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 72.568, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "72.568", + "raw_stderr": "0.872", + "latency": "24.404", + "cost_per_test": "0.034377", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.872, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 84.056, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "84.056", + "raw_stderr": "1.047", + "latency": "24.404", + "cost_per_test": "0.034377", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.047, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-pro-preview/534e0cc9-eed1-44b8-afcb-1e5b42c08f33.json b/data/vals_ai/google/gemini-3-pro-preview/534e0cc9-eed1-44b8-afcb-1e5b42c08f33.json new file mode 100644 index 0000000000000000000000000000000000000000..30cf5d236fb3dc0b2681ada59f1a3503904a8643 --- /dev/null +++ b/data/vals_ai/google/gemini-3-pro-preview/534e0cc9-eed1-44b8-afcb-1e5b42c08f33.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/google_gemini-3-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-pro-preview", + "id": "google/gemini-3-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 90.909, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.909", + "raw_stderr": "2.043", + "latency": "58.765", + "cost_per_test": "0.087749", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.043, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 92.424, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.424", + "raw_stderr": "1.881", + "latency": "63.916", + "cost_per_test": "0.091397", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.881, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 91.666, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.666", + "raw_stderr": "1.389", + "latency": "61.34", + "cost_per_test": "0.089573", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.389, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-pro-preview/7175c56c-4572-40c0-9d45-5eff4d4c41d2.json b/data/vals_ai/google/gemini-3-pro-preview/7175c56c-4572-40c0-9d45-5eff4d4c41d2.json new file mode 100644 index 0000000000000000000000000000000000000000..3536834c85b3664a22846e7ae445b9039617531b --- /dev/null +++ b/data/vals_ai/google/gemini-3-pro-preview/7175c56c-4572-40c0-9d45-5eff4d4c41d2.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/google_gemini-3-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-pro-preview", + "id": "google/gemini-3-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 72.036, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "72.036", + "raw_stderr": "1.9", + "latency": "43.386", + "cost_per_test": "0.061162", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.9, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-pro-preview/7e4fa851-74d0-4acf-ac6f-30f0a0d6155c.json b/data/vals_ai/google/gemini-3-pro-preview/7e4fa851-74d0-4acf-ac6f-30f0a0d6155c.json new file mode 100644 index 0000000000000000000000000000000000000000..f6ce26b1071614be444f7b5d1af2af40346a2591 --- /dev/null +++ b/data/vals_ai/google/gemini-3-pro-preview/7e4fa851-74d0-4acf-ac6f-30f0a0d6155c.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/google_gemini-3-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-pro-preview", + "id": "google/gemini-3-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 59.441, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.441", + "raw_stderr": "1.676", + "latency": "14.982", + "cost_per_test": "0.005289", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.676, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 67.016, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.016", + "raw_stderr": "1.605", + "latency": "21.418", + "cost_per_test": "0.377404", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.605, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 63.675, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.675", + "raw_stderr": "0.946", + "latency": "20.059", + "cost_per_test": "0.179069", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.946, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 64.569, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.569", + "raw_stderr": "1.633", + "latency": "23.778", + "cost_per_test": "0.154515", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.633, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-pro-preview/7fc89186-e045-4300-88b4-d607aed62aac.json b/data/vals_ai/google/gemini-3-pro-preview/7fc89186-e045-4300-88b4-d607aed62aac.json new file mode 100644 index 0000000000000000000000000000000000000000..e1492d5efd289bedeb2f334babb1d3dee2298d88 --- /dev/null +++ b/data/vals_ai/google/gemini-3-pro-preview/7fc89186-e045-4300-88b4-d607aed62aac.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/google_gemini-3-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-pro-preview", + "id": "google/gemini-3-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "20", + "raw_stderr": "4", + "latency": "10383.833", + "cost_per_test": "1.229076", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 4.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-pro-preview/b241fadb-f56f-4f60-bb77-fa6377796b25.json b/data/vals_ai/google/gemini-3-pro-preview/b241fadb-f56f-4f60-bb77-fa6377796b25.json new file mode 100644 index 0000000000000000000000000000000000000000..d480c35ce64344dd3b9cb46fc7b6707030c9147e --- /dev/null +++ b/data/vals_ai/google/gemini-3-pro-preview/b241fadb-f56f-4f60-bb77-fa6377796b25.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/google_gemini-3-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-pro-preview", + "id": "google/gemini-3-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 38.5, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "38.5", + "raw_stderr": "17.904", + "latency": "1079.286", + "cost_per_test": "0.328592", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 17.904, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 39.167, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "39.167", + "raw_stderr": "14.411", + "latency": "721.594", + "cost_per_test": "0.333784", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 14.411, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "ioi:overall:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 38.834, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "38.834", + "raw_stderr": "11.492", + "latency": "900.44", + "cost_per_test": "0.331188", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 11.492, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-pro-preview/bb0fbf07-6295-4c79-97c6-c53aa92eefac.json b/data/vals_ai/google/gemini-3-pro-preview/bb0fbf07-6295-4c79-97c6-c53aa92eefac.json new file mode 100644 index 0000000000000000000000000000000000000000..f33f4c16500cc2c64cf07963ab36f57da847558d --- /dev/null +++ b/data/vals_ai/google/gemini-3-pro-preview/bb0fbf07-6295-4c79-97c6-c53aa92eefac.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/google_gemini-3-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-pro-preview", + "id": "google/gemini-3-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 75.994, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "75.994", + "raw_stderr": "1.204", + "latency": "32.566", + "cost_per_test": "0.050966", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.204, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 69.078, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "69.078", + "raw_stderr": "0.911", + "latency": "25.602", + "cost_per_test": "0.041198", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.911, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 62.162, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.162", + "raw_stderr": "1.367", + "latency": "18.638", + "cost_per_test": "0.031431", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.367, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-pro-preview/c9779717-b575-4d63-8575-01f09f76419d.json b/data/vals_ai/google/gemini-3-pro-preview/c9779717-b575-4d63-8575-01f09f76419d.json new file mode 100644 index 0000000000000000000000000000000000000000..034e69c57ecec28a4815830e95e21c6d5abf331a --- /dev/null +++ b/data/vals_ai/google/gemini-3-pro-preview/c9779717-b575-4d63-8575-01f09f76419d.json @@ -0,0 +1,381 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/google_gemini-3-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-pro-preview", + "id": "google/gemini-3-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 42.857, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "42.857", + "raw_stderr": "7.729", + "latency": "508.663", + "cost_per_test": "1.051476", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 7.729, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 73.563, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "73.563", + "raw_stderr": "2.735", + "latency": "414.978", + "cost_per_test": "0.815353", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.735, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 88.144, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "88.144", + "raw_stderr": "2.327", + "latency": "340.798", + "cost_per_test": "0.649879", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.327, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "485.399", + "cost_per_test": "0.909358", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:overall:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 76.4, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "76.4", + "raw_stderr": "1.901", + "latency": "394.488", + "cost_per_test": "0.771547", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.901, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-pro-preview/cb6a8d7f-108d-4e77-8a08-90d5e484a0c4.json b/data/vals_ai/google/gemini-3-pro-preview/cb6a8d7f-108d-4e77-8a08-90d5e484a0c4.json new file mode 100644 index 0000000000000000000000000000000000000000..0b4194124df733eaded21c73e8cb2293f640c880 --- /dev/null +++ b/data/vals_ai/google/gemini-3-pro-preview/cb6a8d7f-108d-4e77-8a08-90d5e484a0c4.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/google_gemini-3-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-pro-preview", + "id": "google/gemini-3-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 14.3, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "14.3", + "raw_stderr": "3.061", + "latency": "10398.545", + "cost_per_test": "6.417592", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.061, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-pro-preview/de913e7e-5e46-4220-8544-82e148ac2768.json b/data/vals_ai/google/gemini-3-pro-preview/de913e7e-5e46-4220-8544-82e148ac2768.json new file mode 100644 index 0000000000000000000000000000000000000000..5f69ac79dcfd6c03c6d4dd7b010ce70a6e44bb09 --- /dev/null +++ b/data/vals_ai/google/gemini-3-pro-preview/de913e7e-5e46-4220-8544-82e148ac2768.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/google_gemini-3-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-pro-preview", + "id": "google/gemini-3-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 99.068, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "99.068", + "raw_stderr": "0.535", + "latency": "41.655", + "cost_per_test": "0.06154", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:hard:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 70.857, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "70.857", + "raw_stderr": "2.429", + "latency": "131.816", + "cost_per_test": "0.19064", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.429, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:medium:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 89.295, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "89.295", + "raw_stderr": "1.58", + "latency": "88.206", + "cost_per_test": "0.129689", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.58, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:overall:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 86.407, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "86.407", + "raw_stderr": "0.982", + "latency": "87.226", + "cost_per_test": "0.12729", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.982, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-pro-preview/e1652817-b043-4cd4-9170-01300c6d4b0f.json b/data/vals_ai/google/gemini-3-pro-preview/e1652817-b043-4cd4-9170-01300c6d4b0f.json new file mode 100644 index 0000000000000000000000000000000000000000..18282c6f878c4b6246797d7f38c25ec24214188b --- /dev/null +++ b/data/vals_ai/google/gemini-3-pro-preview/e1652817-b043-4cd4-9170-01300c6d4b0f.json @@ -0,0 +1,366 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/google_gemini-3-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-pro-preview", + "id": "google/gemini-3-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 47.619, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "47.619", + "raw_stderr": "6.267", + "latency": "77.028", + "cost_per_test": "0.011059", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.267, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:automata:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 30.158, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "30.158", + "raw_stderr": "5.884", + "latency": "113.22", + "cost_per_test": "0.010942", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.884, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:calculus:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 69.444, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "69.444", + "raw_stderr": "5.961", + "latency": "122.682", + "cost_per_test": "0.036851", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.961, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 65.067, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "65.067", + "raw_stderr": "5.355", + "latency": "101.578", + "cost_per_test": "0.014122", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.355, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:overall:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 47.615, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "47.615", + "raw_stderr": "3.376", + "latency": "97.275", + "cost_per_test": "0.012041", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.376, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-pro-preview/f96755ea-24af-43c3-9c5f-278e6749aed3.json b/data/vals_ai/google/gemini-3-pro-preview/f96755ea-24af-43c3-9c5f-278e6749aed3.json new file mode 100644 index 0000000000000000000000000000000000000000..610065c4b8042a02258d23a3afdc1fdeb2c52272 --- /dev/null +++ b/data/vals_ai/google/gemini-3-pro-preview/f96755ea-24af-43c3-9c5f-278e6749aed3.json @@ -0,0 +1,450 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/google_gemini-3-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-pro-preview", + "id": "google/gemini-3-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 90.365, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.365", + "raw_stderr": "0.595", + "latency": "7.663", + "cost_per_test": "0.009074", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.595, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 84.251, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.251", + "raw_stderr": "0.268", + "latency": "9.49", + "cost_per_test": "0.011919", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.268, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 91.172, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "91.172", + "raw_stderr": "0.548", + "latency": "7.218", + "cost_per_test": "0.010184", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.548, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.025, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.025", + "raw_stderr": "0.368", + "latency": "8.327", + "cost_per_test": "0.01046", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.368, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 84.09, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.09", + "raw_stderr": "0.815", + "latency": "7.44", + "cost_per_test": "0.009151", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.815, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 85.246, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.246", + "raw_stderr": "1.415", + "latency": "9.824", + "cost_per_test": "0.01197", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.415, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3-pro-preview/fe30dd14-eb64-4166-8aaf-352953e226a6.json b/data/vals_ai/google/gemini-3-pro-preview/fe30dd14-eb64-4166-8aaf-352953e226a6.json new file mode 100644 index 0000000000000000000000000000000000000000..1f70216b582395a5718abdcb4c50233af107ff13 --- /dev/null +++ b/data/vals_ai/google/gemini-3-pro-preview/fe30dd14-eb64-4166-8aaf-352953e226a6.json @@ -0,0 +1,105 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/google_gemini-3-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3-pro-preview", + "id": "google/gemini-3-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:google/gemini-3-pro-preview:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 87.514, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.514", + "raw_stderr": "0.795", + "latency": "48.102", + "cost_per_test": "0.010159", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.795, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-flash-lite-preview/12b82b54-63d6-4e21-b1f2-6061b34cfe08.json b/data/vals_ai/google/gemini-3.1-flash-lite-preview/12b82b54-63d6-4e21-b1f2-6061b34cfe08.json new file mode 100644 index 0000000000000000000000000000000000000000..aa1e3cc707cad44588151f6affe0b17a742b7b32 --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-flash-lite-preview/12b82b54-63d6-4e21-b1f2-6061b34cfe08.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/programbench/google_gemini-3.1-flash-lite-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProgramBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-flash-lite-preview", + "id": "google/gemini-3.1-flash-lite-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-flash-lite-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "programbench:almost:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.programbench.almost", + "source_data": { + "dataset_name": "ProgramBench - Almost Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "almost", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Almost Resolved).", + "metric_id": "vals_ai.programbench.almost.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "almost", + "task_name": "Almost Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "176.246", + "cost_per_test": "0.063906", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "programbench:overall:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.programbench.overall", + "source_data": { + "dataset_name": "ProgramBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Overall).", + "metric_id": "vals_ai.programbench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "176.246", + "cost_per_test": "0.063906", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "programbench:partial:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.programbench.partial", + "source_data": { + "dataset_name": "ProgramBench - Raw Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "partial", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Raw Pass Rate).", + "metric_id": "vals_ai.programbench.partial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 14.282, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "partial", + "task_name": "Raw Pass Rate", + "dataset_type": "public", + "industry": "coding", + "raw_score": "14.282", + "raw_stderr": "1.072", + "latency": "176.246", + "cost_per_test": "0.063906", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.072, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "programbench:strict:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.programbench.strict", + "source_data": { + "dataset_name": "ProgramBench - Fully Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "strict", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Fully Resolved).", + "metric_id": "vals_ai.programbench.strict.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "strict", + "task_name": "Fully Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "176.246", + "cost_per_test": "0.063906", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-flash-lite-preview/1de9bc26-ca52-4bd1-bc92-dce40615621f.json b/data/vals_ai/google/gemini-3.1-flash-lite-preview/1de9bc26-ca52-4bd1-bc92-dce40615621f.json new file mode 100644 index 0000000000000000000000000000000000000000..9b6d363a4f0585b4a1f9083dca2530e1e635d761 --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-flash-lite-preview/1de9bc26-ca52-4bd1-bc92-dce40615621f.json @@ -0,0 +1,762 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/google_gemini-3.1-flash-lite-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-flash-lite-preview", + "id": "google/gemini-3.1-flash-lite-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-flash-lite-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 30.969, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "30.969", + "raw_stderr": "4.238", + "latency": "61.502", + "cost_per_test": "0.177463", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 4.238, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 19.303, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "19.303", + "raw_stderr": "0.526", + "latency": "54.778", + "cost_per_test": "0.142482", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.526, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 34.057, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "34.057", + "raw_stderr": "0.849", + "latency": "61.889", + "cost_per_test": "0.196515", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.849, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 37.38, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "37.38", + "raw_stderr": "2.128", + "latency": "54.432", + "cost_per_test": "0.162369", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.128, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 31.261, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "31.261", + "raw_stderr": "0.373", + "latency": "57.997", + "cost_per_test": "0.08654", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.373, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 4.112, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "4.112", + "raw_stderr": "1.206", + "latency": "60.912", + "cost_per_test": "0.141152", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.206, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 36.184, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "36.184", + "raw_stderr": "0.795", + "latency": "42.875", + "cost_per_test": "0.128348", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.795, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 49.932, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "49.932", + "raw_stderr": "0.284", + "latency": "54.088", + "cost_per_test": "0.151645", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.284, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 34.189, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "34.189", + "raw_stderr": "1.174", + "latency": "39.442", + "cost_per_test": "0.094503", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.174, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:overall:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 29.988, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "29.988", + "raw_stderr": "0.643", + "latency": "54.778", + "cost_per_test": "0.142482", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.643, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 11.808, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "11.808", + "raw_stderr": "0.741", + "latency": "60.876", + "cost_per_test": "0.143796", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.741, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-flash-lite-preview/480441f3-deef-49bc-9398-ec6a99e4acda.json b/data/vals_ai/google/gemini-3.1-flash-lite-preview/480441f3-deef-49bc-9398-ec6a99e4acda.json new file mode 100644 index 0000000000000000000000000000000000000000..333bfaf1e2a668fe5e59ba6dc476f78077006688 --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-flash-lite-preview/480441f3-deef-49bc-9398-ec6a99e4acda.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/google_gemini-3.1-flash-lite-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-flash-lite-preview", + "id": "google/gemini-3.1-flash-lite-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-flash-lite-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 79.293, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.293", + "raw_stderr": "2.88", + "latency": "16.085", + "cost_per_test": "0.002329", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.88, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 82.828, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.828", + "raw_stderr": "2.68", + "latency": "16.842", + "cost_per_test": "0.002266", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.68, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 81.06, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.06", + "raw_stderr": "1.967", + "latency": "16.464", + "cost_per_test": "0.002298", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.967, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-flash-lite-preview/4b776856-d89f-424b-ba7c-c6f8ae2fa49e.json b/data/vals_ai/google/gemini-3.1-flash-lite-preview/4b776856-d89f-424b-ba7c-c6f8ae2fa49e.json new file mode 100644 index 0000000000000000000000000000000000000000..9374e5c145c2b7679c21568293aaa99d1d379f74 --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-flash-lite-preview/4b776856-d89f-424b-ba7c-c6f8ae2fa49e.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/google_gemini-3.1-flash-lite-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-flash-lite-preview", + "id": "google/gemini-3.1-flash-lite-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-flash-lite-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.758, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.758", + "raw_stderr": "0.617", + "latency": "14.6", + "cost_per_test": "0.002031", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.617, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:hard:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 55.429, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "55.429", + "raw_stderr": "2.657", + "latency": "43.676", + "cost_per_test": "0.006597", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.657, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:medium:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 86.162, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "86.162", + "raw_stderr": "1.764", + "latency": "33.971", + "cost_per_test": "0.004421", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.764, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:overall:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 80.116, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "80.116", + "raw_stderr": "1.083", + "latency": "30.749", + "cost_per_test": "0.00435", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.083, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-flash-lite-preview/5166082c-b11e-490a-ac01-dbf9f9e18822.json b/data/vals_ai/google/gemini-3.1-flash-lite-preview/5166082c-b11e-490a-ac01-dbf9f9e18822.json new file mode 100644 index 0000000000000000000000000000000000000000..378c8016ec621484cc1a3b01a8303dcb21f2d1fa --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-flash-lite-preview/5166082c-b11e-490a-ac01-dbf9f9e18822.json @@ -0,0 +1,366 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/google_gemini-3.1-flash-lite-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-flash-lite-preview", + "id": "google/gemini-3.1-flash-lite-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-flash-lite-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 47.619, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "47.619", + "raw_stderr": "6.27", + "latency": "18.889", + "cost_per_test": "0.009314", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.27, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:automata:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 42.064, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "42.064", + "raw_stderr": "6.255", + "latency": "14.224", + "cost_per_test": "0.005199", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.255, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:calculus:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 58.333, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "58.333", + "raw_stderr": "6.102", + "latency": "19.833", + "cost_per_test": "0.002188", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.102, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 58.936, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "58.936", + "raw_stderr": "5.549", + "latency": "15.659", + "cost_per_test": "0.007622", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.549, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:overall:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 49.54, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "49.54", + "raw_stderr": "3.484", + "latency": "16.257", + "cost_per_test": "0.007378", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.484, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-flash-lite-preview/5b5c15ad-9363-4bf4-82aa-243ca6f3fc70.json b/data/vals_ai/google/gemini-3.1-flash-lite-preview/5b5c15ad-9363-4bf4-82aa-243ca6f3fc70.json new file mode 100644 index 0000000000000000000000000000000000000000..01239ff1711c28d2afb1b1369c52e421b7f29650 --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-flash-lite-preview/5b5c15ad-9363-4bf4-82aa-243ca6f3fc70.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/google_gemini-3.1-flash-lite-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-flash-lite-preview", + "id": "google/gemini-3.1-flash-lite-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-flash-lite-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 77.027, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "77.027", + "raw_stderr": "1.186", + "latency": "6.323", + "cost_per_test": "0.003233", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.186, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 68.044, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.044", + "raw_stderr": "0.912", + "latency": "5.368", + "cost_per_test": "0.00284", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.912, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 59.062, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.062", + "raw_stderr": "1.386", + "latency": "4.413", + "cost_per_test": "0.002448", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.386, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-flash-lite-preview/5de0c446-816c-414f-a815-ae6f13aabfad.json b/data/vals_ai/google/gemini-3.1-flash-lite-preview/5de0c446-816c-414f-a815-ae6f13aabfad.json new file mode 100644 index 0000000000000000000000000000000000000000..df6b92e596bfc346864b03acac53a1b65458752d --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-flash-lite-preview/5de0c446-816c-414f-a815-ae6f13aabfad.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/google_gemini-3.1-flash-lite-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-flash-lite-preview", + "id": "google/gemini-3.1-flash-lite-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-flash-lite-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 63.902, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "63.902", + "raw_stderr": "1.823", + "latency": "16.793", + "cost_per_test": "0.002195", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.823, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-flash-lite-preview/5e8e4dde-154c-42b4-ad93-b898e1db7f17.json b/data/vals_ai/google/gemini-3.1-flash-lite-preview/5e8e4dde-154c-42b4-ad93-b898e1db7f17.json new file mode 100644 index 0000000000000000000000000000000000000000..fd6ffc8aa1d319ed67606e92ea064d608b1c0d3d --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-flash-lite-preview/5e8e4dde-154c-42b4-ad93-b898e1db7f17.json @@ -0,0 +1,381 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/google_gemini-3.1-flash-lite-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-flash-lite-preview", + "id": "google/gemini-3.1-flash-lite-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-flash-lite-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 21.429, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "21.429", + "raw_stderr": "6.408", + "latency": "156.311", + "cost_per_test": "0.132103", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.408, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 59.77, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "59.77", + "raw_stderr": "3.041", + "latency": "144.518", + "cost_per_test": "0.114825", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.041, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 76.804, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "76.804", + "raw_stderr": "3.038", + "latency": "112.63", + "cost_per_test": "0.087251", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.038, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "111.196", + "cost_per_test": "0.083345", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:overall:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 62.8, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "62.8", + "raw_stderr": "2.164", + "latency": "132.936", + "cost_per_test": "0.105389", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.164, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-flash-lite-preview/5f426679-187d-4cbe-a1d9-4a4ccc648426.json b/data/vals_ai/google/gemini-3.1-flash-lite-preview/5f426679-187d-4cbe-a1d9-4a4ccc648426.json new file mode 100644 index 0000000000000000000000000000000000000000..f931c1eb9ede5d159dfa6eb8b07d582c90c5b0f3 --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-flash-lite-preview/5f426679-187d-4cbe-a1d9-4a4ccc648426.json @@ -0,0 +1,550 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_multimodal_index/google_gemini-3.1-flash-lite-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Multimodal Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-flash-lite-preview", + "id": "google/gemini-3.1-flash-lite-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-flash-lite-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_multimodal_index:corp_fin_v2:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.vals_multimodal_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Multimodal Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (CorpFin v2).", + "metric_id": "vals_ai.vals_multimodal_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 60.839, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "60.839", + "raw_stderr": "1.666", + "latency": "4.708", + "cost_per_test": "0.022927", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.666, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:finance_agent:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.vals_multimodal_index.finance_agent", + "source_data": { + "dataset_name": "Vals Multimodal Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_multimodal_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 30.068, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "30.068", + "raw_stderr": "0.727", + "latency": "53.577", + "cost_per_test": "0.133469", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.727, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:mortgage_tax:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.vals_multimodal_index.mortgage_tax", + "source_data": { + "dataset_name": "Vals Multimodal Index - Mortgage Tax", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "mortgage_tax", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Mortgage Tax).", + "metric_id": "vals_ai.vals_multimodal_index.mortgage_tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 68.044, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "mortgage_tax", + "task_name": "Mortgage Tax", + "dataset_type": "private", + "industry": "index", + "raw_score": "68.044", + "raw_stderr": "0.912", + "latency": "5.368", + "cost_per_test": "0.00284", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.912, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:overall:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.vals_multimodal_index.overall", + "source_data": { + "dataset_name": "Vals Multimodal Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Overall).", + "metric_id": "vals_ai.vals_multimodal_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 41.352, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "41.352", + "raw_stderr": "0.661", + "latency": "152.434", + "cost_per_test": "0.123994", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.661, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:sage:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.vals_multimodal_index.sage", + "source_data": { + "dataset_name": "Vals Multimodal Index - SAGE", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "sage", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SAGE).", + "metric_id": "vals_ai.vals_multimodal_index.sage.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 49.54, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "sage", + "task_name": "SAGE", + "dataset_type": "private", + "industry": "index", + "raw_score": "49.54", + "raw_stderr": "3.484", + "latency": "16.257", + "cost_per_test": "0.00773", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.484, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:swebench:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.vals_multimodal_index.swebench", + "source_data": { + "dataset_name": "Vals Multimodal Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SWE-bench).", + "metric_id": "vals_ai.vals_multimodal_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 57.843, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "57.843", + "raw_stderr": "4.889", + "latency": "136.077", + "cost_per_test": "0.106331" + }, + "uncertainty": { + "standard_error": { + "value": 4.889, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:terminal_bench_2_1:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.vals_multimodal_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Multimodal Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_multimodal_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 34.082, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "34.082", + "raw_stderr": "0.991", + "latency": "278.342", + "cost_per_test": "0.084894", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.991, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:vibe_code_bench:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.vals_multimodal_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Multimodal Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_multimodal_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "0", + "raw_stderr": "0", + "latency": "572.711", + "cost_per_test": "0.51", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-flash-lite-preview/66de9ecb-23c4-4612-9d71-d62cc28be4c8.json b/data/vals_ai/google/gemini-3.1-flash-lite-preview/66de9ecb-23c4-4612-9d71-d62cc28be4c8.json new file mode 100644 index 0000000000000000000000000000000000000000..452bf0fd6367069f86755f11da2e352f435bf62a --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-flash-lite-preview/66de9ecb-23c4-4612-9d71-d62cc28be4c8.json @@ -0,0 +1,105 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/google_gemini-3.1-flash-lite-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-flash-lite-preview", + "id": "google/gemini-3.1-flash-lite-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-flash-lite-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 82.486, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.486", + "raw_stderr": "0.914", + "latency": "12.448", + "cost_per_test": "0.003187", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.914, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-flash-lite-preview/7d92d25d-a0be-4b7e-8af8-0a4cd566774e.json b/data/vals_ai/google/gemini-3.1-flash-lite-preview/7d92d25d-a0be-4b7e-8af8-0a4cd566774e.json new file mode 100644 index 0000000000000000000000000000000000000000..bd3d15190f9d24dd1c53c31a80f60f8ee721ef49 --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-flash-lite-preview/7d92d25d-a0be-4b7e-8af8-0a4cd566774e.json @@ -0,0 +1,696 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/emb/google_gemini-3.1-flash-lite-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - EMB", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-flash-lite-preview", + "id": "google/gemini-3.1-flash-lite-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-flash-lite-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "emb:comps:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.emb.comps", + "source_data": { + "dataset_name": "EMB - Comps", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "comps", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Comps).", + "metric_id": "vals_ai.emb.comps.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 10.48, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "comps", + "task_name": "Comps", + "dataset_type": "private", + "industry": "finance", + "raw_score": "10.48", + "raw_stderr": "3.475", + "latency": "101.863", + "cost_per_test": "0.182456", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.475, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:dataroom:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.emb.dataroom", + "source_data": { + "dataset_name": "EMB - Dataroom", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dataroom", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Dataroom).", + "metric_id": "vals_ai.emb.dataroom.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 15.178, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dataroom", + "task_name": "Dataroom", + "dataset_type": "private", + "industry": "finance", + "raw_score": "15.178", + "raw_stderr": "5.151", + "latency": "84.269", + "cost_per_test": "0.050422", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.151, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:dcf:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.emb.dcf", + "source_data": { + "dataset_name": "EMB - DCF", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dcf", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (DCF).", + "metric_id": "vals_ai.emb.dcf.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 5.098, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dcf", + "task_name": "DCF", + "dataset_type": "private", + "industry": "finance", + "raw_score": "5.098", + "raw_stderr": "1.006", + "latency": "64.7", + "cost_per_test": "0.046823", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.006, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:lbo:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.emb.lbo", + "source_data": { + "dataset_name": "EMB - LBO", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "lbo", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (LBO).", + "metric_id": "vals_ai.emb.lbo.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 8.805, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "lbo", + "task_name": "LBO", + "dataset_type": "private", + "industry": "finance", + "raw_score": "8.805", + "raw_stderr": "2.184", + "latency": "49.222", + "cost_per_test": "0.037241", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.184, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:ma:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.emb.ma", + "source_data": { + "dataset_name": "EMB - M&A", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "ma", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (M&A).", + "metric_id": "vals_ai.emb.ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 6.751, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "ma", + "task_name": "M&A", + "dataset_type": "private", + "industry": "finance", + "raw_score": "6.751", + "raw_stderr": "1.416", + "latency": "55.632", + "cost_per_test": "0.034741", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.416, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:misc:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.emb.misc", + "source_data": { + "dataset_name": "EMB - Misc", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "misc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Misc).", + "metric_id": "vals_ai.emb.misc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 8.699, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "misc", + "task_name": "Misc", + "dataset_type": "private", + "industry": "finance", + "raw_score": "8.699", + "raw_stderr": "2.021", + "latency": "59.445", + "cost_per_test": "0.097216", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.021, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:operating:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.emb.operating", + "source_data": { + "dataset_name": "EMB - Operating", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "operating", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Operating).", + "metric_id": "vals_ai.emb.operating.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 5.395, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "operating", + "task_name": "Operating", + "dataset_type": "private", + "industry": "finance", + "raw_score": "5.395", + "raw_stderr": "1.935", + "latency": "52.473", + "cost_per_test": "0.034484", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.935, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:overall:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.emb.overall", + "source_data": { + "dataset_name": "EMB - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Overall).", + "metric_id": "vals_ai.emb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 8.629, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "8.629", + "raw_stderr": "1.088", + "latency": "67.007", + "cost_per_test": "0.069424", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.088, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:scratch:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.emb.scratch", + "source_data": { + "dataset_name": "EMB - Scratch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "scratch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Scratch).", + "metric_id": "vals_ai.emb.scratch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 9.072, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "scratch", + "task_name": "Scratch", + "dataset_type": "private", + "industry": "finance", + "raw_score": "9.072", + "raw_stderr": "1.194", + "latency": "63.021", + "cost_per_test": "0.064241", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.194, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:template:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.emb.template", + "source_data": { + "dataset_name": "EMB - Template", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "template", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Template).", + "metric_id": "vals_ai.emb.template.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 8.137, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "template", + "task_name": "Template", + "dataset_type": "private", + "industry": "finance", + "raw_score": "8.137", + "raw_stderr": "1.799", + "latency": "70.993", + "cost_per_test": "0.074608", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.799, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-flash-lite-preview/80bc0f89-ba19-4389-afdf-80fbb0cd9e43.json b/data/vals_ai/google/gemini-3.1-flash-lite-preview/80bc0f89-ba19-4389-afdf-80fbb0cd9e43.json new file mode 100644 index 0000000000000000000000000000000000000000..ad66a0798af6e1d050baedffeed55eabb071da16 --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-flash-lite-preview/80bc0f89-ba19-4389-afdf-80fbb0cd9e43.json @@ -0,0 +1,1071 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/google_gemini-3.1-flash-lite-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-flash-lite-preview", + "id": "google/gemini-3.1-flash-lite-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-flash-lite-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.166, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.166", + "raw_stderr": "0.942", + "latency": "21.666", + "cost_per_test": "0.000816", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.942, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.381, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.381", + "raw_stderr": "0.999", + "latency": "6.983", + "cost_per_test": "0.000925", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.999, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.989, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.989", + "raw_stderr": "0.851", + "latency": "12.463", + "cost_per_test": "0.00158", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.851, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.024, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.024", + "raw_stderr": "1.544", + "latency": "34.835", + "cost_per_test": "0.001091", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.544, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.336, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.336", + "raw_stderr": "1.062", + "latency": "6.013", + "cost_per_test": "0.000798", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.062, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.695, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.695", + "raw_stderr": "1.187", + "latency": "14.796", + "cost_per_test": "0.001948", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.187, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.318, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.318", + "raw_stderr": "1.39", + "latency": "6.968", + "cost_per_test": "0.000835", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.39, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.215, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.215", + "raw_stderr": "2.115", + "latency": "36.902", + "cost_per_test": "0.000893", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.115, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.023, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.023", + "raw_stderr": "1.305", + "latency": "10.251", + "cost_per_test": "0.001055", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.305, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.819, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.819", + "raw_stderr": "0.603", + "latency": "10.308", + "cost_per_test": "0.001361", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.603, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.926, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.926", + "raw_stderr": "1.266", + "latency": "5.446", + "cost_per_test": "0.000674", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.266, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.24, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.24", + "raw_stderr": "0.34", + "latency": "14.511", + "cost_per_test": "0.001052", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.34, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.764, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.764", + "raw_stderr": "1.729", + "latency": "6.827", + "cost_per_test": "0.00079", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.729, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.994, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.994", + "raw_stderr": "0.753", + "latency": "10.656", + "cost_per_test": "0.001305", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.753, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.714, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.714", + "raw_stderr": "1.239", + "latency": "19.037", + "cost_per_test": "0.000661", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.239, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-flash-lite-preview/96bf1d30-29d1-456a-9854-69a5056a6c5a.json b/data/vals_ai/google/gemini-3.1-flash-lite-preview/96bf1d30-29d1-456a-9854-69a5056a6c5a.json new file mode 100644 index 0000000000000000000000000000000000000000..2673ee02db9c1ad3684a71c6907b490a66052636 --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-flash-lite-preview/96bf1d30-29d1-456a-9854-69a5056a6c5a.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/google_gemini-3.1-flash-lite-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-flash-lite-preview", + "id": "google/gemini-3.1-flash-lite-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-flash-lite-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 60.425, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.425", + "raw_stderr": "1.398", + "latency": "7.644", + "cost_per_test": "0.00096", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.398, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 71.79, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.79", + "raw_stderr": "0.88", + "latency": "7.644", + "cost_per_test": "0.00096", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.88, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 83.156, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "83.156", + "raw_stderr": "1.07", + "latency": "7.644", + "cost_per_test": "0.00096", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.07, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-flash-lite-preview/a7914be1-d608-49ed-baf1-e75816ec5097.json b/data/vals_ai/google/gemini-3.1-flash-lite-preview/a7914be1-d608-49ed-baf1-e75816ec5097.json new file mode 100644 index 0000000000000000000000000000000000000000..cec528b99179e0b27c7573b93af1f04c6e3792f4 --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-flash-lite-preview/a7914be1-d608-49ed-baf1-e75816ec5097.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/code-migration/google_gemini-3.1-flash-lite-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Code Migration", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-flash-lite-preview", + "id": "google/gemini-3.1-flash-lite-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-flash-lite-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "code-migration:cli:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.code-migration.cli", + "source_data": { + "dataset_name": "Code Migration - CLI", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cli", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (CLI).", + "metric_id": "vals_ai.code-migration.cli.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 5.729, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cli", + "task_name": "CLI", + "dataset_type": "private", + "industry": "coding", + "raw_score": "5.729", + "raw_stderr": "1.363", + "latency": "87.657", + "cost_per_test": "0.052306", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.363, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "code-migration:cobol:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.code-migration.cobol", + "source_data": { + "dataset_name": "Code Migration - COBOL", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cobol", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (COBOL).", + "metric_id": "vals_ai.code-migration.cobol.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 1.25, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cobol", + "task_name": "COBOL", + "dataset_type": "private", + "industry": "coding", + "raw_score": "1.25", + "raw_stderr": "1.25", + "latency": "107.549", + "cost_per_test": "0.055194", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.25, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "code-migration:code_quality:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.code-migration.code_quality", + "source_data": { + "dataset_name": "Code Migration - Code Quality", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "code_quality", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Code Quality).", + "metric_id": "vals_ai.code-migration.code_quality.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 2.664, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "code_quality", + "task_name": "Code Quality", + "dataset_type": "private", + "industry": "coding", + "raw_score": "2.664", + "raw_stderr": "0.038", + "latency": "92.63", + "cost_per_test": "0.053028", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.038, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "code-migration:overall:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.code-migration.overall", + "source_data": { + "dataset_name": "Code Migration - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Overall).", + "metric_id": "vals_ai.code-migration.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 4.609, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "4.609", + "raw_stderr": "1.069", + "latency": "92.63", + "cost_per_test": "0.053028", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.069, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-flash-lite-preview/cac67baf-8280-4300-bd42-fa0581519b1d.json b/data/vals_ai/google/gemini-3.1-flash-lite-preview/cac67baf-8280-4300-bd42-fa0581519b1d.json new file mode 100644 index 0000000000000000000000000000000000000000..b9698d718164cd88cd436e6c80d4d3eaca17800d --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-flash-lite-preview/cac67baf-8280-4300-bd42-fa0581519b1d.json @@ -0,0 +1,450 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/google_gemini-3.1-flash-lite-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-flash-lite-preview", + "id": "google/gemini-3.1-flash-lite-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-flash-lite-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 88.472, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.472", + "raw_stderr": "0.714", + "latency": "2.904", + "cost_per_test": "0.000289", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.714, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.293, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.293", + "raw_stderr": "0.344", + "latency": "2.731", + "cost_per_test": "0.000309", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.344, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 89.371, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.371", + "raw_stderr": "0.57", + "latency": "2.392", + "cost_per_test": "0.000329", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.57, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.764, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.764", + "raw_stderr": "0.406", + "latency": "2.704", + "cost_per_test": "0.000295", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.406, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.514, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.514", + "raw_stderr": "0.833", + "latency": "2.351", + "cost_per_test": "0.000262", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.833, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.168, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.168", + "raw_stderr": "1.576", + "latency": "3.143", + "cost_per_test": "0.000287", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.576, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-flash-lite-preview/dec0b612-02a3-448e-9a8b-2a396457266f.json b/data/vals_ai/google/gemini-3.1-flash-lite-preview/dec0b612-02a3-448e-9a8b-2a396457266f.json new file mode 100644 index 0000000000000000000000000000000000000000..59fe9855268d7a7c67054dc24f70f79e726fecdc --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-flash-lite-preview/dec0b612-02a3-448e-9a8b-2a396457266f.json @@ -0,0 +1,418 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/google_gemini-3.1-flash-lite-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-flash-lite-preview", + "id": "google/gemini-3.1-flash-lite-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-flash-lite-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 60.839, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "60.839", + "raw_stderr": "1.666", + "latency": "4.708", + "cost_per_test": "0.022927", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.666, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 30.068, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "30.068", + "raw_stderr": "0.727", + "latency": "53.577", + "cost_per_test": "0.133469", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.727, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:overall:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 36.2, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "36.2", + "raw_stderr": "0.741", + "latency": "209.083", + "cost_per_test": "0.171524", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.741, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 57.843, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "57.843", + "raw_stderr": "4.889", + "latency": "136.077", + "cost_per_test": "0.106331" + }, + "uncertainty": { + "standard_error": { + "value": 4.889, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 34.082, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "34.082", + "raw_stderr": "0.991", + "latency": "278.342", + "cost_per_test": "0.084894", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.991, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "0", + "raw_stderr": "0", + "latency": "572.711", + "cost_per_test": "0.51", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-flash-lite-preview/e535bbfc-f2c4-4eec-b275-181a420c911f.json b/data/vals_ai/google/gemini-3.1-flash-lite-preview/e535bbfc-f2c4-4eec-b275-181a420c911f.json new file mode 100644 index 0000000000000000000000000000000000000000..ec862186daeefffa046f46270b3a79afe55bb4bb --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-flash-lite-preview/e535bbfc-f2c4-4eec-b275-181a420c911f.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/google_gemini-3.1-flash-lite-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-flash-lite-preview", + "id": "google/gemini-3.1-flash-lite-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-flash-lite-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 56.643, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.643", + "raw_stderr": "1.692", + "latency": "3.379", + "cost_per_test": "0.001726", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.692, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.606, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.606", + "raw_stderr": "1.668", + "latency": "4.003", + "cost_per_test": "0.048652", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.668, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 59.363, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.363", + "raw_stderr": "0.967", + "latency": "4.03", + "cost_per_test": "0.022927", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.967, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.839, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.839", + "raw_stderr": "1.666", + "latency": "4.708", + "cost_per_test": "0.018404", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.666, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-flash-lite-preview/e79447a9-91cd-47d9-a363-70792f9f7d34.json b/data/vals_ai/google/gemini-3.1-flash-lite-preview/e79447a9-91cd-47d9-a363-70792f9f7d34.json new file mode 100644 index 0000000000000000000000000000000000000000..915c6b70c6af99dff623e7bce3ecc8a8660a4516 --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-flash-lite-preview/e79447a9-91cd-47d9-a363-70792f9f7d34.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/google_gemini-3.1-flash-lite-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-flash-lite-preview", + "id": "google/gemini-3.1-flash-lite-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-flash-lite-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66.667", + "raw_stderr": "8.333", + "latency": "6193.118", + "cost_per_test": "1.888892", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 8.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 15.556, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "15.556", + "raw_stderr": "1.111", + "latency": "825.749", + "cost_per_test": "0.251852", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.111, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 41.818, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "41.818", + "raw_stderr": "1.05", + "latency": "450.409", + "cost_per_test": "0.137374", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.05, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 34.082, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "34.082", + "raw_stderr": "0.991", + "latency": "278.342", + "cost_per_test": "0.084894", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.991, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-flash-lite-preview/e8f7242c-c7a8-4685-8d2a-6b9d304d9e44.json b/data/vals_ai/google/gemini-3.1-flash-lite-preview/e8f7242c-c7a8-4685-8d2a-6b9d304d9e44.json new file mode 100644 index 0000000000000000000000000000000000000000..47ff450bc00023a2d9726324e5f2bbd4516e8e00 --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-flash-lite-preview/e8f7242c-c7a8-4685-8d2a-6b9d304d9e44.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/google_gemini-3.1-flash-lite-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-flash-lite-preview", + "id": "google/gemini-3.1-flash-lite-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-flash-lite-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:google/gemini-3.1-flash-lite-preview:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "572.711", + "cost_per_test": "0.51", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-pro-preview/0e917065-4a16-4162-9d8a-b766662266e7.json b/data/vals_ai/google/gemini-3.1-pro-preview/0e917065-4a16-4162-9d8a-b766662266e7.json new file mode 100644 index 0000000000000000000000000000000000000000..008a34ae55ba956c8de9f4b5bf53aea1516c5a27 --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-pro-preview/0e917065-4a16-4162-9d8a-b766662266e7.json @@ -0,0 +1,1224 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_research/google_gemini-3.1-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Legal Research Bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-pro-preview", + "id": "google/gemini-3.1-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_research:administrative_regulatory:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.legal_research.administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 24.242, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "24.242", + "raw_stderr": "5.316", + "latency": "219.629", + "cost_per_test": "1.225685", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.316, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:business_commercial:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.legal_research.business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 23.256, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "23.256", + "raw_stderr": "6.519", + "latency": "216.623", + "cost_per_test": "0.45295", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.519, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:civil_litigation:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.legal_research.civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 11.111, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "11.111", + "raw_stderr": "6.163", + "latency": "225.619", + "cost_per_test": "0.508937", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.163, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:constitutional_civil_rights:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.legal_research.constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 21.053, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "21.053", + "raw_stderr": "6.702", + "latency": "223.214", + "cost_per_test": "0.466165", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.702, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:criminal:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.legal_research.criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 19.298, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "19.298", + "raw_stderr": "5.274", + "latency": "217.466", + "cost_per_test": "0.48486", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.274, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:family:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.legal_research.family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 4.545, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "4.545", + "raw_stderr": "4.545", + "latency": "434.377", + "cost_per_test": "4.138935", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 4.545, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:health:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.legal_research.health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "33.333", + "raw_stderr": "8.754", + "latency": "232.285", + "cost_per_test": "1.804209", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 8.754, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:immigration:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.legal_research.immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 18.182, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "18.182", + "raw_stderr": "12.197", + "latency": "185.396", + "cost_per_test": "0.389491", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 12.197, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:overall:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.legal_research.overall", + "source_data": { + "dataset_name": "Legal Research Bench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Overall).", + "metric_id": "vals_ai.legal_research.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 20.673, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "legal", + "raw_score": "20.673", + "raw_stderr": "2.815", + "latency": "242.05", + "cost_per_test": "1.100145", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.815, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass", + "source_data": { + "dataset_name": "Legal Research Bench - Weighted Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Weighted Pass).", + "metric_id": "vals_ai.legal_research.weighted_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 76.61, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass", + "task_name": "Weighted Pass", + "dataset_type": "private", + "industry": "legal", + "raw_score": "76.61", + "raw_stderr": "1.602", + "latency": "242.05", + "cost_per_test": "1.100145", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.602, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__administrative_regulatory:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.weighted_pass__administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 79.215, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "79.215", + "raw_stderr": "2.633", + "latency": "219.629", + "cost_per_test": "1.225685", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.633, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__business_commercial:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.weighted_pass__business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 80.773, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "80.773", + "raw_stderr": "3.085", + "latency": "216.623", + "cost_per_test": "0.45295", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.085, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__civil_litigation:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.weighted_pass__civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 82.707, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "82.707", + "raw_stderr": "2.966", + "latency": "225.619", + "cost_per_test": "0.508937", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.966, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__constitutional_civil_rights:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 79.637, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "79.637", + "raw_stderr": "2.872", + "latency": "223.214", + "cost_per_test": "0.466165", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.872, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__criminal:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.weighted_pass__criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 76.177, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "76.177", + "raw_stderr": "2.892", + "latency": "217.466", + "cost_per_test": "0.48486", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.892, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__family:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.weighted_pass__family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 58.209, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "58.209", + "raw_stderr": "6.265", + "latency": "434.377", + "cost_per_test": "4.138935", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.265, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__health:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.weighted_pass__health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 83.436, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "83.436", + "raw_stderr": "3.06", + "latency": "232.285", + "cost_per_test": "1.804209", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.06, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__immigration:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.weighted_pass__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 72.894, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "72.894", + "raw_stderr": "9.888", + "latency": "185.396", + "cost_per_test": "0.389491", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 9.888, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-pro-preview/1ef1ad93-c13d-4454-af8e-28281fa95d80.json b/data/vals_ai/google/gemini-3.1-pro-preview/1ef1ad93-c13d-4454-af8e-28281fa95d80.json new file mode 100644 index 0000000000000000000000000000000000000000..26292367385dd5d4a7654425ff9b7489559be202 --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-pro-preview/1ef1ad93-c13d-4454-af8e-28281fa95d80.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/public-benefits-bench/google_gemini-3.1-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Public Benefits Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "social mobility", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-pro-preview", + "id": "google/gemini-3.1-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "public-benefits-bench:overall:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.public-benefits-bench.overall", + "source_data": { + "dataset_name": "Public Benefits Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Public Benefits Bench v1.1 (Overall).", + "metric_id": "vals_ai.public-benefits-bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "score_details": { + "score": 53.789, + "details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "social mobility", + "raw_score": "53.789", + "raw_stderr": "1.297", + "latency": "350.286", + "cost_per_test": "0.376768", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.297, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-pro-preview/30b9c8c4-eb7e-4636-9d25-92b2153cea07.json b/data/vals_ai/google/gemini-3.1-pro-preview/30b9c8c4-eb7e-4636-9d25-92b2153cea07.json new file mode 100644 index 0000000000000000000000000000000000000000..567fd1c5ef74f4efe6a18c176abad72aed769cc5 --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-pro-preview/30b9c8c4-eb7e-4636-9d25-92b2153cea07.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/google_gemini-3.1-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-pro-preview", + "id": "google/gemini-3.1-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.839, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.839", + "raw_stderr": "1.666", + "latency": "19.092", + "cost_per_test": "0.012811", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.666, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 66.9, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.9", + "raw_stderr": "1.607", + "latency": "28.935", + "cost_per_test": "0.387473", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.607, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 64.491, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.491", + "raw_stderr": "0.942", + "latency": "25.388", + "cost_per_test": "0.181957", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.942, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 65.734, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.734", + "raw_stderr": "1.62", + "latency": "28.137", + "cost_per_test": "0.145587", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.62, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-pro-preview/39c1bb93-aa0b-43e1-b16a-fd07bf303678.json b/data/vals_ai/google/gemini-3.1-pro-preview/39c1bb93-aa0b-43e1-b16a-fd07bf303678.json new file mode 100644 index 0000000000000000000000000000000000000000..8c89192cd3e85871557f276b36980d1dc4ef1743 --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-pro-preview/39c1bb93-aa0b-43e1-b16a-fd07bf303678.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/google_gemini-3.1-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-pro-preview", + "id": "google/gemini-3.1-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 32.034, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "32.034", + "raw_stderr": "4.342", + "latency": "1211.906", + "cost_per_test": "3.825202", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 4.342, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-pro-preview/4b9aa92b-304e-4be2-a9ae-1f35494b63c0.json b/data/vals_ai/google/gemini-3.1-pro-preview/4b9aa92b-304e-4be2-a9ae-1f35494b63c0.json new file mode 100644 index 0000000000000000000000000000000000000000..d64f12529367b4d69ebe2595d3b1ddd330bfa281 --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-pro-preview/4b9aa92b-304e-4be2-a9ae-1f35494b63c0.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/code-migration/google_gemini-3.1-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Code Migration", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-pro-preview", + "id": "google/gemini-3.1-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "code-migration:cli:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.code-migration.cli", + "source_data": { + "dataset_name": "Code Migration - CLI", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cli", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (CLI).", + "metric_id": "vals_ai.code-migration.cli.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 10.833, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cli", + "task_name": "CLI", + "dataset_type": "private", + "industry": "coding", + "raw_score": "10.833", + "raw_stderr": "1.468", + "latency": "735.951", + "cost_per_test": "1.521038", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.468, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "code-migration:cobol:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.code-migration.cobol", + "source_data": { + "dataset_name": "Code Migration - COBOL", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cobol", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (COBOL).", + "metric_id": "vals_ai.code-migration.cobol.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 36.753, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cobol", + "task_name": "COBOL", + "dataset_type": "private", + "industry": "coding", + "raw_score": "36.753", + "raw_stderr": "15.105", + "latency": "674.444", + "cost_per_test": "1.895374", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 15.105, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "code-migration:code_quality:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.code-migration.code_quality", + "source_data": { + "dataset_name": "Code Migration - Code Quality", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "code_quality", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Code Quality).", + "metric_id": "vals_ai.code-migration.code_quality.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 2.538, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "code_quality", + "task_name": "Code Quality", + "dataset_type": "private", + "industry": "coding", + "raw_score": "2.538", + "raw_stderr": "0.039", + "latency": "720.574", + "cost_per_test": "1.614622", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.039, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "code-migration:overall:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.code-migration.overall", + "source_data": { + "dataset_name": "Code Migration - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Overall).", + "metric_id": "vals_ai.code-migration.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 17.313, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "17.313", + "raw_stderr": "3.933", + "latency": "720.574", + "cost_per_test": "1.614622", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.933, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-pro-preview/595fd4e4-7602-4f01-9677-483813ceb6ea.json b/data/vals_ai/google/gemini-3.1-pro-preview/595fd4e4-7602-4f01-9677-483813ceb6ea.json new file mode 100644 index 0000000000000000000000000000000000000000..828c2a0e91d8235a457f2fdf82fb1ca5df769eb1 --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-pro-preview/595fd4e4-7602-4f01-9677-483813ceb6ea.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/google_gemini-3.1-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-pro-preview", + "id": "google/gemini-3.1-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 76.114, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "76.114", + "raw_stderr": "1.915", + "latency": "69.139", + "cost_per_test": "0.097954", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.915, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-pro-preview/630462da-4c09-4e34-a791-7aedb8b9b46d.json b/data/vals_ai/google/gemini-3.1-pro-preview/630462da-4c09-4e34-a791-7aedb8b9b46d.json new file mode 100644 index 0000000000000000000000000000000000000000..cbbac2e2cb66d4107ecb5122e8340447e1707f64 --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-pro-preview/630462da-4c09-4e34-a791-7aedb8b9b46d.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/google_gemini-3.1-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-pro-preview", + "id": "google/gemini-3.1-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 26.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "26", + "raw_stderr": "4.386", + "latency": "2166.981", + "cost_per_test": "0.525953", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 4.386, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-pro-preview/63c13650-3d8b-456a-a941-1c4f637e5a23.json b/data/vals_ai/google/gemini-3.1-pro-preview/63c13650-3d8b-456a-a941-1c4f637e5a23.json new file mode 100644 index 0000000000000000000000000000000000000000..64988d668fd535677b0f4e3456ddd0451bbdd265 --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-pro-preview/63c13650-3d8b-456a-a941-1c4f637e5a23.json @@ -0,0 +1,418 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/google_gemini-3.1-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-pro-preview", + "id": "google/gemini-3.1-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 65.734, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "65.734", + "raw_stderr": "1.62", + "latency": "28.137", + "cost_per_test": "0.145587", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.62, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 42.968, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "42.968", + "raw_stderr": "1.931", + "latency": "196.068", + "cost_per_test": "1.378409", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.931, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:overall:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 53.77, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "53.77", + "raw_stderr": "1.591", + "latency": "435.112", + "cost_per_test": "1.350398", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.591, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 72.549, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "72.549", + "raw_stderr": "4.419", + "latency": "329.202", + "cost_per_test": "0.820116" + }, + "uncertainty": { + "standard_error": { + "value": 4.419, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 70.787, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "70.787", + "raw_stderr": "1.124", + "latency": "410.249", + "cost_per_test": "0.582677", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.124, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 34.214, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "34.214", + "raw_stderr": "6.444", + "latency": "1211.906", + "cost_per_test": "3.825202", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.444, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-pro-preview/6ab9ecc1-6950-4c3e-ab4d-050b72773677.json b/data/vals_ai/google/gemini-3.1-pro-preview/6ab9ecc1-6950-4c3e-ab4d-050b72773677.json new file mode 100644 index 0000000000000000000000000000000000000000..43aa02f20d9a59531b8e04b4c75cfcc69d41d9e5 --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-pro-preview/6ab9ecc1-6950-4c3e-ab4d-050b72773677.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/google_gemini-3.1-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-pro-preview", + "id": "google/gemini-3.1-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 95.96, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "95.96", + "raw_stderr": "1.399", + "latency": "67.495", + "cost_per_test": "0.085026", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.399, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 94.949, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.949", + "raw_stderr": "1.556", + "latency": "64.024", + "cost_per_test": "0.077857", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.556, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 95.454, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "95.454", + "raw_stderr": "1.046", + "latency": "65.76", + "cost_per_test": "0.081442", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.046, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-pro-preview/6cb60215-4bc3-4c43-bf52-fb9df9a6a1a6.json b/data/vals_ai/google/gemini-3.1-pro-preview/6cb60215-4bc3-4c43-bf52-fb9df9a6a1a6.json new file mode 100644 index 0000000000000000000000000000000000000000..b219c5e7a970a782ce256a2eac5ab0aa7f706c69 --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-pro-preview/6cb60215-4bc3-4c43-bf52-fb9df9a6a1a6.json @@ -0,0 +1,1071 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/google_gemini-3.1-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-pro-preview", + "id": "google/gemini-3.1-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 95.258, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "95.258", + "raw_stderr": "0.794", + "latency": "15.298", + "cost_per_test": "0.016695", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.794, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.409, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.409", + "raw_stderr": "0.883", + "latency": "19.549", + "cost_per_test": "0.023771", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.883, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.286, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.286", + "raw_stderr": "0.744", + "latency": "36.067", + "cost_per_test": "0.044854", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.744, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.683, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.683", + "raw_stderr": "1.286", + "latency": "21.033", + "cost_per_test": "0.025749", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.286, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.365, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.365", + "raw_stderr": "0.857", + "latency": "14.452", + "cost_per_test": "0.017048", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.857, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.029, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.029", + "raw_stderr": "1.043", + "latency": "64.579", + "cost_per_test": "0.079443", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.043, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.43, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.43", + "raw_stderr": "1.197", + "latency": "19.713", + "cost_per_test": "0.018611", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.197, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.564, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.564", + "raw_stderr": "1.801", + "latency": "15.026", + "cost_per_test": "0.018593", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.801, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.105, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.105", + "raw_stderr": "1.102", + "latency": "22.32", + "cost_per_test": "0.025774", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.102, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 96.003, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "96.003", + "raw_stderr": "0.533", + "latency": "24.464", + "cost_per_test": "0.026993", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.533, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.636, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.636", + "raw_stderr": "1.044", + "latency": "15.066", + "cost_per_test": "0.016267", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.044, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.987, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.987", + "raw_stderr": "0.284", + "latency": "23.912", + "cost_per_test": "0.027545", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.284, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.381, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.381", + "raw_stderr": "1.32", + "latency": "15.572", + "cost_per_test": "0.017428", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.32, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.688, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.688", + "raw_stderr": "0.622", + "latency": "35.667", + "cost_per_test": "0.038713", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.622, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.98, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.98", + "raw_stderr": "0.961", + "latency": "15.965", + "cost_per_test": "0.015685", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.961, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-pro-preview/6ef73c1b-86ff-4e91-85c0-a11f1286c97d.json b/data/vals_ai/google/gemini-3.1-pro-preview/6ef73c1b-86ff-4e91-85c0-a11f1286c97d.json new file mode 100644 index 0000000000000000000000000000000000000000..a8d99c2174d6eaaa2b39e5f2c1f758d8b3ec6dbc --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-pro-preview/6ef73c1b-86ff-4e91-85c0-a11f1286c97d.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/google_gemini-3.1-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-pro-preview", + "id": "google/gemini-3.1-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 76.391, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "76.391", + "raw_stderr": "1.197", + "latency": "24.789", + "cost_per_test": "0.024731", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.197, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 69.396, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "69.396", + "raw_stderr": "0.908", + "latency": "23.616", + "cost_per_test": "0.023542", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.908, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 62.401, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.401", + "raw_stderr": "1.366", + "latency": "22.444", + "cost_per_test": "0.022352", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.366, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-pro-preview/6f6ab7ef-2311-4da1-be00-b73327f9777f.json b/data/vals_ai/google/gemini-3.1-pro-preview/6f6ab7ef-2311-4da1-be00-b73327f9777f.json new file mode 100644 index 0000000000000000000000000000000000000000..176e84d72cc332f7e46d7e17ab36aae66f7566da --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-pro-preview/6f6ab7ef-2311-4da1-be00-b73327f9777f.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/google_gemini-3.1-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-pro-preview", + "id": "google/gemini-3.1-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 91.667, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "91.667", + "raw_stderr": "8.333", + "latency": "9128.048", + "cost_per_test": "12.964574", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 8.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 60.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "60", + "raw_stderr": "1.925", + "latency": "1217.073", + "cost_per_test": "1.72861", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.925, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 75.152, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "75.152", + "raw_stderr": "1.212", + "latency": "663.858", + "cost_per_test": "0.942878", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.212, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 70.787, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "70.787", + "raw_stderr": "1.124", + "latency": "410.249", + "cost_per_test": "0.582677", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.124, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-pro-preview/7de4743c-a141-4059-ae36-b92bb8785534.json b/data/vals_ai/google/gemini-3.1-pro-preview/7de4743c-a141-4059-ae36-b92bb8785534.json new file mode 100644 index 0000000000000000000000000000000000000000..267b64bc3443370932417aec7f5453ece2bd58d5 --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-pro-preview/7de4743c-a141-4059-ae36-b92bb8785534.json @@ -0,0 +1,696 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/emb/google_gemini-3.1-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - EMB", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-pro-preview", + "id": "google/gemini-3.1-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "emb:comps:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.emb.comps", + "source_data": { + "dataset_name": "EMB - Comps", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "comps", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Comps).", + "metric_id": "vals_ai.emb.comps.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 44.505, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "comps", + "task_name": "Comps", + "dataset_type": "private", + "industry": "finance", + "raw_score": "44.505", + "raw_stderr": "7.615", + "latency": "768.417", + "cost_per_test": "7.466867", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 7.615, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:dataroom:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.emb.dataroom", + "source_data": { + "dataset_name": "EMB - Dataroom", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dataroom", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Dataroom).", + "metric_id": "vals_ai.emb.dataroom.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 79.17, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dataroom", + "task_name": "Dataroom", + "dataset_type": "private", + "industry": "finance", + "raw_score": "79.17", + "raw_stderr": "4.883", + "latency": "1109.179", + "cost_per_test": "1.64075", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 4.883, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:dcf:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.emb.dcf", + "source_data": { + "dataset_name": "EMB - DCF", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dcf", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (DCF).", + "metric_id": "vals_ai.emb.dcf.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 46.388, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dcf", + "task_name": "DCF", + "dataset_type": "private", + "industry": "finance", + "raw_score": "46.388", + "raw_stderr": "7.715", + "latency": "584.588", + "cost_per_test": "2.495552", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 7.715, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:lbo:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.emb.lbo", + "source_data": { + "dataset_name": "EMB - LBO", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "lbo", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (LBO).", + "metric_id": "vals_ai.emb.lbo.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 37.967, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "lbo", + "task_name": "LBO", + "dataset_type": "private", + "industry": "finance", + "raw_score": "37.967", + "raw_stderr": "4.006", + "latency": "784.347", + "cost_per_test": "4.762336", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 4.006, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:ma:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.emb.ma", + "source_data": { + "dataset_name": "EMB - M&A", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "ma", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (M&A).", + "metric_id": "vals_ai.emb.ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 58.833, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "ma", + "task_name": "M&A", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.833", + "raw_stderr": "7.829", + "latency": "646.47", + "cost_per_test": "6.070452", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 7.829, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:misc:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.emb.misc", + "source_data": { + "dataset_name": "EMB - Misc", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "misc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Misc).", + "metric_id": "vals_ai.emb.misc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 43.366, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "misc", + "task_name": "Misc", + "dataset_type": "private", + "industry": "finance", + "raw_score": "43.366", + "raw_stderr": "8.142", + "latency": "546.97", + "cost_per_test": "1.744042", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 8.142, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:operating:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.emb.operating", + "source_data": { + "dataset_name": "EMB - Operating", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "operating", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Operating).", + "metric_id": "vals_ai.emb.operating.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 58.087, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "operating", + "task_name": "Operating", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.087", + "raw_stderr": "6.208", + "latency": "506.512", + "cost_per_test": "1.560798", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.208, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:overall:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.emb.overall", + "source_data": { + "dataset_name": "EMB - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Overall).", + "metric_id": "vals_ai.emb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 52.616, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.616", + "raw_stderr": "2.974", + "latency": "708.582", + "cost_per_test": "3.769246", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.974, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:scratch:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.emb.scratch", + "source_data": { + "dataset_name": "EMB - Scratch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "scratch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Scratch).", + "metric_id": "vals_ai.emb.scratch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 57.704, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "scratch", + "task_name": "Scratch", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.704", + "raw_stderr": "2.832", + "latency": "624.193", + "cost_per_test": "3.148008", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.832, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:template:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.emb.template", + "source_data": { + "dataset_name": "EMB - Template", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "template", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Template).", + "metric_id": "vals_ai.emb.template.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 47.213, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "template", + "task_name": "Template", + "dataset_type": "private", + "industry": "finance", + "raw_score": "47.213", + "raw_stderr": "4.677", + "latency": "792.971", + "cost_per_test": "4.390484", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 4.677, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-pro-preview/7e442830-453a-49cd-8b23-439d179dd65b.json b/data/vals_ai/google/gemini-3.1-pro-preview/7e442830-453a-49cd-8b23-439d179dd65b.json new file mode 100644 index 0000000000000000000000000000000000000000..845dfaec594a6ec7bc0601270690f720394ae041 --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-pro-preview/7e442830-453a-49cd-8b23-439d179dd65b.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/google_gemini-3.1-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-pro-preview", + "id": "google/gemini-3.1-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 60.749, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.749", + "raw_stderr": "1.393", + "latency": "41.931", + "cost_per_test": "0.03149", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.393, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 72.882, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "72.882", + "raw_stderr": "0.863", + "latency": "41.931", + "cost_per_test": "0.03149", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.863, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 85.016, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "85.016", + "raw_stderr": "1.019", + "latency": "41.931", + "cost_per_test": "0.03149", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.019, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-pro-preview/8ba1f910-3f8e-4584-84e4-9f189f22b3f1.json b/data/vals_ai/google/gemini-3.1-pro-preview/8ba1f910-3f8e-4584-84e4-9f189f22b3f1.json new file mode 100644 index 0000000000000000000000000000000000000000..c9783161cb214da75b3b06a8a59a2d36110cdcb3 --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-pro-preview/8ba1f910-3f8e-4584-84e4-9f189f22b3f1.json @@ -0,0 +1,381 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/google_gemini-3.1-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-pro-preview", + "id": "google/gemini-3.1-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 42.857, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "42.857", + "raw_stderr": "7.729", + "latency": "386.06", + "cost_per_test": "0.999095", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 7.729, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 77.778, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "77.778", + "raw_stderr": "2.578", + "latency": "338.098", + "cost_per_test": "0.843349", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.578, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 88.66, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "88.66", + "raw_stderr": "2.282", + "latency": "259.783", + "cost_per_test": "0.652263", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.282, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "424.841", + "cost_per_test": "1.025173", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:overall:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 78.8, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "78.8", + "raw_stderr": "1.83", + "latency": "312.261", + "cost_per_test": "0.783381", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.83, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-pro-preview/9bc40fb8-2bf2-45fc-acec-9a995139d100.json b/data/vals_ai/google/gemini-3.1-pro-preview/9bc40fb8-2bf2-45fc-acec-9a995139d100.json new file mode 100644 index 0000000000000000000000000000000000000000..a73a677754c62f593f0ef08901519272655c8190 --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-pro-preview/9bc40fb8-2bf2-45fc-acec-9a995139d100.json @@ -0,0 +1,550 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_multimodal_index/google_gemini-3.1-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Multimodal Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-pro-preview", + "id": "google/gemini-3.1-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_multimodal_index:corp_fin_v2:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.vals_multimodal_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Multimodal Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (CorpFin v2).", + "metric_id": "vals_ai.vals_multimodal_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 65.734, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "65.734", + "raw_stderr": "1.62", + "latency": "28.137", + "cost_per_test": "0.145587", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.62, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:finance_agent:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.vals_multimodal_index.finance_agent", + "source_data": { + "dataset_name": "Vals Multimodal Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_multimodal_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 42.968, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "42.968", + "raw_stderr": "1.931", + "latency": "196.068", + "cost_per_test": "1.378409", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.931, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:mortgage_tax:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.vals_multimodal_index.mortgage_tax", + "source_data": { + "dataset_name": "Vals Multimodal Index - Mortgage Tax", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "mortgage_tax", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Mortgage Tax).", + "metric_id": "vals_ai.vals_multimodal_index.mortgage_tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 69.396, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "mortgage_tax", + "task_name": "Mortgage Tax", + "dataset_type": "private", + "industry": "index", + "raw_score": "69.396", + "raw_stderr": "0.908", + "latency": "23.616", + "cost_per_test": "0.023542", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.908, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:overall:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.vals_multimodal_index.overall", + "source_data": { + "dataset_name": "Vals Multimodal Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Overall).", + "metric_id": "vals_ai.vals_multimodal_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 56.068, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "56.068", + "raw_stderr": "1.406", + "latency": "322.907", + "cost_per_test": "0.977777", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.406, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:sage:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.vals_multimodal_index.sage", + "source_data": { + "dataset_name": "Vals Multimodal Index - SAGE", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "sage", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SAGE).", + "metric_id": "vals_ai.vals_multimodal_index.sage.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 48.677, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "sage", + "task_name": "SAGE", + "dataset_type": "private", + "industry": "index", + "raw_score": "48.677", + "raw_stderr": "3.293", + "latency": "61.174", + "cost_per_test": "0.068908", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.293, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:swebench:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.vals_multimodal_index.swebench", + "source_data": { + "dataset_name": "Vals Multimodal Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SWE-bench).", + "metric_id": "vals_ai.vals_multimodal_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 72.549, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "72.549", + "raw_stderr": "4.419", + "latency": "329.202", + "cost_per_test": "0.820116" + }, + "uncertainty": { + "standard_error": { + "value": 4.419, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:terminal_bench_2_1:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.vals_multimodal_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Multimodal Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_multimodal_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 70.787, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "70.787", + "raw_stderr": "1.124", + "latency": "410.249", + "cost_per_test": "0.582677", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.124, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:vibe_code_bench:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.vals_multimodal_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Multimodal Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_multimodal_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 34.214, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "34.214", + "raw_stderr": "6.444", + "latency": "1211.906", + "cost_per_test": "3.825202", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.444, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-pro-preview/aa160ef8-884a-4e28-b8d7-1733c99cfd72.json b/data/vals_ai/google/gemini-3.1-pro-preview/aa160ef8-884a-4e28-b8d7-1733c99cfd72.json new file mode 100644 index 0000000000000000000000000000000000000000..de6bc30a6d298183680d1db419c1d65d799ac60a --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-pro-preview/aa160ef8-884a-4e28-b8d7-1733c99cfd72.json @@ -0,0 +1,105 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/google_gemini-3.1-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-pro-preview", + "id": "google/gemini-3.1-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 88.208, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.208", + "raw_stderr": "0.775", + "latency": "76.99", + "cost_per_test": "0.064084", + "temperature": "1", + "max_output_tokens": "65535", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.775, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65535, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-pro-preview/b463c33e-c2d1-4a5b-8b58-8e3d604d6ef6.json b/data/vals_ai/google/gemini-3.1-pro-preview/b463c33e-c2d1-4a5b-8b58-8e3d604d6ef6.json new file mode 100644 index 0000000000000000000000000000000000000000..4cd76d50278093a845dd06f899bb4323fd69e1ec --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-pro-preview/b463c33e-c2d1-4a5b-8b58-8e3d604d6ef6.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/programbench/google_gemini-3.1-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProgramBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-pro-preview", + "id": "google/gemini-3.1-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "programbench:almost:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.programbench.almost", + "source_data": { + "dataset_name": "ProgramBench - Almost Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "almost", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Almost Resolved).", + "metric_id": "vals_ai.programbench.almost.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 1.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "almost", + "task_name": "Almost Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "1", + "raw_stderr": "0.705", + "latency": "755.928", + "cost_per_test": "2.629095", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.705, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "programbench:overall:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.programbench.overall", + "source_data": { + "dataset_name": "ProgramBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Overall).", + "metric_id": "vals_ai.programbench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "755.928", + "cost_per_test": "2.629095", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "programbench:partial:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.programbench.partial", + "source_data": { + "dataset_name": "ProgramBench - Raw Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "partial", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Raw Pass Rate).", + "metric_id": "vals_ai.programbench.partial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 39.468, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "partial", + "task_name": "Raw Pass Rate", + "dataset_type": "public", + "industry": "coding", + "raw_score": "39.468", + "raw_stderr": "1.809", + "latency": "755.928", + "cost_per_test": "2.629095", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.809, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "programbench:strict:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.programbench.strict", + "source_data": { + "dataset_name": "ProgramBench - Fully Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "strict", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Fully Resolved).", + "metric_id": "vals_ai.programbench.strict.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "strict", + "task_name": "Fully Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "755.928", + "cost_per_test": "2.629095", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-pro-preview/c3f69fbb-5758-4914-8cd4-d0cbaa27aa15.json b/data/vals_ai/google/gemini-3.1-pro-preview/c3f69fbb-5758-4914-8cd4-d0cbaa27aa15.json new file mode 100644 index 0000000000000000000000000000000000000000..6f38c5e3d02b403356d794b2634568bfdbaa208a --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-pro-preview/c3f69fbb-5758-4914-8cd4-d0cbaa27aa15.json @@ -0,0 +1,195 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/cyber/google_gemini-3.1-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CyberBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "beta", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-pro-preview", + "id": "google/gemini-3.1-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "cyber:overall:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.cyber.overall", + "source_data": { + "dataset_name": "CyberBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (Overall).", + "metric_id": "vals_ai.cyber.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 36.441, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "beta", + "raw_score": "36.441", + "raw_stderr": "4.175", + "latency": "1031.764", + "cost_per_test": "2.10256", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 4.175, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "cyber:patch:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.cyber.patch", + "source_data": { + "dataset_name": "CyberBench - Patch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "patch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (Patch).", + "metric_id": "vals_ai.cyber.patch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 69.492, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "patch", + "task_name": "Patch", + "dataset_type": "private", + "industry": "beta", + "raw_score": "69.492", + "raw_stderr": "5.994", + "latency": "1710.084", + "cost_per_test": "4.045512", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.994, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "cyber:poc:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.cyber.poc", + "source_data": { + "dataset_name": "CyberBench - PoC", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "poc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (PoC).", + "metric_id": "vals_ai.cyber.poc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 3.39, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "poc", + "task_name": "PoC", + "dataset_type": "private", + "industry": "beta", + "raw_score": "3.39", + "raw_stderr": "2.356", + "latency": "353.444", + "cost_per_test": "0.159608", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.356, + "method": "vals_reported" + } + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-pro-preview/d589b740-51b2-4c49-a19a-6e8d8652b396.json b/data/vals_ai/google/gemini-3.1-pro-preview/d589b740-51b2-4c49-a19a-6e8d8652b396.json new file mode 100644 index 0000000000000000000000000000000000000000..4b740c998432043bd31e4766017c717e5ec8b9da --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-pro-preview/d589b740-51b2-4c49-a19a-6e8d8652b396.json @@ -0,0 +1,762 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/google_gemini-3.1-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-pro-preview", + "id": "google/gemini-3.1-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 38.361, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "38.361", + "raw_stderr": "1.026", + "latency": "188.406", + "cost_per_test": "1.417992", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.026, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 30.879, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "30.879", + "raw_stderr": "1.348", + "latency": "208.258", + "cost_per_test": "1.591906", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.348, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 36.66, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "36.66", + "raw_stderr": "2.083", + "latency": "268.379", + "cost_per_test": "2.383117", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.083, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 46.181, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "46.181", + "raw_stderr": "1.857", + "latency": "191.853", + "cost_per_test": "1.695327", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.857, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 64.654, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.654", + "raw_stderr": "0.648", + "latency": "154.84", + "cost_per_test": "0.509894", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.648, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 15.101, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "15.101", + "raw_stderr": "1.726", + "latency": "287.693", + "cost_per_test": "1.714516", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.726, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 45.206, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "45.206", + "raw_stderr": "2.644", + "latency": "131.886", + "cost_per_test": "1.237141", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.644, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 67.018, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.018", + "raw_stderr": "1.376", + "latency": "190.991", + "cost_per_test": "1.723303", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.376, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 52.842, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.842", + "raw_stderr": "1.717", + "latency": "134.603", + "cost_per_test": "0.657507", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.717, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:overall:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 42.982, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "42.982", + "raw_stderr": "1.206", + "latency": "208.258", + "cost_per_test": "1.591906", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.206, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 20.816, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "20.816", + "raw_stderr": "3.014", + "latency": "344.984", + "cost_per_test": "3.095248", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.014, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-pro-preview/e89f4bde-019f-44ee-8114-cb3f2498410c.json b/data/vals_ai/google/gemini-3.1-pro-preview/e89f4bde-019f-44ee-8114-cb3f2498410c.json new file mode 100644 index 0000000000000000000000000000000000000000..16d146396eaeaf94a71c023ec6f9660c41ceafa5 --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-pro-preview/e89f4bde-019f-44ee-8114-cb3f2498410c.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/google_gemini-3.1-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-pro-preview", + "id": "google/gemini-3.1-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 99.068, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "99.068", + "raw_stderr": "0.535", + "latency": "19.291", + "cost_per_test": "0.021104", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:hard:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 76.571, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "76.571", + "raw_stderr": "2.264", + "latency": "175.343", + "cost_per_test": "0.198871", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.264, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:medium:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 89.817, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "89.817", + "raw_stderr": "1.545", + "latency": "72.04", + "cost_per_test": "0.083466", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.545, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:overall:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 88.485, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "88.485", + "raw_stderr": "0.931", + "latency": "88.891", + "cost_per_test": "0.101147", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.931, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-pro-preview/fd5e3a7d-af06-4ee3-857a-1edc5b62f311.json b/data/vals_ai/google/gemini-3.1-pro-preview/fd5e3a7d-af06-4ee3-857a-1edc5b62f311.json new file mode 100644 index 0000000000000000000000000000000000000000..63609fbb8e277c15f01f7b446aa4da170eb8e997 --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-pro-preview/fd5e3a7d-af06-4ee3-857a-1edc5b62f311.json @@ -0,0 +1,450 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/google_gemini-3.1-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-pro-preview", + "id": "google/gemini-3.1-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 89.257, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.257", + "raw_stderr": "0.611", + "latency": "10.23", + "cost_per_test": "0.005926", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.611, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.823, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.823", + "raw_stderr": "0.367", + "latency": "9.143", + "cost_per_test": "0.006094", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.367, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 91.48, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "91.48", + "raw_stderr": "0.53", + "latency": "8.723", + "cost_per_test": "0.006807", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.53, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.398, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.398", + "raw_stderr": "0.329", + "latency": "10.062", + "cost_per_test": "0.006771", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.329, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 84.136, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.136", + "raw_stderr": "0.766", + "latency": "9.066", + "cost_per_test": "0.006197", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.766, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 88.294, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.294", + "raw_stderr": "1.152", + "latency": "13.146", + "cost_per_test": "0.00883", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.152, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.1-pro-preview/fd84db2d-46de-43aa-94dd-a1cf0db500a9.json b/data/vals_ai/google/gemini-3.1-pro-preview/fd84db2d-46de-43aa-94dd-a1cf0db500a9.json new file mode 100644 index 0000000000000000000000000000000000000000..49adf2a982280f5ec56ea963d0c83974de8bd4f5 --- /dev/null +++ b/data/vals_ai/google/gemini-3.1-pro-preview/fd84db2d-46de-43aa-94dd-a1cf0db500a9.json @@ -0,0 +1,366 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/google_gemini-3.1-pro-preview/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.1-pro-preview", + "id": "google/gemini-3.1-pro-preview", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.1-pro-preview", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 39.683, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "39.683", + "raw_stderr": "6.148", + "latency": "63.765", + "cost_per_test": "0.072766", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.148, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:automata:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 32.143, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "32.143", + "raw_stderr": "5.937", + "latency": "62.434", + "cost_per_test": "0.068462", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.937, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:calculus:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 78.395, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "78.395", + "raw_stderr": "5.215", + "latency": "57.524", + "cost_per_test": "0.070287", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.215, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 74.205, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "74.205", + "raw_stderr": "4.952", + "latency": "57.323", + "cost_per_test": "0.065497", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 4.952, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:overall:google/gemini-3.1-pro-preview:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 48.677, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "48.677", + "raw_stderr": "3.293", + "latency": "61.174", + "cost_per_test": "0.068908", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.293, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.5-flash/14867896-68d0-44af-8ebf-abf7c97a377c.json b/data/vals_ai/google/gemini-3.5-flash/14867896-68d0-44af-8ebf-abf7c97a377c.json new file mode 100644 index 0000000000000000000000000000000000000000..195d72aa33e0d58b0d18e57e2fe1c2eb6a02ba10 --- /dev/null +++ b/data/vals_ai/google/gemini-3.5-flash/14867896-68d0-44af-8ebf-abf7c97a377c.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/google_gemini-3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.5-flash", + "id": "google/gemini-3.5-flash", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.5-flash", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 76.574, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "76.574", + "raw_stderr": "1.923", + "latency": "57.075", + "cost_per_test": "0.166341", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.923, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.5-flash/15f98f4b-2989-499f-9d20-950bcb001d18.json b/data/vals_ai/google/gemini-3.5-flash/15f98f4b-2989-499f-9d20-950bcb001d18.json new file mode 100644 index 0000000000000000000000000000000000000000..e319b304cb516cb18fbe8a1a9abf63126da2704b --- /dev/null +++ b/data/vals_ai/google/gemini-3.5-flash/15f98f4b-2989-499f-9d20-950bcb001d18.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/google_gemini-3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.5-flash", + "id": "google/gemini-3.5-flash", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.5-flash", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 48.683, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "48.683", + "raw_stderr": "4.727", + "latency": "893.074", + "cost_per_test": "2.540385", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 4.727, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.5-flash/17b16189-a065-4f20-b90c-9eeab8cb3bb8.json b/data/vals_ai/google/gemini-3.5-flash/17b16189-a065-4f20-b90c-9eeab8cb3bb8.json new file mode 100644 index 0000000000000000000000000000000000000000..984826bffe39ed8b4aef7a4d85cce40190c5e26f --- /dev/null +++ b/data/vals_ai/google/gemini-3.5-flash/17b16189-a065-4f20-b90c-9eeab8cb3bb8.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/google_gemini-3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.5-flash", + "id": "google/gemini-3.5-flash", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.5-flash", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 42.857, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "42.857", + "raw_stderr": "6.216", + "latency": "47.196", + "cost_per_test": "0.057667", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.216, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:automata:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 38.492, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "38.492", + "raw_stderr": "6.367", + "latency": "27.741", + "cost_per_test": "0.051282", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.367, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 68.305, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "68.305", + "raw_stderr": "5.215", + "latency": "22.84", + "cost_per_test": "0.049593", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.215, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:overall:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 49.885, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "49.885", + "raw_stderr": "3.438", + "latency": "32.592", + "cost_per_test": "0.052847", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.438, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.5-flash/22f9fcff-fca3-44fa-a8c6-5c778a70cfcf.json b/data/vals_ai/google/gemini-3.5-flash/22f9fcff-fca3-44fa-a8c6-5c778a70cfcf.json new file mode 100644 index 0000000000000000000000000000000000000000..477a2f4354ad90c0fd8b216869557307a767805a --- /dev/null +++ b/data/vals_ai/google/gemini-3.5-flash/22f9fcff-fca3-44fa-a8c6-5c778a70cfcf.json @@ -0,0 +1,450 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/google_gemini-3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.5-flash", + "id": "google/gemini-3.5-flash", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.5-flash", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.939, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.939", + "raw_stderr": "1.055", + "latency": "27.306", + "cost_per_test": "0.00642", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.055, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 88.675, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.675", + "raw_stderr": "0.289", + "latency": "30.374", + "cost_per_test": "0.006199", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.289, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 76.014, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "76.014", + "raw_stderr": "3.493", + "latency": "50.772", + "cost_per_test": "0.007857", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.493, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.602, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.602", + "raw_stderr": "0.849", + "latency": "34.576", + "cost_per_test": "0.007553", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.849, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.384, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.384", + "raw_stderr": "1.281", + "latency": "29.001", + "cost_per_test": "0.006559", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.281, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.996, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.996", + "raw_stderr": "1.731", + "latency": "35.426", + "cost_per_test": "0.01073", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.731, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.5-flash/2513a14f-de89-44db-a56e-af3121dff4a5.json b/data/vals_ai/google/gemini-3.5-flash/2513a14f-de89-44db-a56e-af3121dff4a5.json new file mode 100644 index 0000000000000000000000000000000000000000..4b054dba75c494de8d77eb031568a47bb5d01e92 --- /dev/null +++ b/data/vals_ai/google/gemini-3.5-flash/2513a14f-de89-44db-a56e-af3121dff4a5.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/google_gemini-3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.5-flash", + "id": "google/gemini-3.5-flash", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.5-flash", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 77.504, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "77.504", + "raw_stderr": "1.177", + "latency": "12.063", + "cost_per_test": "0.019197", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.177, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 68.124, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.124", + "raw_stderr": "0.91", + "latency": "11.428", + "cost_per_test": "0.017873", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.91, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 58.744, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.744", + "raw_stderr": "1.388", + "latency": "10.793", + "cost_per_test": "0.01655", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.388, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.5-flash/29e7c067-ac66-4158-850b-d00154d7149a.json b/data/vals_ai/google/gemini-3.5-flash/29e7c067-ac66-4158-850b-d00154d7149a.json new file mode 100644 index 0000000000000000000000000000000000000000..cf636993fb75c5d31a5c2f3353646a82251401ea --- /dev/null +++ b/data/vals_ai/google/gemini-3.5-flash/29e7c067-ac66-4158-850b-d00154d7149a.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/google_gemini-3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.5-flash", + "id": "google/gemini-3.5-flash", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.5-flash", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 100.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "100", + "raw_stderr": "0", + "latency": "8304.209", + "cost_per_test": "19.316978", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 60.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "60", + "raw_stderr": "1.925", + "latency": "1107.228", + "cost_per_test": "2.575597", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.925, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 80.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "80", + "raw_stderr": "1.05", + "latency": "603.942", + "cost_per_test": "1.404871", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.05, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 74.157, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "74.157", + "raw_stderr": "1.124", + "latency": "373.223", + "cost_per_test": "0.868179", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.124, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.5-flash/2fe7ff49-6fb5-47ec-8d0e-d40a37b45e77.json b/data/vals_ai/google/gemini-3.5-flash/2fe7ff49-6fb5-47ec-8d0e-d40a37b45e77.json new file mode 100644 index 0000000000000000000000000000000000000000..cc40f119eacd66e2342efe2cd8f7f926b9f3e276 --- /dev/null +++ b/data/vals_ai/google/gemini-3.5-flash/2fe7ff49-6fb5-47ec-8d0e-d40a37b45e77.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/google_gemini-3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.5-flash", + "id": "google/gemini-3.5-flash", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.5-flash", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 91.414, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.414", + "raw_stderr": "1.991", + "latency": "26.753", + "cost_per_test": "0.066225", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.991, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 93.939, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.939", + "raw_stderr": "2.142", + "latency": "28.557", + "cost_per_test": "0.061459", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.142, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 92.676, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.676", + "raw_stderr": "1.462", + "latency": "27.655", + "cost_per_test": "0.063842", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.462, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.5-flash/3ef83307-e382-45e0-be8c-206dd067c951.json b/data/vals_ai/google/gemini-3.5-flash/3ef83307-e382-45e0-be8c-206dd067c951.json new file mode 100644 index 0000000000000000000000000000000000000000..8705ea35adb6d9a6daf1d71740436ba8ad7bd4bc --- /dev/null +++ b/data/vals_ai/google/gemini-3.5-flash/3ef83307-e382-45e0-be8c-206dd067c951.json @@ -0,0 +1,550 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_multimodal_index/google_gemini-3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Multimodal Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.5-flash", + "id": "google/gemini-3.5-flash", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.5-flash", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_multimodal_index:corp_fin_v2:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.vals_multimodal_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Multimodal Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (CorpFin v2).", + "metric_id": "vals_ai.vals_multimodal_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 64.569, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "64.569", + "raw_stderr": "1.633", + "latency": "8.593", + "cost_per_test": "0.140208", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.633, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:finance_agent:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.vals_multimodal_index.finance_agent", + "source_data": { + "dataset_name": "Vals Multimodal Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_multimodal_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 58.101, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "58.101", + "raw_stderr": "0.874", + "latency": "316.567", + "cost_per_test": "2.505488", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.874, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:mortgage_tax:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.vals_multimodal_index.mortgage_tax", + "source_data": { + "dataset_name": "Vals Multimodal Index - Mortgage Tax", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "mortgage_tax", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Mortgage Tax).", + "metric_id": "vals_ai.vals_multimodal_index.mortgage_tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 68.124, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "mortgage_tax", + "task_name": "Mortgage Tax", + "dataset_type": "private", + "industry": "index", + "raw_score": "68.124", + "raw_stderr": "0.91", + "latency": "11.428", + "cost_per_test": "0.017873", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.91, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:overall:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.vals_multimodal_index.overall", + "source_data": { + "dataset_name": "Vals Multimodal Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Overall).", + "metric_id": "vals_ai.vals_multimodal_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 62.928, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "62.928", + "raw_stderr": "1.464", + "latency": "266.552", + "cost_per_test": "1.011417", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.464, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:sage:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.vals_multimodal_index.sage", + "source_data": { + "dataset_name": "Vals Multimodal Index - SAGE", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "sage", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SAGE).", + "metric_id": "vals_ai.vals_multimodal_index.sage.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 49.885, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "sage", + "task_name": "SAGE", + "dataset_type": "private", + "industry": "index", + "raw_score": "49.885", + "raw_stderr": "3.438", + "latency": "32.592", + "cost_per_test": "0.052555", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.438, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:swebench:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.vals_multimodal_index.swebench", + "source_data": { + "dataset_name": "Vals Multimodal Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SWE-bench).", + "metric_id": "vals_ai.vals_multimodal_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 75.49, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "75.49", + "raw_stderr": "4.259", + "latency": "230.39", + "cost_per_test": "0.953882" + }, + "uncertainty": { + "standard_error": { + "value": 4.259, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:terminal_bench_2_1:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.vals_multimodal_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Multimodal Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_multimodal_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 74.157, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "74.157", + "raw_stderr": "1.124", + "latency": "373.223", + "cost_per_test": "0.868179", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.124, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:vibe_code_bench:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.vals_multimodal_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Multimodal Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_multimodal_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 54.709, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "54.709", + "raw_stderr": "7.002", + "latency": "893.074", + "cost_per_test": "2.540385", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 7.002, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.5-flash/3f2ec1ef-afb7-49f8-88a9-61f49301771d.json b/data/vals_ai/google/gemini-3.5-flash/3f2ec1ef-afb7-49f8-88a9-61f49301771d.json new file mode 100644 index 0000000000000000000000000000000000000000..be495e274058e9896bd99f635e411c33fee16556 --- /dev/null +++ b/data/vals_ai/google/gemini-3.5-flash/3f2ec1ef-afb7-49f8-88a9-61f49301771d.json @@ -0,0 +1,381 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/google_gemini-3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.5-flash", + "id": "google/gemini-3.5-flash", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.5-flash", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 52.381, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "52.381", + "raw_stderr": "7.8", + "latency": "313.923", + "cost_per_test": "1.795431", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 7.8, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 77.778, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "77.778", + "raw_stderr": "2.578", + "latency": "276.821", + "cost_per_test": "1.342939", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.578, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 86.598, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "86.598", + "raw_stderr": "2.452", + "latency": "210.851", + "cost_per_test": "1.058229", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.452, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "227.077", + "cost_per_test": "1.498598", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:overall:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 78.8, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "78.8", + "raw_stderr": "1.83", + "latency": "254.129", + "cost_per_test": "0.953882", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.83, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.5-flash/58943a08-575b-4ba5-a2ac-e8457f2b8548.json b/data/vals_ai/google/gemini-3.5-flash/58943a08-575b-4ba5-a2ac-e8457f2b8548.json new file mode 100644 index 0000000000000000000000000000000000000000..a4d34f733214b5e772cf4261c512af5f6d7c4a3d --- /dev/null +++ b/data/vals_ai/google/gemini-3.5-flash/58943a08-575b-4ba5-a2ac-e8457f2b8548.json @@ -0,0 +1,3486 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/hlab/google_gemini-3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Harvey's Legal Agent Benchmark", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.5-flash", + "id": "google/gemini-3.5-flash", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.5-flash", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "hlab:antitrust_competition:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "444.69", + "cost_per_test": "2.159387", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:arbitration_international_dispute_resolution:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "573.85", + "cost_per_test": "2.445749", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:banking_finance:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "440.784", + "cost_per_test": "1.899233", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:bankruptcy_restructuring:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "494.656", + "cost_per_test": "2.21075", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:capital_markets:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "413.785", + "cost_per_test": "1.728373", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:corporate_governance:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "518.079", + "cost_per_test": "2.188345", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:corporate_ma:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "421.045", + "cost_per_test": "1.877373", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Criteria Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Criteria Pass Rate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 79.312, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate", + "task_name": "Criteria Pass Rate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.312", + "raw_stderr": "0.489", + "latency": "477.704", + "cost_per_test": "2.11387", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.489, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__antitrust_competition:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 84.42, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.42", + "raw_stderr": "2.187", + "latency": "444.69", + "cost_per_test": "2.159387", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.187, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__arbitration_international_dispute_resolution:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 78.495, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.495", + "raw_stderr": "2.464", + "latency": "573.85", + "cost_per_test": "2.445749", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.464, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__banking_finance:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 80.669, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.669", + "raw_stderr": "2.412", + "latency": "440.784", + "cost_per_test": "1.899233", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.412, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__bankruptcy_restructuring:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 75.778, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "75.778", + "raw_stderr": "2.525", + "latency": "494.656", + "cost_per_test": "2.21075", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.525, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__capital_markets:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 82.745, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.745", + "raw_stderr": "2.371", + "latency": "413.785", + "cost_per_test": "1.728373", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.371, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_governance:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 85.274, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.274", + "raw_stderr": "2.077", + "latency": "518.079", + "cost_per_test": "2.188345", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.077, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_ma:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 86.538, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.538", + "raw_stderr": "1.935", + "latency": "421.045", + "cost_per_test": "1.877373", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.935, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__data_privacy_cybersecurity:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 84.452, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.452", + "raw_stderr": "2.158", + "latency": "460.391", + "cost_per_test": "1.788615", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.158, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__emerging_companies_venture_capital:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 78.498, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.498", + "raw_stderr": "2.404", + "latency": "481.855", + "cost_per_test": "2.122257", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.404, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__employment_labor:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 70.412, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "70.412", + "raw_stderr": "2.799", + "latency": "510.958", + "cost_per_test": "2.207924", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.799, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__energy_natural_resources:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 85.148, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.148", + "raw_stderr": "2.046", + "latency": "458.45", + "cost_per_test": "2.006452", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.046, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__environmental_esg:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 72.76, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "72.76", + "raw_stderr": "2.67", + "latency": "470.749", + "cost_per_test": "2.036146", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.67, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__funds_asset_management:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 76.968, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "76.968", + "raw_stderr": "2.277", + "latency": "522.945", + "cost_per_test": "2.30202", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.277, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__healthcare_life_sciences:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 80.528, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.528", + "raw_stderr": "2.279", + "latency": "575.662", + "cost_per_test": "2.41183", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.279, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__immigration:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 76.727, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "76.727", + "raw_stderr": "2.553", + "latency": "419.572", + "cost_per_test": "1.560451", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.553, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__insurance:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 63.736, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "63.736", + "raw_stderr": "2.915", + "latency": "401.166", + "cost_per_test": "1.769687", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.915, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__intellectual_property:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 90.558, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.558", + "raw_stderr": "1.92", + "latency": "399.364", + "cost_per_test": "1.661407", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.92, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__international_trade_sanctions:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 78.571, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.571", + "raw_stderr": "2.59", + "latency": "414.779", + "cost_per_test": "1.880602", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.59, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__litigation_dispute_resolution:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 80.524, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.524", + "raw_stderr": "2.428", + "latency": "522.008", + "cost_per_test": "2.378166", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.428, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__real_estate:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 86.297, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.297", + "raw_stderr": "1.859", + "latency": "501.914", + "cost_per_test": "2.425857", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.859, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__structured_finance_securitization:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 68.889, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "68.889", + "raw_stderr": "2.613", + "latency": "561.124", + "cost_per_test": "3.015746", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.613, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__tax:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 74.582, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "74.582", + "raw_stderr": "2.522", + "latency": "530.325", + "cost_per_test": "2.60193", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.522, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__trusts_estates_private_client:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 88.55, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.55", + "raw_stderr": "1.971", + "latency": "530.53", + "cost_per_test": "2.192393", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.971, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__white_collar_defense_investigations:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 73.737, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "73.737", + "raw_stderr": "2.558", + "latency": "380.554", + "cost_per_test": "1.771686", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.558, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:data_privacy_cybersecurity:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "460.391", + "cost_per_test": "1.788615", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:emerging_companies_venture_capital:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "481.855", + "cost_per_test": "2.122257", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:employment_labor:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "510.958", + "cost_per_test": "2.207924", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:energy_natural_resources:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "458.45", + "cost_per_test": "2.006452", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:environmental_esg:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "470.749", + "cost_per_test": "2.036146", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:funds_asset_management:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "522.945", + "cost_per_test": "2.30202", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:healthcare_life_sciences:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "575.662", + "cost_per_test": "2.41183", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:immigration:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "419.572", + "cost_per_test": "1.560451", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:insurance:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "401.166", + "cost_per_test": "1.769687", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:intellectual_property:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "399.364", + "cost_per_test": "1.661407", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:international_trade_sanctions:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "414.779", + "cost_per_test": "1.880602", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:litigation_dispute_resolution:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "522.008", + "cost_per_test": "2.378166", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:overall:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.overall", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Overall).", + "metric_id": "vals_ai.hlab.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 2.5, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "2.5", + "raw_stderr": "0.84", + "latency": "477.704", + "cost_per_test": "2.11387", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.84, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:real_estate:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "501.914", + "cost_per_test": "2.425857", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:structured_finance_securitization:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "561.124", + "cost_per_test": "3.015746", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:tax:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "530.325", + "cost_per_test": "2.60193", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:trusts_estates_private_client:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "530.53", + "cost_per_test": "2.192393", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:white_collar_defense_investigations:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.hlab.white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "380.554", + "cost_per_test": "1.771686", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.5-flash/5b525121-f100-4a98-9dfb-e8ff480f4582.json b/data/vals_ai/google/gemini-3.5-flash/5b525121-f100-4a98-9dfb-e8ff480f4582.json new file mode 100644 index 0000000000000000000000000000000000000000..8e1f5144f1d0fecbf4c3cbf9b5b31c3e7692888b --- /dev/null +++ b/data/vals_ai/google/gemini-3.5-flash/5b525121-f100-4a98-9dfb-e8ff480f4582.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/code-migration/google_gemini-3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Code Migration", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.5-flash", + "id": "google/gemini-3.5-flash", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.5-flash", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "code-migration:cli:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.code-migration.cli", + "source_data": { + "dataset_name": "Code Migration - CLI", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cli", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (CLI).", + "metric_id": "vals_ai.code-migration.cli.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 17.175, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cli", + "task_name": "CLI", + "dataset_type": "private", + "industry": "coding", + "raw_score": "17.175", + "raw_stderr": "1.823", + "latency": "1277.58", + "cost_per_test": "6.345007", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.823, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "code-migration:cobol:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.code-migration.cobol", + "source_data": { + "dataset_name": "Code Migration - COBOL", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cobol", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (COBOL).", + "metric_id": "vals_ai.code-migration.cobol.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 55.455, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cobol", + "task_name": "COBOL", + "dataset_type": "private", + "industry": "coding", + "raw_score": "55.455", + "raw_stderr": "15.478", + "latency": "841.077", + "cost_per_test": "3.865467", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 15.478, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "code-migration:code_quality:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.code-migration.code_quality", + "source_data": { + "dataset_name": "Code Migration - Code Quality", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "code_quality", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Code Quality).", + "metric_id": "vals_ai.code-migration.code_quality.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 2.789, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "code_quality", + "task_name": "Code Quality", + "dataset_type": "private", + "industry": "coding", + "raw_score": "2.789", + "raw_stderr": "0.073", + "latency": "1168.454", + "cost_per_test": "5.725122", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.073, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "code-migration:overall:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.code-migration.overall", + "source_data": { + "dataset_name": "Code Migration - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Overall).", + "metric_id": "vals_ai.code-migration.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 26.745, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "26.745", + "raw_stderr": "4.104", + "latency": "1168.454", + "cost_per_test": "5.725122", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 4.104, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.5-flash/73183861-484a-44a4-8f67-069039c9fe29.json b/data/vals_ai/google/gemini-3.5-flash/73183861-484a-44a4-8f67-069039c9fe29.json new file mode 100644 index 0000000000000000000000000000000000000000..7478f98f2e9f3bf9895dc432b0c215e0640d3722 --- /dev/null +++ b/data/vals_ai/google/gemini-3.5-flash/73183861-484a-44a4-8f67-069039c9fe29.json @@ -0,0 +1,762 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/google_gemini-3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.5-flash", + "id": "google/gemini-3.5-flash", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.5-flash", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 54.869, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "54.869", + "raw_stderr": "1.14", + "latency": "317.205", + "cost_per_test": "2.304186", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.14, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 45.706, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "45.706", + "raw_stderr": "0.196", + "latency": "322.279", + "cost_per_test": "2.505488", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.196, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 48.782, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "48.782", + "raw_stderr": "0.311", + "latency": "380.721", + "cost_per_test": "3.1206", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.311, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 63.948, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.948", + "raw_stderr": "2.287", + "latency": "307.82", + "cost_per_test": "2.588355", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.287, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 79.013, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "79.013", + "raw_stderr": "1.341", + "latency": "243.066", + "cost_per_test": "1.266639", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.341, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 29.372, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "29.372", + "raw_stderr": "0.936", + "latency": "424.764", + "cost_per_test": "2.97355", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.936, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 56.832, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.832", + "raw_stderr": "0.418", + "latency": "287.438", + "cost_per_test": "2.622948", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.418, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 77.597, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "77.597", + "raw_stderr": "1.028", + "latency": "304.655", + "cost_per_test": "2.413434", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.028, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 73.97, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "73.97", + "raw_stderr": "1.911", + "latency": "231.63", + "cost_per_test": "1.486992", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.911, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:overall:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 57.861, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.861", + "raw_stderr": "0.231", + "latency": "322.279", + "cost_per_test": "2.505488", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.231, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 36.365, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "36.365", + "raw_stderr": "2.688", + "latency": "447.419", + "cost_per_test": "3.899099", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.688, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.5-flash/84eeeec2-9d28-41b4-bfb7-6618df7cc95b.json b/data/vals_ai/google/gemini-3.5-flash/84eeeec2-9d28-41b4-bfb7-6618df7cc95b.json new file mode 100644 index 0000000000000000000000000000000000000000..2e5739289582344fe1b994fc488adfc8e824b68a --- /dev/null +++ b/data/vals_ai/google/gemini-3.5-flash/84eeeec2-9d28-41b4-bfb7-6618df7cc95b.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/public-benefits-bench/google_gemini-3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Public Benefits Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "social mobility", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.5-flash", + "id": "google/gemini-3.5-flash", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.5-flash", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "public-benefits-bench:overall:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.public-benefits-bench.overall", + "source_data": { + "dataset_name": "Public Benefits Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Public Benefits Bench v1.1 (Overall).", + "metric_id": "vals_ai.public-benefits-bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "score_details": { + "score": 59.472, + "details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "social mobility", + "raw_score": "59.472", + "raw_stderr": "1.277", + "latency": "484.716", + "cost_per_test": "1.113099", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.277, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.5-flash/8927794a-5afa-4698-8c0e-2f63e8644c9c.json b/data/vals_ai/google/gemini-3.5-flash/8927794a-5afa-4698-8c0e-2f63e8644c9c.json new file mode 100644 index 0000000000000000000000000000000000000000..4bb0fc302033f65a3296fc047a04b74030ca5efe --- /dev/null +++ b/data/vals_ai/google/gemini-3.5-flash/8927794a-5afa-4698-8c0e-2f63e8644c9c.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/google_gemini-3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.5-flash", + "id": "google/gemini-3.5-flash", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.5-flash", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 99.068, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "99.068", + "raw_stderr": "0.535", + "latency": "12.711", + "cost_per_test": "0.016532", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:hard:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 73.143, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "73.143", + "raw_stderr": "2.369", + "latency": "144.935", + "cost_per_test": "0.177857", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.369, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:medium:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 90.601, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "90.601", + "raw_stderr": "1.491", + "latency": "35.468", + "cost_per_test": "0.063422", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.491, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:overall:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 87.604, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "87.604", + "raw_stderr": "0.95", + "latency": "64.371", + "cost_per_test": "0.085937", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.95, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.5-flash/9b0b0a0f-13de-49ca-9b98-02e53e6eaaa8.json b/data/vals_ai/google/gemini-3.5-flash/9b0b0a0f-13de-49ca-9b98-02e53e6eaaa8.json new file mode 100644 index 0000000000000000000000000000000000000000..0f7326a03174bb14155cde588ce68c4dc708df4b --- /dev/null +++ b/data/vals_ai/google/gemini-3.5-flash/9b0b0a0f-13de-49ca-9b98-02e53e6eaaa8.json @@ -0,0 +1,1224 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_research/google_gemini-3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Legal Research Bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.5-flash", + "id": "google/gemini-3.5-flash", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.5-flash", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_research:administrative_regulatory:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.legal_research.administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 43.939, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "43.939", + "raw_stderr": "6.156", + "latency": "285.017", + "cost_per_test": "1.356228", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.156, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:business_commercial:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.legal_research.business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 27.907, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "27.907", + "raw_stderr": "6.921", + "latency": "317.652", + "cost_per_test": "1.232191", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.921, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:civil_litigation:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.legal_research.civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 22.222, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "22.222", + "raw_stderr": "8.153", + "latency": "255.111", + "cost_per_test": "1.230964", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 8.153, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:constitutional_civil_rights:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.legal_research.constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 28.947, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "28.947", + "raw_stderr": "7.456", + "latency": "224.844", + "cost_per_test": "0.987781", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 7.456, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:criminal:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.legal_research.criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 26.316, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "26.316", + "raw_stderr": "5.884", + "latency": "256.075", + "cost_per_test": "1.359179", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.884, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:family:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.legal_research.family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 18.182, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "18.182", + "raw_stderr": "8.417", + "latency": "306.75", + "cost_per_test": "1.594003", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 8.417, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:health:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.legal_research.health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 43.333, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "43.333", + "raw_stderr": "9.202", + "latency": "274.578", + "cost_per_test": "1.402227", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 9.202, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:immigration:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.legal_research.immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 18.182, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "18.182", + "raw_stderr": "12.197", + "latency": "219.971", + "cost_per_test": "0.884226", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 12.197, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:overall:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.legal_research.overall", + "source_data": { + "dataset_name": "Legal Research Bench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Overall).", + "metric_id": "vals_ai.legal_research.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 30.769, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "legal", + "raw_score": "30.769", + "raw_stderr": "3.208", + "latency": "277.753", + "cost_per_test": "1.305281", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.208, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass", + "source_data": { + "dataset_name": "Legal Research Bench - Weighted Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Weighted Pass).", + "metric_id": "vals_ai.legal_research.weighted_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 83.93, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass", + "task_name": "Weighted Pass", + "dataset_type": "private", + "industry": "legal", + "raw_score": "83.93", + "raw_stderr": "1.373", + "latency": "277.753", + "cost_per_test": "1.305281", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.373, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__administrative_regulatory:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.weighted_pass__administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 87.126, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "87.126", + "raw_stderr": "2.45", + "latency": "285.017", + "cost_per_test": "1.356228", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.45, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__business_commercial:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.weighted_pass__business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 84.143, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "84.143", + "raw_stderr": "2.921", + "latency": "317.652", + "cost_per_test": "1.232191", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.921, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__civil_litigation:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.weighted_pass__civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 86.308, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "86.308", + "raw_stderr": "2.832", + "latency": "255.111", + "cost_per_test": "1.230964", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.832, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__constitutional_civil_rights:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 85.223, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "85.223", + "raw_stderr": "2.655", + "latency": "224.844", + "cost_per_test": "0.987781", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.655, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__criminal:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.weighted_pass__criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 82.44, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "82.44", + "raw_stderr": "2.626", + "latency": "256.075", + "cost_per_test": "1.359179", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.626, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__family:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.weighted_pass__family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 74.889, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "74.889", + "raw_stderr": "4.464", + "latency": "306.75", + "cost_per_test": "1.594003", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 4.464, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__health:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.weighted_pass__health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 91.474, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "91.474", + "raw_stderr": "1.976", + "latency": "274.578", + "cost_per_test": "1.402227", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.976, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__immigration:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.weighted_pass__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 80.189, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "80.189", + "raw_stderr": "6.859", + "latency": "219.971", + "cost_per_test": "0.884226", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.859, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.5-flash/a633d9f6-0e63-47a6-a623-db6415aa62d9.json b/data/vals_ai/google/gemini-3.5-flash/a633d9f6-0e63-47a6-a623-db6415aa62d9.json new file mode 100644 index 0000000000000000000000000000000000000000..22b0282df81df8f52883d75d3b91440d6f73b3ac --- /dev/null +++ b/data/vals_ai/google/gemini-3.5-flash/a633d9f6-0e63-47a6-a623-db6415aa62d9.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/programbench/google_gemini-3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProgramBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.5-flash", + "id": "google/gemini-3.5-flash", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.5-flash", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "programbench:almost:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.programbench.almost", + "source_data": { + "dataset_name": "ProgramBench - Almost Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "almost", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Almost Resolved).", + "metric_id": "vals_ai.programbench.almost.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 1.5, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "almost", + "task_name": "Almost Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "1.5", + "raw_stderr": "0.862", + "latency": "1104.517", + "cost_per_test": "6.066019", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.862, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "programbench:overall:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.programbench.overall", + "source_data": { + "dataset_name": "ProgramBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Overall).", + "metric_id": "vals_ai.programbench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1104.517", + "cost_per_test": "6.066019", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "programbench:partial:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.programbench.partial", + "source_data": { + "dataset_name": "ProgramBench - Raw Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "partial", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Raw Pass Rate).", + "metric_id": "vals_ai.programbench.partial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 56.356, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "partial", + "task_name": "Raw Pass Rate", + "dataset_type": "public", + "industry": "coding", + "raw_score": "56.356", + "raw_stderr": "1.942", + "latency": "1104.517", + "cost_per_test": "6.066019", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.942, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "programbench:strict:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.programbench.strict", + "source_data": { + "dataset_name": "ProgramBench - Fully Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "strict", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Fully Resolved).", + "metric_id": "vals_ai.programbench.strict.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "strict", + "task_name": "Fully Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1104.517", + "cost_per_test": "6.066019", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.5-flash/b2cad686-af51-46e6-8097-db2c9430685a.json b/data/vals_ai/google/gemini-3.5-flash/b2cad686-af51-46e6-8097-db2c9430685a.json new file mode 100644 index 0000000000000000000000000000000000000000..17b7271d6ff25b91df80332b77353cba1b34d620 --- /dev/null +++ b/data/vals_ai/google/gemini-3.5-flash/b2cad686-af51-46e6-8097-db2c9430685a.json @@ -0,0 +1,418 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/google_gemini-3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.5-flash", + "id": "google/gemini-3.5-flash", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.5-flash", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 64.569, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "64.569", + "raw_stderr": "1.633", + "latency": "8.593", + "cost_per_test": "0.140208", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.633, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 58.101, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "58.101", + "raw_stderr": "0.874", + "latency": "316.567", + "cost_per_test": "2.505488", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.874, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:overall:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 62.748, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "62.748", + "raw_stderr": "1.606", + "latency": "364.369", + "cost_per_test": "1.401628", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.606, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 75.49, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "75.49", + "raw_stderr": "4.259", + "latency": "230.39", + "cost_per_test": "0.953882" + }, + "uncertainty": { + "standard_error": { + "value": 4.259, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 74.157, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "74.157", + "raw_stderr": "1.124", + "latency": "373.223", + "cost_per_test": "0.868179", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.124, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 54.709, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "54.709", + "raw_stderr": "7.002", + "latency": "893.074", + "cost_per_test": "2.540385", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 7.002, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.5-flash/b782ad01-1c0d-46c9-916e-11da299e4fc8.json b/data/vals_ai/google/gemini-3.5-flash/b782ad01-1c0d-46c9-916e-11da299e4fc8.json new file mode 100644 index 0000000000000000000000000000000000000000..0083de814e93bae02597b3599dbbe32bb9bb5c45 --- /dev/null +++ b/data/vals_ai/google/gemini-3.5-flash/b782ad01-1c0d-46c9-916e-11da299e4fc8.json @@ -0,0 +1,696 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/emb/google_gemini-3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - EMB", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.5-flash", + "id": "google/gemini-3.5-flash", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.5-flash", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "emb:comps:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.emb.comps", + "source_data": { + "dataset_name": "EMB - Comps", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "comps", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Comps).", + "metric_id": "vals_ai.emb.comps.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 69.788, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "comps", + "task_name": "Comps", + "dataset_type": "private", + "industry": "finance", + "raw_score": "69.788", + "raw_stderr": "4.035", + "latency": "1282.564", + "cost_per_test": "8.099855", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 4.035, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:dataroom:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.emb.dataroom", + "source_data": { + "dataset_name": "EMB - Dataroom", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dataroom", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Dataroom).", + "metric_id": "vals_ai.emb.dataroom.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 80.921, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dataroom", + "task_name": "Dataroom", + "dataset_type": "private", + "industry": "finance", + "raw_score": "80.921", + "raw_stderr": "5.377", + "latency": "1239.062", + "cost_per_test": "4.451663", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.377, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:dcf:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.emb.dcf", + "source_data": { + "dataset_name": "EMB - DCF", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dcf", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (DCF).", + "metric_id": "vals_ai.emb.dcf.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 58.36, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dcf", + "task_name": "DCF", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.36", + "raw_stderr": "6.889", + "latency": "975.444", + "cost_per_test": "4.056284", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.889, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:lbo:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.emb.lbo", + "source_data": { + "dataset_name": "EMB - LBO", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "lbo", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (LBO).", + "metric_id": "vals_ai.emb.lbo.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 46.47, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "lbo", + "task_name": "LBO", + "dataset_type": "private", + "industry": "finance", + "raw_score": "46.47", + "raw_stderr": "5.599", + "latency": "1383.424", + "cost_per_test": "6.810486", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.599, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:ma:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.emb.ma", + "source_data": { + "dataset_name": "EMB - M&A", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "ma", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (M&A).", + "metric_id": "vals_ai.emb.ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 68.445, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "ma", + "task_name": "M&A", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.445", + "raw_stderr": "6.087", + "latency": "1014.816", + "cost_per_test": "4.118564", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.087, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:misc:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.emb.misc", + "source_data": { + "dataset_name": "EMB - Misc", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "misc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Misc).", + "metric_id": "vals_ai.emb.misc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 62.192, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "misc", + "task_name": "Misc", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.192", + "raw_stderr": "6.979", + "latency": "1022.574", + "cost_per_test": "4.497662", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.979, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:operating:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.emb.operating", + "source_data": { + "dataset_name": "EMB - Operating", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "operating", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Operating).", + "metric_id": "vals_ai.emb.operating.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 58.658, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "operating", + "task_name": "Operating", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.658", + "raw_stderr": "7.143", + "latency": "904.09", + "cost_per_test": "3.646821", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 7.143, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:overall:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.emb.overall", + "source_data": { + "dataset_name": "EMB - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Overall).", + "metric_id": "vals_ai.emb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 63.548, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.548", + "raw_stderr": "2.752", + "latency": "1123.555", + "cost_per_test": "5.173115", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.752, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:scratch:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.emb.scratch", + "source_data": { + "dataset_name": "EMB - Scratch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "scratch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Scratch).", + "metric_id": "vals_ai.emb.scratch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 64.0, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "scratch", + "task_name": "Scratch", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64", + "raw_stderr": "2.417", + "latency": "1075.4", + "cost_per_test": "4.744339", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 2.417, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:template:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.emb.template", + "source_data": { + "dataset_name": "EMB - Template", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "template", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Template).", + "metric_id": "vals_ai.emb.template.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 62.531, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "template", + "task_name": "Template", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.531", + "raw_stderr": "4.267", + "latency": "1171.71", + "cost_per_test": "5.601891", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 4.267, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.5-flash/b82b3cd4-e11f-45de-8be5-256ed714e842.json b/data/vals_ai/google/gemini-3.5-flash/b82b3cd4-e11f-45de-8be5-256ed714e842.json new file mode 100644 index 0000000000000000000000000000000000000000..eb5b8503aec1d664cf6615257d216169d7645d60 --- /dev/null +++ b/data/vals_ai/google/gemini-3.5-flash/b82b3cd4-e11f-45de-8be5-256ed714e842.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/google_gemini-3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.5-flash", + "id": "google/gemini-3.5-flash", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.5-flash", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 61.897, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.897", + "raw_stderr": "1.389", + "latency": "19.694", + "cost_per_test": "0.038136", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.389, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 74.366, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "74.366", + "raw_stderr": "0.846", + "latency": "19.694", + "cost_per_test": "0.038136", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.846, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 86.836, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "86.836", + "raw_stderr": "0.967", + "latency": "19.694", + "cost_per_test": "0.038136", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.967, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.5-flash/ba574f29-b650-4cf7-857f-3f6fd2390d5a.json b/data/vals_ai/google/gemini-3.5-flash/ba574f29-b650-4cf7-857f-3f6fd2390d5a.json new file mode 100644 index 0000000000000000000000000000000000000000..3a344ba97b953ead1446ffd5c984455d967a8db4 --- /dev/null +++ b/data/vals_ai/google/gemini-3.5-flash/ba574f29-b650-4cf7-857f-3f6fd2390d5a.json @@ -0,0 +1,195 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/cyber/google_gemini-3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CyberBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "beta", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.5-flash", + "id": "google/gemini-3.5-flash", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.5-flash", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "cyber:overall:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.cyber.overall", + "source_data": { + "dataset_name": "CyberBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (Overall).", + "metric_id": "vals_ai.cyber.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 69.331, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "beta", + "raw_score": "69.331", + "raw_stderr": "5.79", + "latency": "961.936", + "cost_per_test": "1.9988", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.79, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "cyber:patch:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.cyber.patch", + "source_data": { + "dataset_name": "CyberBench - Patch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "patch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (Patch).", + "metric_id": "vals_ai.cyber.patch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 81.034, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "patch", + "task_name": "Patch", + "dataset_type": "private", + "industry": "beta", + "raw_score": "81.034", + "raw_stderr": "5.148", + "latency": "868.639", + "cost_per_test": "1.003571", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.148, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "cyber:poc:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.cyber.poc", + "source_data": { + "dataset_name": "CyberBench - PoC", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "poc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (PoC).", + "metric_id": "vals_ai.cyber.poc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 57.627, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "poc", + "task_name": "PoC", + "dataset_type": "private", + "industry": "beta", + "raw_score": "57.627", + "raw_stderr": "6.433", + "latency": "1055.232", + "cost_per_test": "2.994029", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.433, + "method": "vals_reported" + } + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.5-flash/c315f395-bbdb-437e-83b3-e51ee5986b7c.json b/data/vals_ai/google/gemini-3.5-flash/c315f395-bbdb-437e-83b3-e51ee5986b7c.json new file mode 100644 index 0000000000000000000000000000000000000000..c7aef882e5ff193f57d6c949562d368dc7aaf9c6 --- /dev/null +++ b/data/vals_ai/google/gemini-3.5-flash/c315f395-bbdb-437e-83b3-e51ee5986b7c.json @@ -0,0 +1,1071 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/google_gemini-3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.5-flash", + "id": "google/gemini-3.5-flash", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.5-flash", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.863, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.863", + "raw_stderr": "0.896", + "latency": "12.325", + "cost_per_test": "0.015872", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.896, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.283, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.283", + "raw_stderr": "0.891", + "latency": "17.416", + "cost_per_test": "0.02368", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.891, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.11, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.11", + "raw_stderr": "0.77", + "latency": "55.325", + "cost_per_test": "0.035628", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.77, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.707, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.707", + "raw_stderr": "1.362", + "latency": "10.896", + "cost_per_test": "0.022972", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.362, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.943, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.943", + "raw_stderr": "0.937", + "latency": "8.784", + "cost_per_test": "0.017242", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.937, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.822, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.822", + "raw_stderr": "1.051", + "latency": "30.126", + "cost_per_test": "0.061489", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.051, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.619, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.619", + "raw_stderr": "1.294", + "latency": "10.563", + "cost_per_test": "0.019367", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.294, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.627, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.627", + "raw_stderr": "1.984", + "latency": "11.228", + "cost_per_test": "0.021196", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.984, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.381, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.381", + "raw_stderr": "1.199", + "latency": "45.82", + "cost_per_test": "0.028158", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.199, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 96.299, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "96.299", + "raw_stderr": "0.533", + "latency": "50.407", + "cost_per_test": "0.023637", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.533, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.554, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.554", + "raw_stderr": "1.118", + "latency": "37.383", + "cost_per_test": "0.016173", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.118, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.515, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.515", + "raw_stderr": "0.306", + "latency": "26.262", + "cost_per_test": "0.02498", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.306, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.575, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.575", + "raw_stderr": "1.477", + "latency": "10.617", + "cost_per_test": "0.01891", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.477, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.072, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.072", + "raw_stderr": "0.663", + "latency": "58.125", + "cost_per_test": "0.029412", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.663, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.351, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.351", + "raw_stderr": "1.045", + "latency": "8.646", + "cost_per_test": "0.015991", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.045, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.5-flash/c86c444a-d396-4a6d-a04f-8b66696fdaa4.json b/data/vals_ai/google/gemini-3.5-flash/c86c444a-d396-4a6d-a04f-8b66696fdaa4.json new file mode 100644 index 0000000000000000000000000000000000000000..3c92ad7525da59145ca39385dcfc848add565c9c --- /dev/null +++ b/data/vals_ai/google/gemini-3.5-flash/c86c444a-d396-4a6d-a04f-8b66696fdaa4.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/google_gemini-3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.5-flash", + "id": "google/gemini-3.5-flash", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.5-flash", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 29.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "29", + "raw_stderr": "4.56", + "latency": "413.679", + "cost_per_test": "1.399975", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 4.56, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.5-flash/d5e8bba6-4a07-4f31-b5ef-c5f2422d3607.json b/data/vals_ai/google/gemini-3.5-flash/d5e8bba6-4a07-4f31-b5ef-c5f2422d3607.json new file mode 100644 index 0000000000000000000000000000000000000000..e03884dc45c67892302f413e497be93516a2ec2f --- /dev/null +++ b/data/vals_ai/google/gemini-3.5-flash/d5e8bba6-4a07-4f31-b5ef-c5f2422d3607.json @@ -0,0 +1,105 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/google_gemini-3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.5-flash", + "id": "google/gemini-3.5-flash", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.5-flash", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 88.266, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.266", + "raw_stderr": "0.774", + "latency": "12.226", + "cost_per_test": "0.025155", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.774, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemini-3.5-flash/fb1a3297-a265-46fa-bc03-e109a27979e3.json b/data/vals_ai/google/gemini-3.5-flash/fb1a3297-a265-46fa-bc03-e109a27979e3.json new file mode 100644 index 0000000000000000000000000000000000000000..5b4a9b27fe6cebf6368ef119efc5e698fcebe0d6 --- /dev/null +++ b/data/vals_ai/google/gemini-3.5-flash/fb1a3297-a265-46fa-bc03-e109a27979e3.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/google_gemini-3.5-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemini-3.5-flash", + "id": "google/gemini-3.5-flash", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemini-3.5-flash", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 62.821, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.821", + "raw_stderr": "1.65", + "latency": "5.149", + "cost_per_test": "0.011121", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.65, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.667", + "raw_stderr": "1.609", + "latency": "12.357", + "cost_per_test": "0.29559", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.609, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 64.686, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.686", + "raw_stderr": "0.942", + "latency": "8.7", + "cost_per_test": "0.140208", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.942, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:google/gemini-3.5-flash:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 64.569, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.569", + "raw_stderr": "1.633", + "latency": "8.593", + "cost_per_test": "0.113913", + "temperature": "1", + "max_output_tokens": "65536", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.633, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 65536, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemma-2-27b-it/cc8f2af6-5804-44a9-a3bc-e9f6d2e67a62.json b/data/vals_ai/google/gemma-2-27b-it/cc8f2af6-5804-44a9-a3bc-e9f6d2e67a62.json new file mode 100644 index 0000000000000000000000000000000000000000..58f641c57c11e0216e172b5394614faf5f424311 --- /dev/null +++ b/data/vals_ai/google/gemma-2-27b-it/cc8f2af6-5804-44a9-a3bc-e9f6d2e67a62.json @@ -0,0 +1,408 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/google_gemma-2-27b-it/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemma-2-27b-it", + "id": "google/gemma-2-27b-it", + "developer": "google", + "additional_details": { + "vals_model_id": "together/google/gemma-2-27b-it", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:together/google/gemma-2-27b-it:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.082, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.082", + "raw_stderr": "1.079", + "latency": "0.576", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.079, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:together/google/gemma-2-27b-it:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 65.128, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "65.128", + "raw_stderr": "0.229", + "latency": "0.544", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.229, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:together/google/gemma-2-27b-it:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 54.941, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "54.941", + "raw_stderr": "0.646", + "latency": "0.505", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.646, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:together/google/gemma-2-27b-it:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 67.978, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "67.978", + "raw_stderr": "0.651", + "latency": "0.887", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.651, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:together/google/gemma-2-27b-it:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 71.531, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "71.531", + "raw_stderr": "1.451", + "latency": "0.587", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.451, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:together/google/gemma-2-27b-it:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 69.21, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "69.21", + "raw_stderr": "2.619", + "latency": "2.222", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.619, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemma-2-9b-it/0fcfbecf-346b-4bbc-ae4e-9990fb33cd99.json b/data/vals_ai/google/gemma-2-9b-it/0fcfbecf-346b-4bbc-ae4e-9990fb33cd99.json new file mode 100644 index 0000000000000000000000000000000000000000..00161fbd6fc1a4aeef64050575c25a255cd8b8a6 --- /dev/null +++ b/data/vals_ai/google/gemma-2-9b-it/0fcfbecf-346b-4bbc-ae4e-9990fb33cd99.json @@ -0,0 +1,408 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/google_gemma-2-9b-it/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemma-2-9b-it", + "id": "google/gemma-2-9b-it", + "developer": "google", + "additional_details": { + "vals_model_id": "together/google/gemma-2-9b-it", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:together/google/gemma-2-9b-it:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 71.712, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "71.712", + "raw_stderr": "1.082", + "latency": "0.389", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.082, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:together/google/gemma-2-9b-it:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 71.075, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "71.075", + "raw_stderr": "0.223", + "latency": "0.4", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.223, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:together/google/gemma-2-9b-it:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 62.036, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "62.036", + "raw_stderr": "0.779", + "latency": "0.461", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.779, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:together/google/gemma-2-9b-it:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 68.138, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "68.138", + "raw_stderr": "0.735", + "latency": "0.579", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.735, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:together/google/gemma-2-9b-it:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 72.476, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "72.476", + "raw_stderr": "1.071", + "latency": "0.388", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.071, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:together/google/gemma-2-9b-it:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 63.391, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "63.391", + "raw_stderr": "3.248", + "latency": "1.259", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.248, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemma-4-31b-it/5056bc91-f5c3-4b7d-acb9-3168fd0d5339.json b/data/vals_ai/google/gemma-4-31b-it/5056bc91-f5c3-4b7d-acb9-3168fd0d5339.json new file mode 100644 index 0000000000000000000000000000000000000000..f429738ea5d1064e576d4bb5273ffff275062756 --- /dev/null +++ b/data/vals_ai/google/gemma-4-31b-it/5056bc91-f5c3-4b7d-acb9-3168fd0d5339.json @@ -0,0 +1,366 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/google_gemma-4-31b-it/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemma-4-31b-it", + "id": "google/gemma-4-31b-it", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemma-4-31b-it", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:google/gemma-4-31b-it:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 50.794, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "50.794", + "raw_stderr": "6.273", + "latency": "106.921", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.273, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:automata:google/gemma-4-31b-it:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 41.666, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "41.666", + "raw_stderr": "6.317", + "latency": "77.76", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.317, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:calculus:google/gemma-4-31b-it:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 59.259, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "59.259", + "raw_stderr": "6.315", + "latency": "88.767", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 6.315, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:google/gemma-4-31b-it:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 72.643, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "72.643", + "raw_stderr": "5.046", + "latency": "94.494", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 5.046, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:overall:google/gemma-4-31b-it:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 55.034, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "55.034", + "raw_stderr": "3.411", + "latency": "93.058", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 3.411, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/google/gemma-4-31b-it/be3c6f96-fe47-45d2-9939-ea4daaf85df9.json b/data/vals_ai/google/gemma-4-31b-it/be3c6f96-fe47-45d2-9939-ea4daaf85df9.json new file mode 100644 index 0000000000000000000000000000000000000000..36b17a43a8ce811ad4687b30ba8d38f8d6e5d681 --- /dev/null +++ b/data/vals_ai/google/gemma-4-31b-it/be3c6f96-fe47-45d2-9939-ea4daaf85df9.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/google_gemma-4-31b-it/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gemma-4-31b-it", + "id": "google/gemma-4-31b-it", + "developer": "google", + "additional_details": { + "vals_model_id": "google/gemma-4-31b-it", + "vals_provider": "Google" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:google/gemma-4-31b-it:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 65.66, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.66", + "raw_stderr": "1.339", + "latency": "55.497", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.339, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:google/gemma-4-31b-it:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 61.368, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.368", + "raw_stderr": "0.967", + "latency": "41.928", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 0.967, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:google/gemma-4-31b-it:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 57.075, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.075", + "raw_stderr": "1.396", + "latency": "28.359", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "Google" + }, + "uncertainty": { + "standard_error": { + "value": 1.396, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/langston/nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561-thinking/11622b35-095e-4287-9713-6d215cd172aa.json b/data/vals_ai/langston/nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561-thinking/11622b35-095e-4287-9713-6d215cd172aa.json new file mode 100644 index 0000000000000000000000000000000000000000..b11f3e68901e0033a0675f2ac92002c70525c4dd --- /dev/null +++ b/data/vals_ai/langston/nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561-thinking/11622b35-095e-4287-9713-6d215cd172aa.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/langston_nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking", + "id": "langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking", + "developer": "langston", + "additional_details": { + "vals_model_id": "together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 64.646, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "64.646", + "raw_stderr": "3.419", + "latency": "66.301", + "cost_per_test": "0", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.419, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 60.101, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "60.101", + "raw_stderr": "3.512", + "latency": "88.877", + "cost_per_test": "0", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.512, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 62.374, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "62.374", + "raw_stderr": "2.451", + "latency": "77.589", + "cost_per_test": "0", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.451, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/langston/nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561-thinking/5d641ee5-7262-4cc3-9c18-746a56cdc1f7.json b/data/vals_ai/langston/nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561-thinking/5d641ee5-7262-4cc3-9c18-746a56cdc1f7.json new file mode 100644 index 0000000000000000000000000000000000000000..04a6bf74c1fc0adbc944118757ad5d3fe2c35dd9 --- /dev/null +++ b/data/vals_ai/langston/nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561-thinking/5d641ee5-7262-4cc3-9c18-746a56cdc1f7.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/langston_nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking", + "id": "langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking", + "developer": "langston", + "additional_details": { + "vals_model_id": "together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 52.085, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.085", + "raw_stderr": "1.428", + "latency": "32.67", + "cost_per_test": "0", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.428, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 63.736, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.736", + "raw_stderr": "0.943", + "latency": "32.67", + "cost_per_test": "0", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.943, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 75.388, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "75.388", + "raw_stderr": "1.232", + "latency": "32.67", + "cost_per_test": "0", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.232, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/langston/nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561-thinking/99f5ae2b-c555-4dd6-8520-5b255c7cd1b5.json b/data/vals_ai/langston/nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561-thinking/99f5ae2b-c555-4dd6-8520-5b255c7cd1b5.json new file mode 100644 index 0000000000000000000000000000000000000000..d945a87c1864ec715fc80bc1fa832a2ae249709b --- /dev/null +++ b/data/vals_ai/langston/nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561-thinking/99f5ae2b-c555-4dd6-8520-5b255c7cd1b5.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/langston_nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking", + "id": "langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking", + "developer": "langston", + "additional_details": { + "vals_model_id": "together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 95.031, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "95.031", + "raw_stderr": "1.211", + "latency": "72.746", + "cost_per_test": "0", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.211, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 20.286, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "20.286", + "raw_stderr": "2.149", + "latency": "232.563", + "cost_per_test": "0", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.149, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 59.791, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "59.791", + "raw_stderr": "2.505", + "latency": "151.89", + "cost_per_test": "0", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.505, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 58.369, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "58.369", + "raw_stderr": "1.172", + "latency": "152.4", + "cost_per_test": "0", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.172, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/langston/nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561-thinking/fbc8c580-b3ca-42ea-bdc7-1643387de057.json b/data/vals_ai/langston/nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561-thinking/fbc8c580-b3ca-42ea-bdc7-1643387de057.json new file mode 100644 index 0000000000000000000000000000000000000000..3a59f21ec388b7f659eefa873e4eb20d1699a5d4 --- /dev/null +++ b/data/vals_ai/langston/nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561-thinking/fbc8c580-b3ca-42ea-bdc7-1643387de057.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/langston_nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking", + "id": "langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking", + "developer": "langston", + "additional_details": { + "vals_model_id": "together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.053, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.053", + "raw_stderr": "1.326", + "latency": "36.516", + "cost_per_test": "0", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.326, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.665, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.665", + "raw_stderr": "1.546", + "latency": "29.982", + "cost_per_test": "0", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.546, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 70.848, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.848", + "raw_stderr": "1.367", + "latency": "52.813", + "cost_per_test": "0", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.367, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.146, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.146", + "raw_stderr": "2.162", + "latency": "34.963", + "cost_per_test": "0", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.162, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.924, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.924", + "raw_stderr": "1.371", + "latency": "34.704", + "cost_per_test": "0", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.371, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 53.87, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "53.87", + "raw_stderr": "1.603", + "latency": "83.21", + "cost_per_test": "0", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.603, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.839, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.839", + "raw_stderr": "1.546", + "latency": "20.723", + "cost_per_test": "0", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.546, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 64.829, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "64.829", + "raw_stderr": "2.446", + "latency": "19.588", + "cost_per_test": "0", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.446, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 55.223, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "55.223", + "raw_stderr": "1.499", + "latency": "24.625", + "cost_per_test": "0", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.499, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 56.477, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "56.477", + "raw_stderr": "1.349", + "latency": "44.146", + "cost_per_test": "0", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.349, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.701, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.701", + "raw_stderr": "1.454", + "latency": "30.305", + "cost_per_test": "0", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.454, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 69.577, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.577", + "raw_stderr": "0.441", + "latency": "36.351", + "cost_per_test": "0", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.441, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 66.533, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.533", + "raw_stderr": "2.116", + "latency": "26.197", + "cost_per_test": "0", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.116, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 64.896, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "64.896", + "raw_stderr": "1.331", + "latency": "50.202", + "cost_per_test": "0", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.331, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.068, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.068", + "raw_stderr": "1.494", + "latency": "20.936", + "cost_per_test": "0", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.494, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/langston/nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561/0dd08b73-86a8-4145-bea5-71623542a34e.json b/data/vals_ai/langston/nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561/0dd08b73-86a8-4145-bea5-71623542a34e.json new file mode 100644 index 0000000000000000000000000000000000000000..e5d2dfe27973c1d70552a90a903d14379725d14c --- /dev/null +++ b/data/vals_ai/langston/nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561/0dd08b73-86a8-4145-bea5-71623542a34e.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/langston_nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561", + "id": "langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561", + "developer": "langston", + "additional_details": { + "vals_model_id": "together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.753, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.753", + "raw_stderr": "1.628", + "latency": "15.727", + "cost_per_test": "0", + "temperature": "0", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.628, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.636, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.636", + "raw_stderr": "1.397", + "latency": "21.458", + "cost_per_test": "0", + "temperature": "0", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.397, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 65.724, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "65.724", + "raw_stderr": "1.423", + "latency": "19.731", + "cost_per_test": "0", + "temperature": "0", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.423, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.341, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.341", + "raw_stderr": "2.162", + "latency": "16.809", + "cost_per_test": "0", + "temperature": "0", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.162, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.858, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.858", + "raw_stderr": "1.531", + "latency": "15.644", + "cost_per_test": "0", + "temperature": "0", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.531, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 52.219, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "52.219", + "raw_stderr": "1.606", + "latency": "23.061", + "cost_per_test": "0", + "temperature": "0", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.606, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 67.848, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "67.848", + "raw_stderr": "1.709", + "latency": "11.844", + "cost_per_test": "0", + "temperature": "0", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.709, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.115, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.115", + "raw_stderr": "2.216", + "latency": "21.861", + "cost_per_test": "0", + "temperature": "0", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.216, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 64.124, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "64.124", + "raw_stderr": "1.446", + "latency": "17.7", + "cost_per_test": "0", + "temperature": "0", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.446, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.87, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.87", + "raw_stderr": "1.189", + "latency": "15.477", + "cost_per_test": "0", + "temperature": "0", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.189, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 66.775, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.775", + "raw_stderr": "1.619", + "latency": "12.585", + "cost_per_test": "0", + "temperature": "0", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.619, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 70.778, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.778", + "raw_stderr": "0.45", + "latency": "17.45", + "cost_per_test": "0", + "temperature": "0", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.45, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 61.723, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "61.723", + "raw_stderr": "2.21", + "latency": "15.884", + "cost_per_test": "0", + "temperature": "0", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.21, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 66.975, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.975", + "raw_stderr": "1.328", + "latency": "18.493", + "cost_per_test": "0", + "temperature": "0", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.328, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.935, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.935", + "raw_stderr": "1.681", + "latency": "18.032", + "cost_per_test": "0", + "temperature": "0", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.681, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/langston/nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561/672ada0b-12e7-4394-8004-fb6617df63c8.json b/data/vals_ai/langston/nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561/672ada0b-12e7-4394-8004-fb6617df63c8.json new file mode 100644 index 0000000000000000000000000000000000000000..d5a66a5b0473a829c4c9079d3d05e8a9cae89c80 --- /dev/null +++ b/data/vals_ai/langston/nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561/672ada0b-12e7-4394-8004-fb6617df63c8.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/langston_nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561", + "id": "langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561", + "developer": "langston", + "additional_details": { + "vals_model_id": "together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 44.444, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "44.444", + "raw_stderr": "3.456", + "latency": "16.209", + "cost_per_test": "0", + "temperature": "0", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.456, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 51.01, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "51.01", + "raw_stderr": "3.523", + "latency": "16.755", + "cost_per_test": "0", + "temperature": "0", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.523, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 47.727, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "47.727", + "raw_stderr": "2.468", + "latency": "16.482", + "cost_per_test": "0", + "temperature": "0", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.468, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/langston/nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561/a7f04368-2322-475b-aa34-cff8bec3e99a.json b/data/vals_ai/langston/nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561/a7f04368-2322-475b-aa34-cff8bec3e99a.json new file mode 100644 index 0000000000000000000000000000000000000000..fdee2c1337758fe7cb7e49c106da1899fa930acf --- /dev/null +++ b/data/vals_ai/langston/nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561/a7f04368-2322-475b-aa34-cff8bec3e99a.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/langston_nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561", + "id": "langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561", + "developer": "langston", + "additional_details": { + "vals_model_id": "together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 48.078, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "48.078", + "raw_stderr": "1.429", + "latency": "18.308", + "cost_per_test": "0", + "temperature": "0", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.429, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 60.22, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.22", + "raw_stderr": "0.959", + "latency": "18.308", + "cost_per_test": "0", + "temperature": "0", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.959, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 72.363, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "72.363", + "raw_stderr": "1.279", + "latency": "18.308", + "cost_per_test": "0", + "temperature": "0", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.279, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/langston/nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561/fa3927bf-e580-4533-a8f0-83efce5bf6a4.json b/data/vals_ai/langston/nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561/fa3927bf-e580-4533-a8f0-83efce5bf6a4.json new file mode 100644 index 0000000000000000000000000000000000000000..e4008e98b8415339e1933831f2e01c779b3cd4dc --- /dev/null +++ b/data/vals_ai/langston/nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561/fa3927bf-e580-4533-a8f0-83efce5bf6a4.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/langston_nim_nvidia_llama-3.3-nemotron-super-49b-v1-42e84561/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561", + "id": "langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561", + "developer": "langston", + "additional_details": { + "vals_model_id": "together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 83.54, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "83.54", + "raw_stderr": "2.066", + "latency": "42.114", + "cost_per_test": "0", + "temperature": "0", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.066, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 3.714, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "3.714", + "raw_stderr": "1.011", + "latency": "46.083", + "cost_per_test": "0", + "temperature": "0", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.011, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 21.671, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "21.671", + "raw_stderr": "2.105", + "latency": "44.488", + "cost_per_test": "0", + "temperature": "0", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.105, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:together/langston/nim/nvidia/llama-3.3-nemotron-super-49b-v1-42e84561:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 36.308, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "36.308", + "raw_stderr": "1.039", + "latency": "44.228", + "cost_per_test": "0", + "temperature": "0", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.039, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta-llama/Llama-2-70b-hf/cdbf5041-9d63-44d4-8662-87980ed99754.json b/data/vals_ai/meta-llama/Llama-2-70b-hf/cdbf5041-9d63-44d4-8662-87980ed99754.json new file mode 100644 index 0000000000000000000000000000000000000000..46a29bb9fafc4f939cdd93cc4244bfa109279dc1 --- /dev/null +++ b/data/vals_ai/meta-llama/Llama-2-70b-hf/cdbf5041-9d63-44d4-8662-87980ed99754.json @@ -0,0 +1,408 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/meta-llama_Llama-2-70b-hf/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Llama-2-70b-hf", + "id": "meta-llama/Llama-2-70b-hf", + "developer": "meta-llama", + "additional_details": { + "vals_model_id": "together/meta-llama/Llama-2-70b-hf", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:together/meta-llama/Llama-2-70b-hf:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 73.2, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "73.2", + "raw_stderr": "1.076", + "latency": "11.731", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.076, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:together/meta-llama/Llama-2-70b-hf:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 58.277, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "58.277", + "raw_stderr": "0.136", + "latency": "11.031", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.136, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:together/meta-llama/Llama-2-70b-hf:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 53.892, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "53.892", + "raw_stderr": "0.355", + "latency": "29.839", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.355, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:together/meta-llama/Llama-2-70b-hf:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 62.434, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "62.434", + "raw_stderr": "0.711", + "latency": "14.204", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.711, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:together/meta-llama/Llama-2-70b-hf:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 64.494, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "64.494", + "raw_stderr": "1.369", + "latency": "12.007", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.369, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:together/meta-llama/Llama-2-70b-hf:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 62.305, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "62.305", + "raw_stderr": "3.074", + "latency": "6.414", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.074, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta-llama/Llama-3.2-11B-Vision-Instruct-Turbo/a1df5eae-9efc-48a0-b8ca-ad777c10e545.json b/data/vals_ai/meta-llama/Llama-3.2-11B-Vision-Instruct-Turbo/a1df5eae-9efc-48a0-b8ca-ad777c10e545.json new file mode 100644 index 0000000000000000000000000000000000000000..b09ee78d8d7fa2024198cfeb1c0a03de4d8f178b --- /dev/null +++ b/data/vals_ai/meta-llama/Llama-3.2-11B-Vision-Instruct-Turbo/a1df5eae-9efc-48a0-b8ca-ad777c10e545.json @@ -0,0 +1,99 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/meta-llama_Llama-3.2-11B-Vision-Instruct-Turbo/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Llama-3.2-11B-Vision-Instruct-Turbo", + "id": "meta-llama/Llama-3.2-11B-Vision-Instruct-Turbo", + "developer": "meta-llama", + "additional_details": { + "vals_model_id": "together/meta-llama/Llama-3.2-11B-Vision-Instruct-Turbo", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:together/meta-llama/Llama-3.2-11B-Vision-Instruct-Turbo:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 38.821, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "38.821", + "raw_stderr": "1.171", + "latency": "38.07", + "cost_per_test": "0.000171", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.171, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta-llama/Llama-3.2-11B-Vision-Instruct-Turbo/c218cca2-9c64-4d4e-9a8e-f723932059eb.json b/data/vals_ai/meta-llama/Llama-3.2-11B-Vision-Instruct-Turbo/c218cca2-9c64-4d4e-9a8e-f723932059eb.json new file mode 100644 index 0000000000000000000000000000000000000000..92c452352346163af81f1c4e1617c5d54ed9eaee --- /dev/null +++ b/data/vals_ai/meta-llama/Llama-3.2-11B-Vision-Instruct-Turbo/c218cca2-9c64-4d4e-9a8e-f723932059eb.json @@ -0,0 +1,216 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/meta-llama_Llama-3.2-11B-Vision-Instruct-Turbo/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Llama-3.2-11B-Vision-Instruct-Turbo", + "id": "meta-llama/Llama-3.2-11B-Vision-Instruct-Turbo", + "developer": "meta-llama", + "additional_details": { + "vals_model_id": "together/meta-llama/Llama-3.2-11B-Vision-Instruct-Turbo", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:together/meta-llama/Llama-3.2-11B-Vision-Instruct-Turbo:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 21.383, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "21.383", + "raw_stderr": "2.264", + "latency": "2.76", + "cost_per_test": "9.7e-05", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.264, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:together/meta-llama/Llama-3.2-11B-Vision-Instruct-Turbo:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 23.291, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "23.291", + "raw_stderr": "1.649", + "latency": "2.452", + "cost_per_test": "9.2e-05", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.649, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:together/meta-llama/Llama-3.2-11B-Vision-Instruct-Turbo:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 25.199, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "25.199", + "raw_stderr": "2.397", + "latency": "2.143", + "cost_per_test": "8.6e-05", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.397, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo/1f761080-4e83-4c3a-ab41-aff78f4e497e.json b/data/vals_ai/meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo/1f761080-4e83-4c3a-ab41-aff78f4e497e.json new file mode 100644 index 0000000000000000000000000000000000000000..b93c7bc789819740683a84cb55963c9b555f2e51 --- /dev/null +++ b/data/vals_ai/meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo/1f761080-4e83-4c3a-ab41-aff78f4e497e.json @@ -0,0 +1,216 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/meta-llama_Llama-3.2-90B-Vision-Instruct-Turbo/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Llama-3.2-90B-Vision-Instruct-Turbo", + "id": "meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo", + "developer": "meta-llama", + "additional_details": { + "vals_model_id": "together/meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:together/meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 39.269, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "39.269", + "raw_stderr": "2.695", + "latency": "4.318", + "cost_per_test": "0.000523", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.695, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:together/meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 38.792, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "38.792", + "raw_stderr": "1.901", + "latency": "5.068", + "cost_per_test": "0.000576", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.901, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:together/meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 38.315, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "38.315", + "raw_stderr": "2.683", + "latency": "5.819", + "cost_per_test": "0.00063", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.683, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo/599474d6-c3b6-4a36-9723-5c7c41dbd363.json b/data/vals_ai/meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo/599474d6-c3b6-4a36-9723-5c7c41dbd363.json new file mode 100644 index 0000000000000000000000000000000000000000..10574e119e69a17b0a81f33595e283cbbb5db0dc --- /dev/null +++ b/data/vals_ai/meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo/599474d6-c3b6-4a36-9723-5c7c41dbd363.json @@ -0,0 +1,99 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/meta-llama_Llama-3.2-90B-Vision-Instruct-Turbo/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Llama-3.2-90B-Vision-Instruct-Turbo", + "id": "meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo", + "developer": "meta-llama", + "additional_details": { + "vals_model_id": "together/meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:together/meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 48.065, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "48.065", + "raw_stderr": "1.201", + "latency": "56.329", + "cost_per_test": "0.008287", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.201, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta-llama/Llama-3.3-70B-Instruct-Turbo/047eca9b-d4a3-4634-a348-133f0eead4f0.json b/data/vals_ai/meta-llama/Llama-3.3-70B-Instruct-Turbo/047eca9b-d4a3-4634-a348-133f0eead4f0.json new file mode 100644 index 0000000000000000000000000000000000000000..bb045c6acbad4d27de7cdcb0d94496a3327b673e --- /dev/null +++ b/data/vals_ai/meta-llama/Llama-3.3-70B-Instruct-Turbo/047eca9b-d4a3-4634-a348-133f0eead4f0.json @@ -0,0 +1,225 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/meta-llama_Llama-3.3-70B-Instruct-Turbo/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Llama-3.3-70B-Instruct-Turbo", + "id": "meta-llama/Llama-3.3-70B-Instruct-Turbo", + "developer": "meta-llama", + "additional_details": { + "vals_model_id": "together/meta-llama/Llama-3.3-70B-Instruct-Turbo", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:together/meta-llama/Llama-3.3-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 48.485, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "48.485", + "raw_stderr": "3.552", + "latency": "7.474", + "cost_per_test": "0.002331", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.552, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:together/meta-llama/Llama-3.3-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 51.515, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "51.515", + "raw_stderr": "3.552", + "latency": "7.211", + "cost_per_test": "0.000886", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.552, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:together/meta-llama/Llama-3.3-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 50.0, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "50", + "raw_stderr": "2.512", + "latency": "7.342", + "cost_per_test": "0.001608", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.512, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta-llama/Llama-3.3-70B-Instruct-Turbo/15a8de08-8b9d-405f-bbc8-8ce5a8ff3c04.json b/data/vals_ai/meta-llama/Llama-3.3-70B-Instruct-Turbo/15a8de08-8b9d-405f-bbc8-8ce5a8ff3c04.json new file mode 100644 index 0000000000000000000000000000000000000000..015d0706e142cd310d148ab4c0542c275b816abe --- /dev/null +++ b/data/vals_ai/meta-llama/Llama-3.3-70B-Instruct-Turbo/15a8de08-8b9d-405f-bbc8-8ce5a8ff3c04.json @@ -0,0 +1,216 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/meta-llama_Llama-3.3-70B-Instruct-Turbo/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Llama-3.3-70B-Instruct-Turbo", + "id": "meta-llama/Llama-3.3-70B-Instruct-Turbo", + "developer": "meta-llama", + "additional_details": { + "vals_model_id": "together/meta-llama/Llama-3.3-70B-Instruct-Turbo", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:together/meta-llama/Llama-3.3-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 47.343, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "47.343", + "raw_stderr": "1.428", + "latency": "3.836", + "cost_per_test": "0.000532", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.428, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:together/meta-llama/Llama-3.3-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 59.444, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.444", + "raw_stderr": "0.962", + "latency": "3.836", + "cost_per_test": "0.000532", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.962, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:together/meta-llama/Llama-3.3-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 71.545, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.545", + "raw_stderr": "1.29", + "latency": "3.836", + "cost_per_test": "0.000532", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.29, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta-llama/Llama-3.3-70B-Instruct-Turbo/514888e3-2f06-4084-9ae0-c72b7aefec8d.json b/data/vals_ai/meta-llama/Llama-3.3-70B-Instruct-Turbo/514888e3-2f06-4084-9ae0-c72b7aefec8d.json new file mode 100644 index 0000000000000000000000000000000000000000..9dea28a2e72ed74ed8731d8160c9d2291b58e19c --- /dev/null +++ b/data/vals_ai/meta-llama/Llama-3.3-70B-Instruct-Turbo/514888e3-2f06-4084-9ae0-c72b7aefec8d.json @@ -0,0 +1,981 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/meta-llama_Llama-3.3-70B-Instruct-Turbo/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Llama-3.3-70B-Instruct-Turbo", + "id": "meta-llama/Llama-3.3-70B-Instruct-Turbo", + "developer": "meta-llama", + "additional_details": { + "vals_model_id": "together/meta-llama/Llama-3.3-70B-Instruct-Turbo", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:together/meta-llama/Llama-3.3-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.216, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.216", + "raw_stderr": "1.326", + "latency": "4.749", + "cost_per_test": "0.001606", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.326, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:together/meta-llama/Llama-3.3-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.384, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.384", + "raw_stderr": "1.573", + "latency": "3.293", + "cost_per_test": "0.001252", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.573, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:together/meta-llama/Llama-3.3-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 70.406, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.406", + "raw_stderr": "1.357", + "latency": "4.962", + "cost_per_test": "0.001937", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.357, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:together/meta-llama/Llama-3.3-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 72.927, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.927", + "raw_stderr": "2.194", + "latency": "4.266", + "cost_per_test": "0.001692", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.194, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:together/meta-llama/Llama-3.3-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.858, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.858", + "raw_stderr": "1.381", + "latency": "3.751", + "cost_per_test": "0.001456", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.381, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:together/meta-llama/Llama-3.3-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 52.735, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "52.735", + "raw_stderr": "1.604", + "latency": "5.892", + "cost_per_test": "0.001797", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.604, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:together/meta-llama/Llama-3.3-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.105, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.105", + "raw_stderr": "1.55", + "latency": "4.147", + "cost_per_test": "0.001362", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.55, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:together/meta-llama/Llama-3.3-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 65.092, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "65.092", + "raw_stderr": "2.442", + "latency": "1.88", + "cost_per_test": "0.00206", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.442, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:together/meta-llama/Llama-3.3-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 51.771, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "51.771", + "raw_stderr": "1.506", + "latency": "4.355", + "cost_per_test": "0.001661", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.506, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:together/meta-llama/Llama-3.3-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.5, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.5", + "raw_stderr": "1.17", + "latency": "3.932", + "cost_per_test": "0.001958", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.17, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:together/meta-llama/Llama-3.3-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.429, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.429", + "raw_stderr": "1.486", + "latency": "3.776", + "cost_per_test": "0.001113", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.486, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:together/meta-llama/Llama-3.3-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 69.86, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.86", + "raw_stderr": "0.442", + "latency": "4.179", + "cost_per_test": "0.001565", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.442, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:together/meta-llama/Llama-3.3-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 61.523, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "61.523", + "raw_stderr": "2.178", + "latency": "4.873", + "cost_per_test": "0.001224", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.178, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:together/meta-llama/Llama-3.3-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 69.284, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.284", + "raw_stderr": "1.28", + "latency": "4.451", + "cost_per_test": "0.001542", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.28, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:together/meta-llama/Llama-3.3-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.815, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.815", + "raw_stderr": "1.516", + "latency": "4.183", + "cost_per_test": "0.001251", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.516, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta-llama/Llama-3.3-70B-Instruct-Turbo/9988362a-95a3-401c-85c7-9463707cc800.json b/data/vals_ai/meta-llama/Llama-3.3-70B-Instruct-Turbo/9988362a-95a3-401c-85c7-9463707cc800.json new file mode 100644 index 0000000000000000000000000000000000000000..bbfeb9a5d434f7ffd85b4bc8dba99b09c5548bd6 --- /dev/null +++ b/data/vals_ai/meta-llama/Llama-3.3-70B-Instruct-Turbo/9988362a-95a3-401c-85c7-9463707cc800.json @@ -0,0 +1,408 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/meta-llama_Llama-3.3-70B-Instruct-Turbo/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Llama-3.3-70B-Instruct-Turbo", + "id": "meta-llama/Llama-3.3-70B-Instruct-Turbo", + "developer": "meta-llama", + "additional_details": { + "vals_model_id": "together/meta-llama/Llama-3.3-70B-Instruct-Turbo", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:together/meta-llama/Llama-3.3-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.695, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.695", + "raw_stderr": "1.01", + "latency": "0.606", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.01, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:together/meta-llama/Llama-3.3-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.425, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.425", + "raw_stderr": "0.194", + "latency": "0.56", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.194, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:together/meta-llama/Llama-3.3-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.206, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.206", + "raw_stderr": "0.591", + "latency": "0.546", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.591, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:together/meta-llama/Llama-3.3-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.18, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.18", + "raw_stderr": "0.771", + "latency": "0.548", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.771, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:together/meta-llama/Llama-3.3-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.51, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.51", + "raw_stderr": "1.023", + "latency": "0.467", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.023, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:together/meta-llama/Llama-3.3-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 67.065, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "67.065", + "raw_stderr": "3.525", + "latency": "0.563", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.525, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta-llama/Llama-3.3-70B-Instruct-Turbo/cce23d3e-97ad-458d-ac0d-d91976552ffc.json b/data/vals_ai/meta-llama/Llama-3.3-70B-Instruct-Turbo/cce23d3e-97ad-458d-ac0d-d91976552ffc.json new file mode 100644 index 0000000000000000000000000000000000000000..c4fe17600b3d197eb7c66bdf4edd96956c40062c --- /dev/null +++ b/data/vals_ai/meta-llama/Llama-3.3-70B-Instruct-Turbo/cce23d3e-97ad-458d-ac0d-d91976552ffc.json @@ -0,0 +1,288 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/meta-llama_Llama-3.3-70B-Instruct-Turbo/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Llama-3.3-70B-Instruct-Turbo", + "id": "meta-llama/Llama-3.3-70B-Instruct-Turbo", + "developer": "meta-llama", + "additional_details": { + "vals_model_id": "together/meta-llama/Llama-3.3-70B-Instruct-Turbo", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:together/meta-llama/Llama-3.3-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 83.54, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "83.54", + "raw_stderr": "2.066", + "latency": "2.421", + "cost_per_test": "0.000567", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.066, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:together/meta-llama/Llama-3.3-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 4.857, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "4.857", + "raw_stderr": "1.149", + "latency": "3.58", + "cost_per_test": "0.000931", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.149, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:together/meta-llama/Llama-3.3-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 20.627, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "20.627", + "raw_stderr": "2.068", + "latency": "2.835", + "cost_per_test": "0.000732", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.068, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:together/meta-llama/Llama-3.3-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 36.341, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "36.341", + "raw_stderr": "1.047", + "latency": "2.945", + "cost_per_test": "0.000743", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.047, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct/5cb56b9c-6bbe-45a4-994e-1694b5b16c2b.json b/data/vals_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct/5cb56b9c-6bbe-45a4-994e-1694b5b16c2b.json new file mode 100644 index 0000000000000000000000000000000000000000..44fa954d879cce4169ed91e03e03504442d37425 --- /dev/null +++ b/data/vals_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct/5cb56b9c-6bbe-45a4-994e-1694b5b16c2b.json @@ -0,0 +1,216 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/meta-llama_Llama-4-Scout-17B-16E-Instruct/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Llama-4-Scout-17B-16E-Instruct", + "id": "meta-llama/Llama-4-Scout-17B-16E-Instruct", + "developer": "meta-llama", + "additional_details": { + "vals_model_id": "together/meta-llama/Llama-4-Scout-17B-16E-Instruct", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 43.009, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "43.009", + "raw_stderr": "1.416", + "latency": "7.042", + "cost_per_test": "0.000378", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.416, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 55.192, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "55.192", + "raw_stderr": "0.975", + "latency": "7.042", + "cost_per_test": "0.000378", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.975, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 67.375, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.375", + "raw_stderr": "1.341", + "latency": "7.042", + "cost_per_test": "0.000378", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.341, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct/8ffc6edf-04a6-4b53-bb52-0f3af93c5bc2.json b/data/vals_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct/8ffc6edf-04a6-4b53-bb52-0f3af93c5bc2.json new file mode 100644 index 0000000000000000000000000000000000000000..5b20def8b7c087d69972c621d4dabf76dca93012 --- /dev/null +++ b/data/vals_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct/8ffc6edf-04a6-4b53-bb52-0f3af93c5bc2.json @@ -0,0 +1,225 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/meta-llama_Llama-4-Scout-17B-16E-Instruct/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Llama-4-Scout-17B-16E-Instruct", + "id": "meta-llama/Llama-4-Scout-17B-16E-Instruct", + "developer": "meta-llama", + "additional_details": { + "vals_model_id": "together/meta-llama/Llama-4-Scout-17B-16E-Instruct", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 40.909, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "40.909", + "raw_stderr": "3.494", + "latency": "10.759", + "cost_per_test": "0.000746", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.494, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 53.03, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "53.03", + "raw_stderr": "3.55", + "latency": "13.424", + "cost_per_test": "0.000563", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.55, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 46.97, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "46.97", + "raw_stderr": "2.491", + "latency": "12.092", + "cost_per_test": "0.000654", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.491, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct/996aaf99-514f-496b-822d-5250a0aa6d89.json b/data/vals_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct/996aaf99-514f-496b-822d-5250a0aa6d89.json new file mode 100644 index 0000000000000000000000000000000000000000..3f0c9a821c0d44751b6b7c6c3430db33685675ff --- /dev/null +++ b/data/vals_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct/996aaf99-514f-496b-822d-5250a0aa6d89.json @@ -0,0 +1,216 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/meta-llama_Llama-4-Scout-17B-16E-Instruct/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Llama-4-Scout-17B-16E-Instruct", + "id": "meta-llama/Llama-4-Scout-17B-16E-Instruct", + "developer": "meta-llama", + "additional_details": { + "vals_model_id": "together/meta-llama/Llama-4-Scout-17B-16E-Instruct", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 63.037, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.037", + "raw_stderr": "1.361", + "latency": "2.328", + "cost_per_test": "0.000554", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.361, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 57.75, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.75", + "raw_stderr": "0.979", + "latency": "2.426", + "cost_per_test": "0.000572", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.979, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 52.464, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.464", + "raw_stderr": "1.408", + "latency": "2.525", + "cost_per_test": "0.000589", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.408, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct/a188bb09-2679-43ee-a4ac-bb8c9993d234.json b/data/vals_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct/a188bb09-2679-43ee-a4ac-bb8c9993d234.json new file mode 100644 index 0000000000000000000000000000000000000000..4b7f6ef822032b256a580dd7ea5f1559b6fbbf5c --- /dev/null +++ b/data/vals_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct/a188bb09-2679-43ee-a4ac-bb8c9993d234.json @@ -0,0 +1,99 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/meta-llama_Llama-4-Scout-17B-16E-Instruct/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Llama-4-Scout-17B-16E-Instruct", + "id": "meta-llama/Llama-4-Scout-17B-16E-Instruct", + "developer": "meta-llama", + "additional_details": { + "vals_model_id": "together/meta-llama/Llama-4-Scout-17B-16E-Instruct", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 58.752, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "58.752", + "raw_stderr": "1.183", + "latency": "10.506", + "cost_per_test": "0.000835", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.183, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct/a8b39241-53ad-47f0-92c4-b8359aa6025c.json b/data/vals_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct/a8b39241-53ad-47f0-92c4-b8359aa6025c.json new file mode 100644 index 0000000000000000000000000000000000000000..b23afc295b464b870ee46af85e05c283d771b5b7 --- /dev/null +++ b/data/vals_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct/a8b39241-53ad-47f0-92c4-b8359aa6025c.json @@ -0,0 +1,346 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/meta-llama_Llama-4-Scout-17B-16E-Instruct/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Llama-4-Scout-17B-16E-Instruct", + "id": "meta-llama/Llama-4-Scout-17B-16E-Instruct", + "developer": "meta-llama", + "additional_details": { + "vals_model_id": "together/meta-llama/Llama-4-Scout-17B-16E-Instruct", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 28.571, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "28.571", + "raw_stderr": "5.669", + "latency": "32.558", + "cost_per_test": "0.002381", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 5.669, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 14.682, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "14.682", + "raw_stderr": "4.524", + "latency": "20.79", + "cost_per_test": "0.003204", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 4.524, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 5.247, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "5.247", + "raw_stderr": "3.146", + "latency": "68.888", + "cost_per_test": "0.014759", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.146, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 61.249, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "61.249", + "raw_stderr": "5.472", + "latency": "45.238", + "cost_per_test": "0.005317", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 5.472, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 34.834, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "34.834", + "raw_stderr": "3.029", + "latency": "32.862", + "cost_per_test": "0.003634", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.029, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct/b02e0460-78af-4774-bfaf-cef26f96de1c.json b/data/vals_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct/b02e0460-78af-4774-bfaf-cef26f96de1c.json new file mode 100644 index 0000000000000000000000000000000000000000..0e1df9fb5284489fcd8c2badbfbbbd5157dbcb2d --- /dev/null +++ b/data/vals_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct/b02e0460-78af-4774-bfaf-cef26f96de1c.json @@ -0,0 +1,276 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/meta-llama_Llama-4-Scout-17B-16E-Instruct/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Llama-4-Scout-17B-16E-Instruct", + "id": "meta-llama/Llama-4-Scout-17B-16E-Instruct", + "developer": "meta-llama", + "additional_details": { + "vals_model_id": "together/meta-llama/Llama-4-Scout-17B-16E-Instruct", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 52.681, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.681", + "raw_stderr": "1.705", + "latency": "2.476", + "cost_per_test": "0.000367", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.705, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 43.124, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "43.124", + "raw_stderr": "1.691", + "latency": "20.149", + "cost_per_test": "0.023152", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.691, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 46.776, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "46.776", + "raw_stderr": "0.98", + "latency": "9.697", + "cost_per_test": "0.011639", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.98, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 44.522, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "44.522", + "raw_stderr": "1.697", + "latency": "6.467", + "cost_per_test": "0.011399", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.697, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct/bfe06a6b-76a7-495d-9bd1-64111cb7c311.json b/data/vals_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct/bfe06a6b-76a7-495d-9bd1-64111cb7c311.json new file mode 100644 index 0000000000000000000000000000000000000000..1c4ded15b2f910cd46a5af26d94110616b41883c --- /dev/null +++ b/data/vals_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct/bfe06a6b-76a7-495d-9bd1-64111cb7c311.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/meta-llama_Llama-4-Scout-17B-16E-Instruct/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Llama-4-Scout-17B-16E-Instruct", + "id": "meta-llama/Llama-4-Scout-17B-16E-Instruct", + "developer": "meta-llama", + "additional_details": { + "vals_model_id": "together/meta-llama/Llama-4-Scout-17B-16E-Instruct", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 50.593, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "50.593", + "raw_stderr": "1.901", + "latency": "11.323", + "cost_per_test": "0.0017", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.901, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct/c9409a89-ef5f-4290-947b-ea5c1998a305.json b/data/vals_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct/c9409a89-ef5f-4290-947b-ea5c1998a305.json new file mode 100644 index 0000000000000000000000000000000000000000..921f5612f8220b57c1039f1eb3b8e20a40f21e33 --- /dev/null +++ b/data/vals_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct/c9409a89-ef5f-4290-947b-ea5c1998a305.json @@ -0,0 +1,981 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/meta-llama_Llama-4-Scout-17B-16E-Instruct/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Llama-4-Scout-17B-16E-Instruct", + "id": "meta-llama/Llama-4-Scout-17B-16E-Instruct", + "developer": "meta-llama", + "additional_details": { + "vals_model_id": "together/meta-llama/Llama-4-Scout-17B-16E-Instruct", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.937, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.937", + "raw_stderr": "1.336", + "latency": "5.538", + "cost_per_test": "0.000512", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.336, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.665, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.665", + "raw_stderr": "1.528", + "latency": "4.955", + "cost_per_test": "0.000424", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.528, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.353, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.353", + "raw_stderr": "1.281", + "latency": "6.093", + "cost_per_test": "0.000616", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.281, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 69.756, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.756", + "raw_stderr": "2.268", + "latency": "4.796", + "cost_per_test": "0.000542", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.268, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.91, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.91", + "raw_stderr": "1.404", + "latency": "4.802", + "cost_per_test": "0.000468", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.404, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 60.991, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "60.991", + "raw_stderr": "1.567", + "latency": "7.111", + "cost_per_test": "0.000583", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.567, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.883, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.883", + "raw_stderr": "1.572", + "latency": "4.621", + "cost_per_test": "0.000437", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.572, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 58.53, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "58.53", + "raw_stderr": "2.524", + "latency": "5.104", + "cost_per_test": "0.000675", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.524, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 47.048, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "47.048", + "raw_stderr": "1.504", + "latency": "5.019", + "cost_per_test": "0.00053", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.504, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.944, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.944", + "raw_stderr": "1.163", + "latency": "5.799", + "cost_per_test": "0.0006", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.163, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 66.126, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.126", + "raw_stderr": "1.557", + "latency": "3.433", + "cost_per_test": "0.000353", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.557, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 69.632, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.632", + "raw_stderr": "0.445", + "latency": "5.342", + "cost_per_test": "0.000507", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.445, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 57.315, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "57.315", + "raw_stderr": "2.214", + "latency": "4.895", + "cost_per_test": "0.00041", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.214, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.828, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.828", + "raw_stderr": "1.171", + "latency": "8.434", + "cost_per_test": "0.000539", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.171, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.564, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.564", + "raw_stderr": "1.521", + "latency": "4.182", + "cost_per_test": "0.000414", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.521, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct/cf18d680-d0ab-4292-a86d-8bd56b161569.json b/data/vals_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct/cf18d680-d0ab-4292-a86d-8bd56b161569.json new file mode 100644 index 0000000000000000000000000000000000000000..df28bc1ad37abc6f2412f7b16f914c75eea5be5f --- /dev/null +++ b/data/vals_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct/cf18d680-d0ab-4292-a86d-8bd56b161569.json @@ -0,0 +1,408 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/meta-llama_Llama-4-Scout-17B-16E-Instruct/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Llama-4-Scout-17B-16E-Instruct", + "id": "meta-llama/Llama-4-Scout-17B-16E-Instruct", + "developer": "meta-llama", + "additional_details": { + "vals_model_id": "together/meta-llama/Llama-4-Scout-17B-16E-Instruct", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 73.85, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "73.85", + "raw_stderr": "1.086", + "latency": "0.226", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.086, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 72.037, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "72.037", + "raw_stderr": "0.231", + "latency": "0.205", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.231, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 66.836, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "66.836", + "raw_stderr": "0.727", + "latency": "0.234", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.727, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 72.036, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "72.036", + "raw_stderr": "0.479", + "latency": "0.392", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.479, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.837, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.837", + "raw_stderr": "0.91", + "latency": "0.194", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.91, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 69.618, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "69.618", + "raw_stderr": "1.776", + "latency": "1.099", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.776, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct/f2f81fcf-726e-45ff-9def-3cd78d4bd617.json b/data/vals_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct/f2f81fcf-726e-45ff-9def-3cd78d4bd617.json new file mode 100644 index 0000000000000000000000000000000000000000..fabb53cb6de6d012caf7f80dbe5afce888cc2cbf --- /dev/null +++ b/data/vals_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct/f2f81fcf-726e-45ff-9def-3cd78d4bd617.json @@ -0,0 +1,288 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/meta-llama_Llama-4-Scout-17B-16E-Instruct/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Llama-4-Scout-17B-16E-Instruct", + "id": "meta-llama/Llama-4-Scout-17B-16E-Instruct", + "developer": "meta-llama", + "additional_details": { + "vals_model_id": "together/meta-llama/Llama-4-Scout-17B-16E-Instruct", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 87.578, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "87.578", + "raw_stderr": "1.838", + "latency": "14.945", + "cost_per_test": "0.000527", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.838, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 4.286, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "4.286", + "raw_stderr": "1.083", + "latency": "15.586", + "cost_per_test": "0.000694", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.083, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 23.76, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "23.76", + "raw_stderr": "2.175", + "latency": "19.756", + "cost_per_test": "0.000625", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.175, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:together/meta-llama/Llama-4-Scout-17B-16E-Instruct:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 38.541, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "38.541", + "raw_stderr": "1.016", + "latency": "16.762", + "cost_per_test": "0.000615", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.016, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo/04d0bf01-f2eb-46c7-a171-d66f7bdb8107.json b/data/vals_ai/meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo/04d0bf01-f2eb-46c7-a171-d66f7bdb8107.json new file mode 100644 index 0000000000000000000000000000000000000000..8ad18f0302a348271fa0d9ac28e27e13eb4b8606 --- /dev/null +++ b/data/vals_ai/meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo/04d0bf01-f2eb-46c7-a171-d66f7bdb8107.json @@ -0,0 +1,216 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/meta-llama_Meta-Llama-3.1-405B-Instruct-Turbo/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Meta-Llama-3.1-405B-Instruct-Turbo", + "id": "meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo", + "developer": "meta-llama", + "additional_details": { + "vals_model_id": "together/meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:together/meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 49.305, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "49.305", + "raw_stderr": "1.43", + "latency": "23.551", + "cost_per_test": "0.001904", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.43, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:together/meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 60.875, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.875", + "raw_stderr": "0.959", + "latency": "23.551", + "cost_per_test": "0.001904", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.959, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:together/meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 72.445, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "72.445", + "raw_stderr": "1.278", + "latency": "23.551", + "cost_per_test": "0.001904", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.278, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo/9df14b71-c25d-4e5a-abf5-73dc60e3842b.json b/data/vals_ai/meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo/9df14b71-c25d-4e5a-abf5-73dc60e3842b.json new file mode 100644 index 0000000000000000000000000000000000000000..9ee02d3b9d3c5b8d325a1dba0761c0a0830d3294 --- /dev/null +++ b/data/vals_ai/meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo/9df14b71-c25d-4e5a-abf5-73dc60e3842b.json @@ -0,0 +1,276 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/meta-llama_Meta-Llama-3.1-70B-Instruct-Turbo/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Meta-Llama-3.1-70B-Instruct-Turbo", + "id": "meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo", + "developer": "meta-llama", + "additional_details": { + "vals_model_id": "together/meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:together/meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 48.252, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "48.252", + "raw_stderr": "1.706", + "latency": "0.073", + "cost_per_test": "0.001373", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.706, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:together/meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 30.653, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "30.653", + "raw_stderr": "1.574", + "latency": "0.046", + "cost_per_test": "0.100898", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.574, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:together/meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 38.85, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "38.85", + "raw_stderr": "0.95", + "latency": "0.062", + "cost_per_test": "0.053099", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.95, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:together/meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 37.646, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "37.646", + "raw_stderr": "1.654", + "latency": "0.068", + "cost_per_test": "0.057026", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.654, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo/aa8010a9-402e-46d0-a9a2-8f76a5530896.json b/data/vals_ai/meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo/aa8010a9-402e-46d0-a9a2-8f76a5530896.json new file mode 100644 index 0000000000000000000000000000000000000000..e0f5304d0da89995103292b71dbc7762a65d7a95 --- /dev/null +++ b/data/vals_ai/meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo/aa8010a9-402e-46d0-a9a2-8f76a5530896.json @@ -0,0 +1,216 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/meta-llama_Meta-Llama-3.1-70B-Instruct-Turbo/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Meta-Llama-3.1-70B-Instruct-Turbo", + "id": "meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo", + "developer": "meta-llama", + "additional_details": { + "vals_model_id": "together/meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:together/meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 44.154, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "44.154", + "raw_stderr": "1.42", + "latency": "4.327", + "cost_per_test": "0.000488", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.42, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:together/meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 56.174, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.174", + "raw_stderr": "0.973", + "latency": "4.327", + "cost_per_test": "0.000488", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.973, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:together/meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 68.193, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.193", + "raw_stderr": "1.332", + "latency": "4.327", + "cost_per_test": "0.000488", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.332, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo/fb88386e-0f51-4995-b1ec-e1d07643b2be.json b/data/vals_ai/meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo/fb88386e-0f51-4995-b1ec-e1d07643b2be.json new file mode 100644 index 0000000000000000000000000000000000000000..999a61a01a43bd40b3e59db59ea3e7081e883b20 --- /dev/null +++ b/data/vals_ai/meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo/fb88386e-0f51-4995-b1ec-e1d07643b2be.json @@ -0,0 +1,276 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/meta-llama_Meta-Llama-3.1-8B-Instruct-Turbo/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Meta-Llama-3.1-8B-Instruct-Turbo", + "id": "meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo", + "developer": "meta-llama", + "additional_details": { + "vals_model_id": "together/meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:together/meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 50.816, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "50.816", + "raw_stderr": "1.707", + "latency": "0.012", + "cost_per_test": "0.000288", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.707, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:together/meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 25.641, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "25.641", + "raw_stderr": "1.491", + "latency": "0.049", + "cost_per_test": "0.020359", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.491, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:together/meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 37.801, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "37.801", + "raw_stderr": "0.934", + "latency": "0.038", + "cost_per_test": "0.010777", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.934, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:together/meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 36.946, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "36.946", + "raw_stderr": "1.648", + "latency": "0.053", + "cost_per_test": "0.011684", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.648, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo/fed3901e-72c0-461c-9089-670b7acc1bfd.json b/data/vals_ai/meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo/fed3901e-72c0-461c-9089-670b7acc1bfd.json new file mode 100644 index 0000000000000000000000000000000000000000..74b781ce5b471777f65be905496465a7fb6f0914 --- /dev/null +++ b/data/vals_ai/meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo/fed3901e-72c0-461c-9089-670b7acc1bfd.json @@ -0,0 +1,216 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/meta-llama_Meta-Llama-3.1-8B-Instruct-Turbo/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Meta-Llama-3.1-8B-Instruct-Turbo", + "id": "meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo", + "developer": "meta-llama", + "additional_details": { + "vals_model_id": "together/meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:together/meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 21.504, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "21.504", + "raw_stderr": "1.175", + "latency": "2.458", + "cost_per_test": "0.000107", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.175, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:together/meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 32.338, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "32.338", + "raw_stderr": "0.92", + "latency": "2.458", + "cost_per_test": "0.000107", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.92, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:together/meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 43.173, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "43.173", + "raw_stderr": "1.416", + "latency": "2.458", + "cost_per_test": "0.000107", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.416, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark/09f5566e-59d1-4163-b5b8-888bd2562a6c.json b/data/vals_ai/meta/muse_spark/09f5566e-59d1-4163-b5b8-888bd2562a6c.json new file mode 100644 index 0000000000000000000000000000000000000000..f36564e21a96ccc8a13c4ceafd7a4cb7ab3daf9c --- /dev/null +++ b/data/vals_ai/meta/muse_spark/09f5566e-59d1-4163-b5b8-888bd2562a6c.json @@ -0,0 +1,225 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/meta_muse_spark/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark", + "id": "meta/muse_spark", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:meta/muse_spark:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 88.384, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.384", + "raw_stderr": "2.277", + "latency": "94.376", + "cost_per_test": "0.003188", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 2.277, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:meta/muse_spark:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 90.909, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.909", + "raw_stderr": "2.093", + "latency": "94.785", + "cost_per_test": "0.002959", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 2.093, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:meta/muse_spark:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 89.646, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.646", + "raw_stderr": "1.546", + "latency": "94.58", + "cost_per_test": "0.003074", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.546, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark/1d32590b-27b9-450f-ab3c-0633472140dc.json b/data/vals_ai/meta/muse_spark/1d32590b-27b9-450f-ab3c-0633472140dc.json new file mode 100644 index 0000000000000000000000000000000000000000..e5f6a11cf9b53cf9d4ba90370fda25c412849f59 --- /dev/null +++ b/data/vals_ai/meta/muse_spark/1d32590b-27b9-450f-ab3c-0633472140dc.json @@ -0,0 +1,216 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/meta_muse_spark/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark", + "id": "meta/muse_spark", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:meta/muse_spark:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 65.576, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.576", + "raw_stderr": "1.359", + "latency": "57.522", + "cost_per_test": "0.001922", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.359, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:meta/muse_spark:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 77.678, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "77.678", + "raw_stderr": "0.806", + "latency": "57.522", + "cost_per_test": "0.001922", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.806, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:meta/muse_spark:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 89.779, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "89.779", + "raw_stderr": "0.866", + "latency": "57.522", + "cost_per_test": "0.001922", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.866, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark/21f0af09-6486-44eb-9b92-eefe289b1538.json b/data/vals_ai/meta/muse_spark/21f0af09-6486-44eb-9b92-eefe289b1538.json new file mode 100644 index 0000000000000000000000000000000000000000..b4ad7e259e2bbbe3a2a0da13869f02af381bfc63 --- /dev/null +++ b/data/vals_ai/meta/muse_spark/21f0af09-6486-44eb-9b92-eefe289b1538.json @@ -0,0 +1,96 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/meta_muse_spark/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark", + "id": "meta/muse_spark", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:meta/muse_spark:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 17.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "17", + "raw_stderr": "3.756", + "latency": "478.958", + "cost_per_test": "0.067086", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 3.756, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark/2371e989-36e6-45db-8435-37e92722a9c3.json b/data/vals_ai/meta/muse_spark/2371e989-36e6-45db-8435-37e92722a9c3.json new file mode 100644 index 0000000000000000000000000000000000000000..3f89c02d972c3236d48e59dc9668a4b3637c073d --- /dev/null +++ b/data/vals_ai/meta/muse_spark/2371e989-36e6-45db-8435-37e92722a9c3.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/meta_muse_spark/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark", + "id": "meta/muse_spark", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:meta/muse_spark:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 85.902, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "85.902", + "raw_stderr": "1.847", + "latency": "191.473", + "cost_per_test": "0.007681", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.847, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark/2f67d2b1-c394-45e6-a5dd-210129f4040b.json b/data/vals_ai/meta/muse_spark/2f67d2b1-c394-45e6-a5dd-210129f4040b.json new file mode 100644 index 0000000000000000000000000000000000000000..3a9da8ba4081128f96038ff15b67088593187dd7 --- /dev/null +++ b/data/vals_ai/meta/muse_spark/2f67d2b1-c394-45e6-a5dd-210129f4040b.json @@ -0,0 +1,276 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/meta_muse_spark/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark", + "id": "meta/muse_spark", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:meta/muse_spark:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 64.802, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.802", + "raw_stderr": "1.63", + "latency": "41.977", + "cost_per_test": "0.001586", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.63, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:meta/muse_spark:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.023, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.023", + "raw_stderr": "1.672", + "latency": "38.935", + "cost_per_test": "0.022947", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.672, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:meta/muse_spark:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 65.113, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.113", + "raw_stderr": "0.936", + "latency": "40.057", + "cost_per_test": "0.011749", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.936, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:meta/muse_spark:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 70.513, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "70.513", + "raw_stderr": "1.557", + "latency": "39.26", + "cost_per_test": "0.010715", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.557, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark/79c055b9-fc03-49db-970f-19076fc8845f.json b/data/vals_ai/meta/muse_spark/79c055b9-fc03-49db-970f-19076fc8845f.json new file mode 100644 index 0000000000000000000000000000000000000000..df19b63f9d9d3dc0fa18917ecff4ce586851387c --- /dev/null +++ b/data/vals_ai/meta/muse_spark/79c055b9-fc03-49db-970f-19076fc8845f.json @@ -0,0 +1,981 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/meta_muse_spark/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark", + "id": "meta/muse_spark", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:meta/muse_spark:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.724, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.724", + "raw_stderr": "0.906", + "latency": "34.668", + "cost_per_test": "0.001406", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.906, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:meta/muse_spark:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.888, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.888", + "raw_stderr": "0.972", + "latency": "37.05", + "cost_per_test": "0.001453", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.972, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:meta/muse_spark:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.519, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.519", + "raw_stderr": "0.828", + "latency": "50.435", + "cost_per_test": "0.001905", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.828, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:meta/muse_spark:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.0, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90", + "raw_stderr": "1.482", + "latency": "51.459", + "cost_per_test": "0.001545", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.482, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:meta/muse_spark:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.81, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.81", + "raw_stderr": "1.041", + "latency": "34.11", + "cost_per_test": "0.001362", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.041, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:meta/muse_spark:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.623, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.623", + "raw_stderr": "1.159", + "latency": "79.232", + "cost_per_test": "0.003009", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.159, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:meta/muse_spark:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.807, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.807", + "raw_stderr": "1.377", + "latency": "59.9", + "cost_per_test": "0.001615", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.377, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:meta/muse_spark:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.84, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.84", + "raw_stderr": "2.016", + "latency": "47.223", + "cost_per_test": "0.001956", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 2.016, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:meta/muse_spark:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.023, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.023", + "raw_stderr": "1.305", + "latency": "47.886", + "cost_per_test": "0.001906", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.305, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:meta/muse_spark:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 95.041, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "95.041", + "raw_stderr": "0.591", + "latency": "56.02", + "cost_per_test": "0.001713", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.591, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:meta/muse_spark:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.632, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.632", + "raw_stderr": "1.186", + "latency": "32.647", + "cost_per_test": "0.001278", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.186, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:meta/muse_spark:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.316, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.316", + "raw_stderr": "0.327", + "latency": "46.242", + "cost_per_test": "0.00167", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.327, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:meta/muse_spark:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.371, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.371", + "raw_stderr": "1.582", + "latency": "34.617", + "cost_per_test": "0.001328", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.582, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:meta/muse_spark:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.687, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.687", + "raw_stderr": "0.722", + "latency": "50.858", + "cost_per_test": "0.001656", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.722, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:meta/muse_spark:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.466, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.466", + "raw_stderr": "1.211", + "latency": "31.285", + "cost_per_test": "0.001246", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.211, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark/8ac68e80-e26d-4471-884d-2c13716f0fab.json b/data/vals_ai/meta/muse_spark/8ac68e80-e26d-4471-884d-2c13716f0fab.json new file mode 100644 index 0000000000000000000000000000000000000000..e172ea668977b55cc8a409f7bfdf3c79ce065efe --- /dev/null +++ b/data/vals_ai/meta/muse_spark/8ac68e80-e26d-4471-884d-2c13716f0fab.json @@ -0,0 +1,351 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/meta_muse_spark/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark", + "id": "meta/muse_spark", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:meta/muse_spark:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 40.476, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "40.476", + "raw_stderr": "7.666", + "latency": "581.828", + "cost_per_test": "0.565926", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 7.666, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:meta/muse_spark:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 72.797, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "72.797", + "raw_stderr": "2.76", + "latency": "512.022", + "cost_per_test": "0.471902", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 2.76, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:meta/muse_spark:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 85.052, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "85.052", + "raw_stderr": "2.567", + "latency": "452.319", + "cost_per_test": "0.379817", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 2.567, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:meta/muse_spark:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "607.768", + "cost_per_test": "0.536316", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:meta/muse_spark:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 74.4, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "74.4", + "raw_stderr": "1.954", + "latency": "495.294", + "cost_per_test": "0.444473", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.954, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark/8b0c4e3d-3653-4f6e-b17d-351bac04a964.json b/data/vals_ai/meta/muse_spark/8b0c4e3d-3653-4f6e-b17d-351bac04a964.json new file mode 100644 index 0000000000000000000000000000000000000000..e310fc245944541bf4b415da39641075a28013ce --- /dev/null +++ b/data/vals_ai/meta/muse_spark/8b0c4e3d-3653-4f6e-b17d-351bac04a964.json @@ -0,0 +1,99 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/meta_muse_spark/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark", + "id": "meta/muse_spark", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:meta/muse_spark:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 87.399, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.399", + "raw_stderr": "0.798", + "latency": "48.272", + "cost_per_test": "0.001746", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.798, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark/bf4673a2-2511-4e02-855e-06c05c43ddd7.json b/data/vals_ai/meta/muse_spark/bf4673a2-2511-4e02-855e-06c05c43ddd7.json new file mode 100644 index 0000000000000000000000000000000000000000..c9d1c557d5b0d3c04374a521086604c33f09a07a --- /dev/null +++ b/data/vals_ai/meta/muse_spark/bf4673a2-2511-4e02-855e-06c05c43ddd7.json @@ -0,0 +1,414 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/meta_muse_spark/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark", + "id": "meta/muse_spark", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:meta/muse_spark:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 89.47, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.47", + "raw_stderr": "0.814", + "latency": "28.603", + "cost_per_test": "0.00096", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.814, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:meta/muse_spark:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.813, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.813", + "raw_stderr": "0.201", + "latency": "31.722", + "cost_per_test": "0.001106", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.201, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:meta/muse_spark:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 88.799, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.799", + "raw_stderr": "0.577", + "latency": "38.275", + "cost_per_test": "0.001456", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.577, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:meta/muse_spark:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 84.217, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.217", + "raw_stderr": "0.416", + "latency": "34.203", + "cost_per_test": "0.001178", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.416, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:meta/muse_spark:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.101, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.101", + "raw_stderr": "0.886", + "latency": "28.855", + "cost_per_test": "0.001036", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.886, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:meta/muse_spark:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.9, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.9", + "raw_stderr": "1.582", + "latency": "43.56", + "cost_per_test": "0.001334", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.582, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark/c0842b3d-ef36-481b-9138-fcf668f52d47.json b/data/vals_ai/meta/muse_spark/c0842b3d-ef36-481b-9138-fcf668f52d47.json new file mode 100644 index 0000000000000000000000000000000000000000..4fb6af11e5e0ade7caf251cdafa720dcdd48c668 --- /dev/null +++ b/data/vals_ai/meta/muse_spark/c0842b3d-ef36-481b-9138-fcf668f52d47.json @@ -0,0 +1,96 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/meta_muse_spark/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark", + "id": "meta/muse_spark", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:meta/muse_spark:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 19.674, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "19.674", + "raw_stderr": "3.966", + "latency": "981.298", + "cost_per_test": "0.579705", + "max_output_tokens": "131072", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 3.966, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark_1_1/0c87624a-9a86-4184-b8a1-99053c2c7000.json b/data/vals_ai/meta/muse_spark_1_1/0c87624a-9a86-4184-b8a1-99053c2c7000.json new file mode 100644 index 0000000000000000000000000000000000000000..4ce5d0fca220c23323eed966f06c92ccd0d532d8 --- /dev/null +++ b/data/vals_ai/meta/muse_spark_1_1/0c87624a-9a86-4184-b8a1-99053c2c7000.json @@ -0,0 +1,676 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/emb/meta_muse_spark_1_1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - EMB", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark_1_1", + "id": "meta/muse_spark_1_1", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark_1_1", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "emb:comps:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.emb.comps", + "source_data": { + "dataset_name": "EMB - Comps", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "comps", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Comps).", + "metric_id": "vals_ai.emb.comps.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 52.72, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "comps", + "task_name": "Comps", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.72", + "raw_stderr": "7.879", + "latency": "948.276", + "cost_per_test": "1.857162", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 7.879, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:dataroom:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.emb.dataroom", + "source_data": { + "dataset_name": "EMB - Dataroom", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dataroom", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Dataroom).", + "metric_id": "vals_ai.emb.dataroom.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 78.618, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dataroom", + "task_name": "Dataroom", + "dataset_type": "private", + "industry": "finance", + "raw_score": "78.618", + "raw_stderr": "5.76", + "latency": "706.309", + "cost_per_test": "0.749924", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 5.76, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:dcf:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.emb.dcf", + "source_data": { + "dataset_name": "EMB - DCF", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dcf", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (DCF).", + "metric_id": "vals_ai.emb.dcf.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 46.328, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dcf", + "task_name": "DCF", + "dataset_type": "private", + "industry": "finance", + "raw_score": "46.328", + "raw_stderr": "8.065", + "latency": "834.416", + "cost_per_test": "0.996697", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 8.065, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:lbo:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.emb.lbo", + "source_data": { + "dataset_name": "EMB - LBO", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "lbo", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (LBO).", + "metric_id": "vals_ai.emb.lbo.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 49.767, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "lbo", + "task_name": "LBO", + "dataset_type": "private", + "industry": "finance", + "raw_score": "49.767", + "raw_stderr": "5.366", + "latency": "1292.905", + "cost_per_test": "1.737996", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 5.366, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:ma:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.emb.ma", + "source_data": { + "dataset_name": "EMB - M&A", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "ma", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (M&A).", + "metric_id": "vals_ai.emb.ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 61.504, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "ma", + "task_name": "M&A", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.504", + "raw_stderr": "9.309", + "latency": "808.487", + "cost_per_test": "0.691326", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 9.309, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:misc:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.emb.misc", + "source_data": { + "dataset_name": "EMB - Misc", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "misc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Misc).", + "metric_id": "vals_ai.emb.misc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 48.03, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "misc", + "task_name": "Misc", + "dataset_type": "private", + "industry": "finance", + "raw_score": "48.03", + "raw_stderr": "9.095", + "latency": "779.582", + "cost_per_test": "0.767546", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 9.095, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:operating:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.emb.operating", + "source_data": { + "dataset_name": "EMB - Operating", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "operating", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Operating).", + "metric_id": "vals_ai.emb.operating.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 57.562, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "operating", + "task_name": "Operating", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.562", + "raw_stderr": "7.506", + "latency": "801.963", + "cost_per_test": "0.788153", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 7.506, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:overall:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.emb.overall", + "source_data": { + "dataset_name": "EMB - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Overall).", + "metric_id": "vals_ai.emb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 56.361, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.361", + "raw_stderr": "3.153", + "latency": "891.512", + "cost_per_test": "1.112498", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 3.153, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:scratch:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.emb.scratch", + "source_data": { + "dataset_name": "EMB - Scratch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "scratch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Scratch).", + "metric_id": "vals_ai.emb.scratch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 63.285, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "scratch", + "task_name": "Scratch", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.285", + "raw_stderr": "2.57", + "latency": "868.085", + "cost_per_test": "1.015739", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 2.57, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:template:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.emb.template", + "source_data": { + "dataset_name": "EMB - Template", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "template", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Template).", + "metric_id": "vals_ai.emb.template.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 49.41, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "template", + "task_name": "Template", + "dataset_type": "private", + "industry": "finance", + "raw_score": "49.41", + "raw_stderr": "5.188", + "latency": "914.94", + "cost_per_test": "1.209257", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 5.188, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark_1_1/0fd43575-835c-4567-90c7-47b474ec257c.json b/data/vals_ai/meta/muse_spark_1_1/0fd43575-835c-4567-90c7-47b474ec257c.json new file mode 100644 index 0000000000000000000000000000000000000000..7b443e87fea153ea32cb6939dafec2125cba2243 --- /dev/null +++ b/data/vals_ai/meta/muse_spark_1_1/0fd43575-835c-4567-90c7-47b474ec257c.json @@ -0,0 +1,536 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_multimodal_index/meta_muse_spark_1_1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Multimodal Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark_1_1", + "id": "meta/muse_spark_1_1", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark_1_1", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_multimodal_index:corp_fin_v2:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.vals_multimodal_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Multimodal Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (CorpFin v2).", + "metric_id": "vals_ai.vals_multimodal_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 73.077, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "73.077", + "raw_stderr": "1.514", + "latency": "25.409", + "cost_per_test": "0.10923", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.514, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:finance_agent:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.vals_multimodal_index.finance_agent", + "source_data": { + "dataset_name": "Vals Multimodal Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_multimodal_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 57.109, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "57.109", + "raw_stderr": "1.089", + "latency": "438.872", + "cost_per_test": "0.71378", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.089, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:mortgage_tax:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.vals_multimodal_index.mortgage_tax", + "source_data": { + "dataset_name": "Vals Multimodal Index - Mortgage Tax", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "mortgage_tax", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Mortgage Tax).", + "metric_id": "vals_ai.vals_multimodal_index.mortgage_tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 66.136, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "mortgage_tax", + "task_name": "Mortgage Tax", + "dataset_type": "private", + "industry": "index", + "raw_score": "66.136", + "raw_stderr": "0.928", + "latency": "24.818", + "cost_per_test": "0.022312", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.928, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:overall:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.vals_multimodal_index.overall", + "source_data": { + "dataset_name": "Vals Multimodal Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Overall).", + "metric_id": "vals_ai.vals_multimodal_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 66.742, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "66.742", + "raw_stderr": "1.151", + "latency": "288.194", + "cost_per_test": "0.366218", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.151, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:sage:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.vals_multimodal_index.sage", + "source_data": { + "dataset_name": "Vals Multimodal Index - SAGE", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "sage", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SAGE).", + "metric_id": "vals_ai.vals_multimodal_index.sage.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 45.58, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "sage", + "task_name": "SAGE", + "dataset_type": "private", + "industry": "index", + "raw_score": "45.58", + "raw_stderr": "3.276", + "latency": "49.948", + "cost_per_test": "0.025502", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 3.276, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:swebench:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.vals_multimodal_index.swebench", + "source_data": { + "dataset_name": "Vals Multimodal Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SWE-bench).", + "metric_id": "vals_ai.vals_multimodal_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 78.431, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "78.431", + "raw_stderr": "4.072", + "latency": "238.11", + "cost_per_test": "0.352022" + }, + "uncertainty": { + "standard_error": { + "value": 4.072, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:terminal_bench_2_1:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.vals_multimodal_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Multimodal Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_multimodal_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 69.288, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "69.288", + "raw_stderr": "0.991", + "latency": "624.772", + "cost_per_test": "0.417396", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.991, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:vibe_code_bench:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.vals_multimodal_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Multimodal Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_multimodal_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 72.412, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "72.412", + "raw_stderr": "5.168", + "latency": "615.428", + "cost_per_test": "0.923283", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 5.168, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark_1_1/16aa35fe-4d74-4370-a5ce-c19f7b2447fb.json b/data/vals_ai/meta/muse_spark_1_1/16aa35fe-4d74-4370-a5ce-c19f7b2447fb.json new file mode 100644 index 0000000000000000000000000000000000000000..f45a9e5956e4a4170af75243ea000d0d61578343 --- /dev/null +++ b/data/vals_ai/meta/muse_spark_1_1/16aa35fe-4d74-4370-a5ce-c19f7b2447fb.json @@ -0,0 +1,408 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/meta_muse_spark_1_1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark_1_1", + "id": "meta/muse_spark_1_1", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark_1_1", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 73.077, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "73.077", + "raw_stderr": "1.514", + "latency": "25.409", + "cost_per_test": "0.10923", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.514, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 57.109, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "57.109", + "raw_stderr": "1.089", + "latency": "438.872", + "cost_per_test": "0.71378", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.089, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_index:overall:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 68.405, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "68.405", + "raw_stderr": "1.272", + "latency": "388.518", + "cost_per_test": "0.503142", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.272, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 78.431, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "78.431", + "raw_stderr": "4.072", + "latency": "238.11", + "cost_per_test": "0.352022" + }, + "uncertainty": { + "standard_error": { + "value": 4.072, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 69.288, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "69.288", + "raw_stderr": "0.991", + "latency": "624.772", + "cost_per_test": "0.417396", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.991, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 72.412, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "72.412", + "raw_stderr": "5.168", + "latency": "615.428", + "cost_per_test": "0.923283", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 5.168, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark_1_1/220ba97f-ffed-43ec-ab9c-b9a76c51445a.json b/data/vals_ai/meta/muse_spark_1_1/220ba97f-ffed-43ec-ab9c-b9a76c51445a.json new file mode 100644 index 0000000000000000000000000000000000000000..1816cc9d21a49da9885577c74426481d85e55959 --- /dev/null +++ b/data/vals_ai/meta/muse_spark_1_1/220ba97f-ffed-43ec-ab9c-b9a76c51445a.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/meta_muse_spark_1_1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark_1_1", + "id": "meta/muse_spark_1_1", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark_1_1", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 90.404, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.404", + "raw_stderr": "2.616", + "latency": "67.934", + "cost_per_test": "0.030784", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 2.616, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 91.919, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.919", + "raw_stderr": "3.296", + "latency": "82.463", + "cost_per_test": "0.037824", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 3.296, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 91.162, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.162", + "raw_stderr": "2.104", + "latency": "75.198", + "cost_per_test": "0.034304", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 2.104, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark_1_1/2b8bde1e-7b2d-4ad2-ad10-4180beb45800.json b/data/vals_ai/meta/muse_spark_1_1/2b8bde1e-7b2d-4ad2-ad10-4180beb45800.json new file mode 100644 index 0000000000000000000000000000000000000000..3ab99d67f91df00cc8420228c30fd94a659667c3 --- /dev/null +++ b/data/vals_ai/meta/muse_spark_1_1/2b8bde1e-7b2d-4ad2-ad10-4180beb45800.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/meta_muse_spark_1_1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark_1_1", + "id": "meta/muse_spark_1_1", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark_1_1", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.758, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.758", + "raw_stderr": "0.617", + "latency": "18.987", + "cost_per_test": "0.00869", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.617, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "lcb:hard:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 69.143, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "69.143", + "raw_stderr": "2.469", + "latency": "238.402", + "cost_per_test": "0.100507", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 2.469, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "lcb:medium:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 89.817, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "89.817", + "raw_stderr": "1.545", + "latency": "82.371", + "cost_per_test": "0.030294", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.545, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "lcb:overall:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 85.906, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "85.906", + "raw_stderr": "0.992", + "latency": "113.253", + "cost_per_test": "0.046497", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.992, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark_1_1/32416658-82f5-4434-b860-0c18a762ebc5.json b/data/vals_ai/meta/muse_spark_1_1/32416658-82f5-4434-b860-0c18a762ebc5.json new file mode 100644 index 0000000000000000000000000000000000000000..0d22e4551f8e1381f36b9b11a12aae6a5097eb87 --- /dev/null +++ b/data/vals_ai/meta/muse_spark_1_1/32416658-82f5-4434-b860-0c18a762ebc5.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/meta_muse_spark_1_1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark_1_1", + "id": "meta/muse_spark_1_1", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark_1_1", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 39.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "39", + "raw_stderr": "4.902", + "latency": "1056.902", + "cost_per_test": "0.519718", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 4.902, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark_1_1/3c3d04e9-2bd6-4734-bf1c-99a881f5c25c.json b/data/vals_ai/meta/muse_spark_1_1/3c3d04e9-2bd6-4734-bf1c-99a881f5c25c.json new file mode 100644 index 0000000000000000000000000000000000000000..c054854cced56fb0c3fb25909e59adb1c0ee4374 --- /dev/null +++ b/data/vals_ai/meta/muse_spark_1_1/3c3d04e9-2bd6-4734-bf1c-99a881f5c25c.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/meta_muse_spark_1_1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark_1_1", + "id": "meta/muse_spark_1_1", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark_1_1", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 88.888, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "88.888", + "raw_stderr": "1.95", + "latency": "63.335", + "cost_per_test": "0.034628", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.95, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark_1_1/5eb7f362-724e-4116-909c-ae33b80498aa.json b/data/vals_ai/meta/muse_spark_1_1/5eb7f362-724e-4116-909c-ae33b80498aa.json new file mode 100644 index 0000000000000000000000000000000000000000..53fa753d0c23b2507901879a1694dac325cc2bbe --- /dev/null +++ b/data/vals_ai/meta/muse_spark_1_1/5eb7f362-724e-4116-909c-ae33b80498aa.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/meta_muse_spark_1_1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark_1_1", + "id": "meta/muse_spark_1_1", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark_1_1", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 74.483, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "74.483", + "raw_stderr": "1.229", + "latency": "30.7", + "cost_per_test": "0.025379", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.229, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 66.136, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.136", + "raw_stderr": "0.928", + "latency": "24.818", + "cost_per_test": "0.022312", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.928, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 57.79, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.79", + "raw_stderr": "1.392", + "latency": "18.935", + "cost_per_test": "0.019246", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.392, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark_1_1/68a3adaf-d542-4fe0-b2bf-58e45d6014e1.json b/data/vals_ai/meta/muse_spark_1_1/68a3adaf-d542-4fe0-b2bf-58e45d6014e1.json new file mode 100644 index 0000000000000000000000000000000000000000..1d524b558e96c5187c1d6012904f051344958e90 --- /dev/null +++ b/data/vals_ai/meta/muse_spark_1_1/68a3adaf-d542-4fe0-b2bf-58e45d6014e1.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/meta_muse_spark_1_1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark_1_1", + "id": "meta/muse_spark_1_1", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark_1_1", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.979, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.979", + "raw_stderr": "0.896", + "latency": "15.731", + "cost_per_test": "0.008217", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.896, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.55, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.55", + "raw_stderr": "0.817", + "latency": "20.041", + "cost_per_test": "0.009893", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.817, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.375, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.375", + "raw_stderr": "0.744", + "latency": "22.142", + "cost_per_test": "0.011437", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.744, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.732, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.732", + "raw_stderr": "1.498", + "latency": "18.999", + "cost_per_test": "0.009764", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.498, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.232, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.232", + "raw_stderr": "1.014", + "latency": "15.915", + "cost_per_test": "0.008745", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.014, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.132, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.132", + "raw_stderr": "1.043", + "latency": "49.792", + "cost_per_test": "0.024833", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.043, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.374, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.374", + "raw_stderr": "1.309", + "latency": "19.769", + "cost_per_test": "0.0099", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.309, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.79, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.79", + "raw_stderr": "2.057", + "latency": "16.33", + "cost_per_test": "0.00874", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 2.057, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.476, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.476", + "raw_stderr": "1.288", + "latency": "25.461", + "cost_per_test": "0.012445", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.288, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 96.003, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "96.003", + "raw_stderr": "0.547", + "latency": "18.126", + "cost_per_test": "0.009214", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.547, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.281, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.281", + "raw_stderr": "1.18", + "latency": "16.9", + "cost_per_test": "0.008698", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.18, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.726, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.726", + "raw_stderr": "0.318", + "latency": "21.03", + "cost_per_test": "0.010704", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.318, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.174, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.174", + "raw_stderr": "1.507", + "latency": "18.405", + "cost_per_test": "0.009215", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.507, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.841, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.841", + "raw_stderr": "0.69", + "latency": "21.651", + "cost_per_test": "0.011016", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.69, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.218, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.218", + "raw_stderr": "1.187", + "latency": "15.16", + "cost_per_test": "0.007738", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.187, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark_1_1/773ec7a6-8ecb-401f-adbe-416c49f23103.json b/data/vals_ai/meta/muse_spark_1_1/773ec7a6-8ecb-401f-adbe-416c49f23103.json new file mode 100644 index 0000000000000000000000000000000000000000..9ff2b9c1439f7bd72ed089d31500521fce35a6df --- /dev/null +++ b/data/vals_ai/meta/muse_spark_1_1/773ec7a6-8ecb-401f-adbe-416c49f23103.json @@ -0,0 +1,366 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/meta_muse_spark_1_1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark_1_1", + "id": "meta/muse_spark_1_1", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark_1_1", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 39.683, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "39.683", + "raw_stderr": "6.148", + "latency": "52.873", + "cost_per_test": "0.025583", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 6.148, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "sage:automata:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 28.174, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "28.174", + "raw_stderr": "5.654", + "latency": "58.277", + "cost_per_test": "0.026942", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 5.654, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "sage:calculus:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 74.691, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "74.691", + "raw_stderr": "5.469", + "latency": "56.224", + "cost_per_test": "0.05685", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 5.469, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 68.884, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "68.884", + "raw_stderr": "5.179", + "latency": "38.693", + "cost_per_test": "0.02398", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 5.179, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "sage:overall:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 45.58, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "45.58", + "raw_stderr": "3.276", + "latency": "49.948", + "cost_per_test": "0.025502", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 3.276, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark_1_1/98badcee-b0a3-4f65-9d05-177f68c578be.json b/data/vals_ai/meta/muse_spark_1_1/98badcee-b0a3-4f65-9d05-177f68c578be.json new file mode 100644 index 0000000000000000000000000000000000000000..a32fc0b30b91067240597ebfb1f54bf8e4c3a8af --- /dev/null +++ b/data/vals_ai/meta/muse_spark_1_1/98badcee-b0a3-4f65-9d05-177f68c578be.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/meta_muse_spark_1_1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark_1_1", + "id": "meta/muse_spark_1_1", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark_1_1", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 66.434, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.434", + "raw_stderr": "1.612", + "latency": "18.683", + "cost_per_test": "0.009465", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.612, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 74.359, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "74.359", + "raw_stderr": "1.491", + "latency": "39.377", + "cost_per_test": "0.230768", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.491, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 71.29, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.29", + "raw_stderr": "0.889", + "latency": "27.823", + "cost_per_test": "0.10923", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.889, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 73.077, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "73.077", + "raw_stderr": "1.514", + "latency": "25.409", + "cost_per_test": "0.087456", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.514, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark_1_1/994cb92e-3965-4984-bedb-2b1de68911ec.json b/data/vals_ai/meta/muse_spark_1_1/994cb92e-3965-4984-bedb-2b1de68911ec.json new file mode 100644 index 0000000000000000000000000000000000000000..451e3b2a3691c9ee1b0ec4483cdeab6b651c10da --- /dev/null +++ b/data/vals_ai/meta/muse_spark_1_1/994cb92e-3965-4984-bedb-2b1de68911ec.json @@ -0,0 +1,371 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/meta_muse_spark_1_1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark_1_1", + "id": "meta/muse_spark_1_1", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark_1_1", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 52.381, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "52.381", + "raw_stderr": "7.8", + "latency": "339.169", + "cost_per_test": "0.682916", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 7.8, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 82.375, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "82.375", + "raw_stderr": "2.363", + "latency": "227.782", + "cost_per_test": "0.381795", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 2.363, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 88.144, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "88.144", + "raw_stderr": "2.327", + "latency": "151.063", + "cost_per_test": "0.235316", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 2.327, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66.667", + "raw_stderr": "33.333", + "latency": "518.434", + "cost_per_test": "0.676282", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "swebench:overall:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 82.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "82", + "raw_stderr": "1.72", + "latency": "209.115", + "cost_per_test": "0.352022", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.72, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark_1_1/a0eceef9-df05-4e9c-88cb-b33471334aef.json b/data/vals_ai/meta/muse_spark_1_1/a0eceef9-df05-4e9c-88cb-b33471334aef.json new file mode 100644 index 0000000000000000000000000000000000000000..6aec21577436ccfc9bd6441c772ab7a2d035473f --- /dev/null +++ b/data/vals_ai/meta/muse_spark_1_1/a0eceef9-df05-4e9c-88cb-b33471334aef.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/public-benefits-bench/meta_muse_spark_1_1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Public Benefits Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "social mobility", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark_1_1", + "id": "meta/muse_spark_1_1", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark_1_1", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "public-benefits-bench:overall:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.public-benefits-bench.overall", + "source_data": { + "dataset_name": "Public Benefits Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Public Benefits Bench v1.1 (Overall).", + "metric_id": "vals_ai.public-benefits-bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "score_details": { + "score": 61.299, + "details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "social mobility", + "raw_score": "61.299", + "raw_stderr": "1.267", + "latency": "524.666", + "cost_per_test": "0.278428", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.267, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark_1_1/a51716d3-2041-4901-9c97-70d62c21443e.json b/data/vals_ai/meta/muse_spark_1_1/a51716d3-2041-4901-9c97-70d62c21443e.json new file mode 100644 index 0000000000000000000000000000000000000000..f6122d24b691b9e7eaa14de65180e9714351d68d --- /dev/null +++ b/data/vals_ai/meta/muse_spark_1_1/a51716d3-2041-4901-9c97-70d62c21443e.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/code-migration/meta_muse_spark_1_1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Code Migration", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark_1_1", + "id": "meta/muse_spark_1_1", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark_1_1", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "code-migration:cli:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.code-migration.cli", + "source_data": { + "dataset_name": "Code Migration - CLI", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cli", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (CLI).", + "metric_id": "vals_ai.code-migration.cli.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 18.629, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cli", + "task_name": "CLI", + "dataset_type": "private", + "industry": "coding", + "raw_score": "18.629", + "raw_stderr": "2.08", + "latency": "2739.514", + "cost_per_test": "5.536818", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 2.08, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "code-migration:cobol:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.code-migration.cobol", + "source_data": { + "dataset_name": "Code Migration - COBOL", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cobol", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (COBOL).", + "metric_id": "vals_ai.code-migration.cobol.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 68.571, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cobol", + "task_name": "COBOL", + "dataset_type": "private", + "industry": "coding", + "raw_score": "68.571", + "raw_stderr": "15.028", + "latency": "1186.716", + "cost_per_test": "1.630862", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 15.028, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "code-migration:code_quality:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.code-migration.code_quality", + "source_data": { + "dataset_name": "Code Migration - Code Quality", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "code_quality", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Code Quality).", + "metric_id": "vals_ai.code-migration.code_quality.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 2.805, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "code_quality", + "task_name": "Code Quality", + "dataset_type": "private", + "industry": "coding", + "raw_score": "2.805", + "raw_stderr": "0.085", + "latency": "2351.314", + "cost_per_test": "4.560329", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.085, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "code-migration:overall:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.code-migration.overall", + "source_data": { + "dataset_name": "Code Migration - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Overall).", + "metric_id": "vals_ai.code-migration.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 31.115, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "31.115", + "raw_stderr": "4.068", + "latency": "2351.314", + "cost_per_test": "4.560329", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 4.068, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark_1_1/b4c8aef5-4dc6-420e-b41e-ac192c3378d1.json b/data/vals_ai/meta/muse_spark_1_1/b4c8aef5-4dc6-420e-b41e-ac192c3378d1.json new file mode 100644 index 0000000000000000000000000000000000000000..a5cd0280f73eaf8b876c0ba25908efba4e0cb52f --- /dev/null +++ b/data/vals_ai/meta/muse_spark_1_1/b4c8aef5-4dc6-420e-b41e-ac192c3378d1.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/meta_muse_spark_1_1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark_1_1", + "id": "meta/muse_spark_1_1", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark_1_1", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 66.885, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.885", + "raw_stderr": "1.346", + "latency": "32.816", + "cost_per_test": "0.014921", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.346, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 79.722, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "79.722", + "raw_stderr": "0.77", + "latency": "32.816", + "cost_per_test": "0.014921", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.77, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 92.559, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "92.559", + "raw_stderr": "0.75", + "latency": "32.816", + "cost_per_test": "0.014921", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.75, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark_1_1/ccc9d412-4ab3-45dc-95da-6bb58b52079e.json b/data/vals_ai/meta/muse_spark_1_1/ccc9d412-4ab3-45dc-95da-6bb58b52079e.json new file mode 100644 index 0000000000000000000000000000000000000000..61f7d8b171efe981e986323498f972894e12a038 --- /dev/null +++ b/data/vals_ai/meta/muse_spark_1_1/ccc9d412-4ab3-45dc-95da-6bb58b52079e.json @@ -0,0 +1,103 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/meta_muse_spark_1_1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark_1_1", + "id": "meta/muse_spark_1_1", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark_1_1", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 86.59, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.59", + "raw_stderr": "0.819", + "latency": "31.715", + "cost_per_test": "0.01361", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.819, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark_1_1/d25eb8f7-0fb1-4a02-ae2f-12d0fb13ac1d.json b/data/vals_ai/meta/muse_spark_1_1/d25eb8f7-0fb1-4a02-ae2f-12d0fb13ac1d.json new file mode 100644 index 0000000000000000000000000000000000000000..40d7998386b1766f45d54508592175298fa72b73 --- /dev/null +++ b/data/vals_ai/meta/muse_spark_1_1/d25eb8f7-0fb1-4a02-ae2f-12d0fb13ac1d.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/meta_muse_spark_1_1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark_1_1", + "id": "meta/muse_spark_1_1", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark_1_1", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 75.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "75", + "raw_stderr": "0", + "latency": "13901.169", + "cost_per_test": "0.168085", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 52.222, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "52.222", + "raw_stderr": "2.94", + "latency": "1853.489", + "cost_per_test": "0.642939", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 2.94, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 78.182, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "78.182", + "raw_stderr": "1.05", + "latency": "1010.994", + "cost_per_test": "0.312504", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.05, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 69.288, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "69.288", + "raw_stderr": "0.991", + "latency": "624.772", + "cost_per_test": "0.417396", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.991, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark_1_1/da320a9d-466b-4eb9-a3dc-761a48397c9b.json b/data/vals_ai/meta/muse_spark_1_1/da320a9d-466b-4eb9-a3dc-761a48397c9b.json new file mode 100644 index 0000000000000000000000000000000000000000..89699af0b937a194d0b6be624a787c21d6f95c61 --- /dev/null +++ b/data/vals_ai/meta/muse_spark_1_1/da320a9d-466b-4eb9-a3dc-761a48397c9b.json @@ -0,0 +1,438 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/meta_muse_spark_1_1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark_1_1", + "id": "meta/muse_spark_1_1", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark_1_1", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.328, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.328", + "raw_stderr": "0.742", + "latency": "18.313", + "cost_per_test": "0.008371", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.742, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.876, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.876", + "raw_stderr": "0.199", + "latency": "14.792", + "cost_per_test": "0.00777", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.199, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 88.993, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.993", + "raw_stderr": "0.571", + "latency": "16.813", + "cost_per_test": "0.009923", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.571, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 84.981, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.981", + "raw_stderr": "0.425", + "latency": "17.148", + "cost_per_test": "0.008535", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.425, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.147, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.147", + "raw_stderr": "1.152", + "latency": "12.902", + "cost_per_test": "0.006965", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.152, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.562, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.562", + "raw_stderr": "1.509", + "latency": "22.919", + "cost_per_test": "0.009646", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.509, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark_1_1/eec8b525-460f-49be-b840-329cc8db3fb9.json b/data/vals_ai/meta/muse_spark_1_1/eec8b525-460f-49be-b840-329cc8db3fb9.json new file mode 100644 index 0000000000000000000000000000000000000000..e1f641d0127f30447d5fa72d44d33981bf2b8090 --- /dev/null +++ b/data/vals_ai/meta/muse_spark_1_1/eec8b525-460f-49be-b840-329cc8db3fb9.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/meta_muse_spark_1_1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark_1_1", + "id": "meta/muse_spark_1_1", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark_1_1", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 72.158, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "72.158", + "raw_stderr": "3.327", + "latency": "615.428", + "cost_per_test": "0.923283", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 3.327, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark_1_1/fa82a17a-348e-4ecc-85c1-95bec6883d96.json b/data/vals_ai/meta/muse_spark_1_1/fa82a17a-348e-4ecc-85c1-95bec6883d96.json new file mode 100644 index 0000000000000000000000000000000000000000..941fa004b76a946495e0ec4b72f695c7116430dd --- /dev/null +++ b/data/vals_ai/meta/muse_spark_1_1/fa82a17a-348e-4ecc-85c1-95bec6883d96.json @@ -0,0 +1,3386 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/hlab/meta_muse_spark_1_1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Harvey's Legal Agent Benchmark", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark_1_1", + "id": "meta/muse_spark_1_1", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark_1_1", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "hlab:antitrust_competition:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 40.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "40", + "raw_stderr": "24.495", + "latency": "708.483", + "cost_per_test": "0.782306", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 24.495, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:arbitration_international_dispute_resolution:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "774.512", + "cost_per_test": "0.973952", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:banking_finance:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "553.192", + "cost_per_test": "0.664525", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:bankruptcy_restructuring:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "838.671", + "cost_per_test": "0.878241", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:capital_markets:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "557.868", + "cost_per_test": "0.645246", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:corporate_governance:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "1321.071", + "cost_per_test": "1.030941", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:corporate_ma:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "778.885", + "cost_per_test": "0.847903", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Criteria Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Criteria Pass Rate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 92.859, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate", + "task_name": "Criteria Pass Rate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "92.859", + "raw_stderr": "0.324", + "latency": "740.851", + "cost_per_test": "0.800257", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.324, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__antitrust_competition:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 94.203, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "94.203", + "raw_stderr": "1.409", + "latency": "708.483", + "cost_per_test": "0.782306", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.409, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__arbitration_international_dispute_resolution:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 94.265, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "94.265", + "raw_stderr": "1.394", + "latency": "774.512", + "cost_per_test": "0.973952", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.394, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__banking_finance:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 94.796, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "94.796", + "raw_stderr": "1.357", + "latency": "553.192", + "cost_per_test": "0.664525", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.357, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__bankruptcy_restructuring:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 92.042, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "92.042", + "raw_stderr": "1.595", + "latency": "838.671", + "cost_per_test": "0.878241", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.595, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__capital_markets:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 92.549, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "92.549", + "raw_stderr": "1.648", + "latency": "557.868", + "cost_per_test": "0.645246", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.648, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_governance:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 96.918, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "96.918", + "raw_stderr": "1.013", + "latency": "1321.071", + "cost_per_test": "1.030941", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.013, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_ma:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 96.154, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "96.154", + "raw_stderr": "1.09", + "latency": "778.885", + "cost_per_test": "0.847903", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.09, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__data_privacy_cybersecurity:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 93.694, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "93.694", + "raw_stderr": "1.635", + "latency": "654.122", + "cost_per_test": "0.83785", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.635, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__emerging_companies_venture_capital:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 91.809, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "91.809", + "raw_stderr": "1.605", + "latency": "798.356", + "cost_per_test": "0.754855", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.605, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__employment_labor:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 87.266, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.266", + "raw_stderr": "2.044", + "latency": "452.276", + "cost_per_test": "0.589655", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 2.044, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__energy_natural_resources:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 96.37, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "96.37", + "raw_stderr": "1.076", + "latency": "1031.535", + "cost_per_test": "1.224794", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.076, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__environmental_esg:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 92.529, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "92.529", + "raw_stderr": "1.999", + "latency": "496.385", + "cost_per_test": "0.594901", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.999, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__funds_asset_management:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 92.128, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "92.128", + "raw_stderr": "1.456", + "latency": "1163.578", + "cost_per_test": "0.99674", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.456, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__healthcare_life_sciences:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 92.739, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "92.739", + "raw_stderr": "1.493", + "latency": "1001.377", + "cost_per_test": "0.991405", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.493, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__immigration:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 93.213, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "93.213", + "raw_stderr": "1.696", + "latency": "529.328", + "cost_per_test": "0.483627", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.696, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__insurance:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 87.931, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.931", + "raw_stderr": "3.038", + "latency": "491.636", + "cost_per_test": "0.58862", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 3.038, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__intellectual_property:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 95.406, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "95.406", + "raw_stderr": "1.247", + "latency": "561.361", + "cost_per_test": "0.701418", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.247, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__international_trade_sanctions:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 90.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90", + "raw_stderr": "4.286", + "latency": "372.412", + "cost_per_test": "0.355364", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 4.286, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__litigation_dispute_resolution:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 93.633, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "93.633", + "raw_stderr": "1.497", + "latency": "748.114", + "cost_per_test": "0.879933", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.497, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__real_estate:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 95.627, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "95.627", + "raw_stderr": "1.106", + "latency": "935.002", + "cost_per_test": "0.95827", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.106, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__structured_finance_securitization:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 87.302, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.302", + "raw_stderr": "1.879", + "latency": "565.425", + "cost_per_test": "0.630492", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.879, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__tax:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 91.304, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "91.304", + "raw_stderr": "1.632", + "latency": "698.262", + "cost_per_test": "0.714588", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.632, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__trusts_estates_private_client:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 91.603, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "91.603", + "raw_stderr": "1.717", + "latency": "613.884", + "cost_per_test": "0.719941", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.717, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__white_collar_defense_investigations:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 89.562, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.562", + "raw_stderr": "1.777", + "latency": "532.972", + "cost_per_test": "0.739764", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.777, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:data_privacy_cybersecurity:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "654.122", + "cost_per_test": "0.83785", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:emerging_companies_venture_capital:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "798.356", + "cost_per_test": "0.754855", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:employment_labor:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "452.276", + "cost_per_test": "0.589655", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:energy_natural_resources:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "1031.535", + "cost_per_test": "1.224794", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:environmental_esg:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "496.385", + "cost_per_test": "0.594901", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:funds_asset_management:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 40.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "40", + "raw_stderr": "24.495", + "latency": "1163.578", + "cost_per_test": "0.99674", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 24.495, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:healthcare_life_sciences:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "1001.377", + "cost_per_test": "0.991405", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:immigration:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 25.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "25", + "raw_stderr": "25", + "latency": "529.328", + "cost_per_test": "0.483627", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 25.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:insurance:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "491.636", + "cost_per_test": "0.58862", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:intellectual_property:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "561.361", + "cost_per_test": "0.701418", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:international_trade_sanctions:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "372.412", + "cost_per_test": "0.355364", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:litigation_dispute_resolution:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "748.114", + "cost_per_test": "0.879933", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:overall:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.overall", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Overall).", + "metric_id": "vals_ai.hlab.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "3.219", + "latency": "740.851", + "cost_per_test": "0.800257", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 3.219, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:real_estate:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "935.002", + "cost_per_test": "0.95827", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:structured_finance_securitization:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "565.425", + "cost_per_test": "0.630492", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:tax:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "698.262", + "cost_per_test": "0.714588", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:trusts_estates_private_client:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "613.884", + "cost_per_test": "0.719941", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:white_collar_defense_investigations:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.hlab.white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "532.972", + "cost_per_test": "0.739764", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark_1_1/fa8e07d6-330a-4f72-ae23-4f6fa5971faa.json b/data/vals_ai/meta/muse_spark_1_1/fa8e07d6-330a-4f72-ae23-4f6fa5971faa.json new file mode 100644 index 0000000000000000000000000000000000000000..249f4277b341ee13f49a0088ec37961a45b31bbc --- /dev/null +++ b/data/vals_ai/meta/muse_spark_1_1/fa8e07d6-330a-4f72-ae23-4f6fa5971faa.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/programbench/meta_muse_spark_1_1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProgramBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark_1_1", + "id": "meta/muse_spark_1_1", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark_1_1", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "programbench:almost:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.programbench.almost", + "source_data": { + "dataset_name": "ProgramBench - Almost Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "almost", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Almost Resolved).", + "metric_id": "vals_ai.programbench.almost.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 1.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "almost", + "task_name": "Almost Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "1", + "raw_stderr": "0.705", + "latency": "19626.22", + "cost_per_test": "9.158845", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.705, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "programbench:overall:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.programbench.overall", + "source_data": { + "dataset_name": "ProgramBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Overall).", + "metric_id": "vals_ai.programbench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "19626.22", + "cost_per_test": "9.158845", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "programbench:partial:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.programbench.partial", + "source_data": { + "dataset_name": "ProgramBench - Raw Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "partial", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Raw Pass Rate).", + "metric_id": "vals_ai.programbench.partial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 16.157, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "partial", + "task_name": "Raw Pass Rate", + "dataset_type": "public", + "industry": "coding", + "raw_score": "16.157", + "raw_stderr": "2.022", + "latency": "19626.22", + "cost_per_test": "9.158845", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 2.022, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "programbench:strict:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.programbench.strict", + "source_data": { + "dataset_name": "ProgramBench - Fully Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "strict", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Fully Resolved).", + "metric_id": "vals_ai.programbench.strict.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "strict", + "task_name": "Fully Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "19626.22", + "cost_per_test": "9.158845", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark_1_1/fb0f0da9-1d02-440f-b1de-9d70b2af908d.json b/data/vals_ai/meta/muse_spark_1_1/fb0f0da9-1d02-440f-b1de-9d70b2af908d.json new file mode 100644 index 0000000000000000000000000000000000000000..9c6747827057ede677d391aff9bae5db8b0fe586 --- /dev/null +++ b/data/vals_ai/meta/muse_spark_1_1/fb0f0da9-1d02-440f-b1de-9d70b2af908d.json @@ -0,0 +1,740 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/meta_muse_spark_1_1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark_1_1", + "id": "meta/muse_spark_1_1", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark_1_1", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 49.66, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "49.66", + "raw_stderr": "2.337", + "latency": "331.405", + "cost_per_test": "0.707474", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 2.337, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 44.919, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "44.919", + "raw_stderr": "1.004", + "latency": "414.77", + "cost_per_test": "0.723391", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.004, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 47.162, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "47.162", + "raw_stderr": "0.42", + "latency": "515.058", + "cost_per_test": "0.973188", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.42, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 63.449, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.449", + "raw_stderr": "1.377", + "latency": "569.177", + "cost_per_test": "0.877309", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.377, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 76.197, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "76.197", + "raw_stderr": "0.67", + "latency": "240.689", + "cost_per_test": "0.282783", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.67, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 34.139, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "34.139", + "raw_stderr": "2.315", + "latency": "380.427", + "cost_per_test": "0.668285", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 2.315, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 66.365, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.365", + "raw_stderr": "1.941", + "latency": "409.17", + "cost_per_test": "0.746224", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.941, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 76.845, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "76.845", + "raw_stderr": "0.97", + "latency": "418.705", + "cost_per_test": "0.78633", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.97, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 67.233, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.233", + "raw_stderr": "1.314", + "latency": "217.075", + "cost_per_test": "0.458463", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.314, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:overall:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 57.207, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.207", + "raw_stderr": "0.789", + "latency": "414.77", + "cost_per_test": "0.723391", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 0.789, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 33.81, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "33.81", + "raw_stderr": "2.472", + "latency": "674.897", + "cost_per_test": "1.036058", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 2.472, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/meta/muse_spark_1_1/fbc6bc11-416b-4bb7-ba4e-dd0cdca51069.json b/data/vals_ai/meta/muse_spark_1_1/fbc6bc11-416b-4bb7-ba4e-dd0cdca51069.json new file mode 100644 index 0000000000000000000000000000000000000000..a11fc13b60c8914dce04f5fc29f5824b2db9f053 --- /dev/null +++ b/data/vals_ai/meta/muse_spark_1_1/fbc6bc11-416b-4bb7-ba4e-dd0cdca51069.json @@ -0,0 +1,1188 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_research/meta_muse_spark_1_1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Legal Research Bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "muse_spark_1_1", + "id": "meta/muse_spark_1_1", + "developer": "meta", + "additional_details": { + "vals_model_id": "meta/muse_spark_1_1", + "vals_provider": "Meta" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_research:administrative_regulatory:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.legal_research.administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 37.879, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "37.879", + "raw_stderr": "6.017", + "latency": "282.392", + "cost_per_test": "0.382532", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 6.017, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:business_commercial:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.legal_research.business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 51.163, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "51.163", + "raw_stderr": "7.713", + "latency": "319.151", + "cost_per_test": "0.363621", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 7.713, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:civil_litigation:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.legal_research.civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 48.148, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "48.148", + "raw_stderr": "9.799", + "latency": "308.847", + "cost_per_test": "0.367043", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 9.799, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:constitutional_civil_rights:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.legal_research.constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 28.947, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "28.947", + "raw_stderr": "7.456", + "latency": "310.439", + "cost_per_test": "0.303515", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 7.456, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:criminal:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.legal_research.criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 36.842, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "36.842", + "raw_stderr": "6.446", + "latency": "320.962", + "cost_per_test": "0.345173", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 6.446, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:family:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.legal_research.family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 22.727, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "22.727", + "raw_stderr": "9.145", + "latency": "348.059", + "cost_per_test": "0.406411", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 9.145, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:health:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.legal_research.health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 40.0, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "40", + "raw_stderr": "9.097", + "latency": "298.421", + "cost_per_test": "0.506452", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 9.097, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:immigration:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.legal_research.immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 9.091, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "9.091", + "raw_stderr": "9.091", + "latency": "305.943", + "cost_per_test": "0.448482", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 9.091, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:overall:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.legal_research.overall", + "source_data": { + "dataset_name": "Legal Research Bench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Overall).", + "metric_id": "vals_ai.legal_research.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 37.981, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "legal", + "raw_score": "37.981", + "raw_stderr": "3.373", + "latency": "312.994", + "cost_per_test": "0.377379", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 3.373, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass", + "source_data": { + "dataset_name": "Legal Research Bench - Weighted Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Weighted Pass).", + "metric_id": "vals_ai.legal_research.weighted_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 84.621, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass", + "task_name": "Weighted Pass", + "dataset_type": "private", + "industry": "legal", + "raw_score": "84.621", + "raw_stderr": "1.419", + "latency": "312.994", + "cost_per_test": "0.377379", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 1.419, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__administrative_regulatory:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.weighted_pass__administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 83.623, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "83.623", + "raw_stderr": "2.846", + "latency": "282.392", + "cost_per_test": "0.382532", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 2.846, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__business_commercial:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.weighted_pass__business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 89.091, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "89.091", + "raw_stderr": "2.687", + "latency": "319.151", + "cost_per_test": "0.363621", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 2.687, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__civil_litigation:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.weighted_pass__civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 88.953, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "88.953", + "raw_stderr": "3.252", + "latency": "308.847", + "cost_per_test": "0.367043", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 3.252, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__constitutional_civil_rights:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 85.619, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "85.619", + "raw_stderr": "2.661", + "latency": "310.439", + "cost_per_test": "0.303515", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 2.661, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__criminal:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.weighted_pass__criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 85.445, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "85.445", + "raw_stderr": "2.33", + "latency": "320.962", + "cost_per_test": "0.345173", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 2.33, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__family:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.weighted_pass__family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 73.686, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "73.686", + "raw_stderr": "4.833", + "latency": "348.059", + "cost_per_test": "0.406411", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 4.833, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__health:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.weighted_pass__health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 87.936, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "87.936", + "raw_stderr": "2.581", + "latency": "298.421", + "cost_per_test": "0.506452", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 2.581, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__immigration:meta/muse_spark_1_1:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.weighted_pass__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 82.55, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "82.55", + "raw_stderr": "6.309", + "latency": "305.943", + "cost_per_test": "0.448482", + "max_output_tokens": "256000", + "reasoning_effort": "xhigh", + "provider": "Meta" + }, + "uncertainty": { + "standard_error": { + "value": 6.309, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.1/28cd72b5-dbe4-45bb-bbe4-8bc637faa83d.json b/data/vals_ai/minimax/MiniMax-M2.1/28cd72b5-dbe4-45bb-bbe4-8bc637faa83d.json new file mode 100644 index 0000000000000000000000000000000000000000..30836ed6fc3cbcc0f570960ffe4b61ada5110cd3 --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.1/28cd72b5-dbe4-45bb-bbe4-8bc637faa83d.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/minimax_MiniMax-M2.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.1", + "id": "minimax/MiniMax-M2.1", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.1", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 75.758, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.758", + "raw_stderr": "3.088", + "latency": "110.423", + "cost_per_test": "0.009533", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 3.088, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 81.313, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.313", + "raw_stderr": "2.93", + "latency": "160.401", + "cost_per_test": "0.013312", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.93, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 78.536, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.536", + "raw_stderr": "2.128", + "latency": "135.412", + "cost_per_test": "0.011422", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.128, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.1/2b9b75fc-822f-44a9-b814-2aa1660fc381.json b/data/vals_ai/minimax/MiniMax-M2.1/2b9b75fc-822f-44a9-b814-2aa1660fc381.json new file mode 100644 index 0000000000000000000000000000000000000000..ae625c1e17ad7bc006541e86d9061d84be6f7eed --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.1/2b9b75fc-822f-44a9-b814-2aa1660fc381.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/minimax_MiniMax-M2.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.1", + "id": "minimax/MiniMax-M2.1", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.1", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.584, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.584", + "raw_stderr": "0.924", + "latency": "27.098", + "cost_per_test": "0.0022", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.924, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.114, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.114", + "raw_stderr": "1.069", + "latency": "38.449", + "cost_per_test": "0.003023", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.069, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.343, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.343", + "raw_stderr": "0.836", + "latency": "54.018", + "cost_per_test": "0.004516", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.836, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.073, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.073", + "raw_stderr": "1.67", + "latency": "38.487", + "cost_per_test": "0.003263", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.67, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.284, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.284", + "raw_stderr": "1.019", + "latency": "21.514", + "cost_per_test": "0.001944", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.019, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.837, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.837", + "raw_stderr": "1.239", + "latency": "105.718", + "cost_per_test": "0.008321", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.239, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.741, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.741", + "raw_stderr": "1.29", + "latency": "25.643", + "cost_per_test": "0.001883", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.29, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.528, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.528", + "raw_stderr": "2.067", + "latency": "38.904", + "cost_per_test": "0.003043", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.067, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.928, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.928", + "raw_stderr": "1.229", + "latency": "43.414", + "cost_per_test": "0.003182", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.229, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.746, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.746", + "raw_stderr": "0.706", + "latency": "55.213", + "cost_per_test": "0.004308", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.706, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.039, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.039", + "raw_stderr": "1.14", + "latency": "26.158", + "cost_per_test": "0.001852", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.14, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.049, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.049", + "raw_stderr": "0.333", + "latency": "40.957", + "cost_per_test": "0.003225", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.972, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.972", + "raw_stderr": "1.555", + "latency": "23.972", + "cost_per_test": "0.001774", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.555, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.53, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.53", + "raw_stderr": "0.849", + "latency": "52.388", + "cost_per_test": "0.003995", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.849, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.97, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.97", + "raw_stderr": "1.157", + "latency": "22.426", + "cost_per_test": "0.001849", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.157, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.1/30f6d4b5-8e39-4068-8ece-144a3ba6eecf.json b/data/vals_ai/minimax/MiniMax-M2.1/30f6d4b5-8e39-4068-8ece-144a3ba6eecf.json new file mode 100644 index 0000000000000000000000000000000000000000..291abcbe0f969efdaf2cf7eb3d507b7eb4f8285a --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.1/30f6d4b5-8e39-4068-8ece-144a3ba6eecf.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/minimax_MiniMax-M2.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.1", + "id": "minimax/MiniMax-M2.1", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.1", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 80.777, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "80.777", + "raw_stderr": "1.831", + "latency": "53.161", + "cost_per_test": "0.005087", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.831, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.1/44290625-3122-4735-9f73-def55e6c5681.json b/data/vals_ai/minimax/MiniMax-M2.1/44290625-3122-4735-9f73-def55e6c5681.json new file mode 100644 index 0000000000000000000000000000000000000000..ab2349937b711f25eb7bb80dfbe3be39c01410f5 --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.1/44290625-3122-4735-9f73-def55e6c5681.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/minimax_MiniMax-M2.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.1", + "id": "minimax/MiniMax-M2.1", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.1", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 96.273, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "96.273", + "raw_stderr": "1.056", + "latency": "36.87", + "cost_per_test": "0.002892", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.056, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 62.571, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "62.571", + "raw_stderr": "2.587", + "latency": "529.6", + "cost_per_test": "0.038707", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.587, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 86.423, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "86.423", + "raw_stderr": "1.75", + "latency": "171.747", + "cost_per_test": "0.013857", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.75, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 81.756, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "81.756", + "raw_stderr": "1.099", + "latency": "246.072", + "cost_per_test": "0.018485", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.099, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.1/8935e12a-9e26-4dd2-abf0-3e27509f95b4.json b/data/vals_ai/minimax/MiniMax-M2.1/8935e12a-9e26-4dd2-abf0-3e27509f95b4.json new file mode 100644 index 0000000000000000000000000000000000000000..21ee9ca37317e226f336d5f4446d6743c15233e3 --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.1/8935e12a-9e26-4dd2-abf0-3e27509f95b4.json @@ -0,0 +1,438 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/minimax_MiniMax-M2.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.1", + "id": "minimax/MiniMax-M2.1", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.1", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 85.318, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.318", + "raw_stderr": "0.804", + "latency": "9.637", + "cost_per_test": "0.000695", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.804, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 73.592, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "73.592", + "raw_stderr": "0.263", + "latency": "9.034", + "cost_per_test": "0.000753", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.263, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 71.973, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "71.973", + "raw_stderr": "0.807", + "latency": "10.116", + "cost_per_test": "0.001045", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.807, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 75.448, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "75.448", + "raw_stderr": "0.539", + "latency": "11.491", + "cost_per_test": "0.000937", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.539, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 71.806, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "71.806", + "raw_stderr": "1.597", + "latency": "7.218", + "cost_per_test": "0.000605", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.597, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 74.551, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "74.551", + "raw_stderr": "1.826", + "latency": "21.452", + "cost_per_test": "0.001587", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.826, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.1/8e48f343-d959-4df6-99a5-043adb645fa0.json b/data/vals_ai/minimax/MiniMax-M2.1/8e48f343-d959-4df6-99a5-043adb645fa0.json new file mode 100644 index 0000000000000000000000000000000000000000..f087f94ed45d03028da44239499b26c43680566d --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.1/8e48f343-d959-4df6-99a5-043adb645fa0.json @@ -0,0 +1,371 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/minimax_MiniMax-M2.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.1", + "id": "minimax/MiniMax-M2.1", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.1", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 40.476, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "40.476", + "raw_stderr": "7.666", + "latency": "551.344", + "cost_per_test": "0.73469", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 7.666, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 72.797, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "72.797", + "raw_stderr": "2.76", + "latency": "430.408", + "cost_per_test": "0.508754", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.76, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 85.567, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "85.567", + "raw_stderr": "2.53", + "latency": "287.029", + "cost_per_test": "0.275394", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.53, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "650.542", + "cost_per_test": "0.817428", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 74.8, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "74.8", + "raw_stderr": "1.944", + "latency": "386.256", + "cost_per_test": "0.439041", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.944, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.1/957e221e-fd42-4cbb-a39d-bd1cc04b2789.json b/data/vals_ai/minimax/MiniMax-M2.1/957e221e-fd42-4cbb-a39d-bd1cc04b2789.json new file mode 100644 index 0000000000000000000000000000000000000000..54ee80c0dd44b0765b02bb5ae49875edbc105190 --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.1/957e221e-fd42-4cbb-a39d-bd1cc04b2789.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/minimax_MiniMax-M2.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.1", + "id": "minimax/MiniMax-M2.1", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.1", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.305, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.305", + "raw_stderr": "1.663", + "latency": "15.39", + "cost_per_test": "0.001629", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.663, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 54.312, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "54.312", + "raw_stderr": "1.701", + "latency": "28.337", + "cost_per_test": "0.041668", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.701, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 58.896, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.896", + "raw_stderr": "0.968", + "latency": "21.969", + "cost_per_test": "0.022461", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.968, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.072, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.072", + "raw_stderr": "1.665", + "latency": "22.18", + "cost_per_test": "0.024085", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.665, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.1/f34bf8e0-40ff-44eb-a10f-1498f1a8aefe.json b/data/vals_ai/minimax/MiniMax-M2.1/f34bf8e0-40ff-44eb-a10f-1498f1a8aefe.json new file mode 100644 index 0000000000000000000000000000000000000000..c3beb2bd6dfb8c85006fa978c74b50b1e4fac958 --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.1/f34bf8e0-40ff-44eb-a10f-1498f1a8aefe.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/minimax_MiniMax-M2.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.1", + "id": "minimax/MiniMax-M2.1", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.1", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 54.456, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "54.456", + "raw_stderr": "1.424", + "latency": "113.258", + "cost_per_test": "0.009137", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.424, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 66.353, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.353", + "raw_stderr": "0.925", + "latency": "113.258", + "cost_per_test": "0.009137", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.925, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 78.25, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "78.25", + "raw_stderr": "1.18", + "latency": "113.258", + "cost_per_test": "0.009137", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.18, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.1/f73f138f-6bf1-4f7c-984f-8e9b5b1be326.json b/data/vals_ai/minimax/MiniMax-M2.1/f73f138f-6bf1-4f7c-984f-8e9b5b1be326.json new file mode 100644 index 0000000000000000000000000000000000000000..d635ac7ff5dabb5a23aecd57fd66c3d584ba8a3e --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.1/f73f138f-6bf1-4f7c-984f-8e9b5b1be326.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/minimax_MiniMax-M2.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.1", + "id": "minimax/MiniMax-M2.1", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.1", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "4937.493", + "cost_per_test": "0.233531", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 4.667, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "4.667", + "raw_stderr": "2.6", + "latency": "5483.31", + "cost_per_test": "0.405159", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.6, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:minimax/MiniMax-M2.1:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 2.334, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "2.334", + "raw_stderr": "1.3", + "latency": "5210.402", + "cost_per_test": "0.319345", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.3, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.5/14daf1cb-b5b8-4fed-b462-41e5c5d0dc9d.json b/data/vals_ai/minimax/MiniMax-M2.5/14daf1cb-b5b8-4fed-b462-41e5c5d0dc9d.json new file mode 100644 index 0000000000000000000000000000000000000000..36205d169bafb50802433734fa63e721ccbd8238 --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.5/14daf1cb-b5b8-4fed-b462-41e5c5d0dc9d.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/minimax_MiniMax-M2.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.5", + "id": "minimax/MiniMax-M2.5", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.5", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.758, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.758", + "raw_stderr": "0.617", + "latency": "44.105", + "cost_per_test": "0.004989", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.617, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 57.143, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "57.143", + "raw_stderr": "2.645", + "latency": "642.154", + "cost_per_test": "0.068525", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.645, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 81.723, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "81.723", + "raw_stderr": "1.975", + "latency": "249.775", + "cost_per_test": "0.025642", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.975, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 79.208, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "79.208", + "raw_stderr": "1.119", + "latency": "312.011", + "cost_per_test": "0.033052", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.119, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.5/2d749d71-c3ce-46d9-a0a2-ccfbe58e5d05.json b/data/vals_ai/minimax/MiniMax-M2.5/2d749d71-c3ce-46d9-a0a2-ccfbe58e5d05.json new file mode 100644 index 0000000000000000000000000000000000000000..e69050171e501a6bef654b3753bf2b68bea9f4aa --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.5/2d749d71-c3ce-46d9-a0a2-ccfbe58e5d05.json @@ -0,0 +1,438 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/minimax_MiniMax-M2.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.5", + "id": "minimax/MiniMax-M2.5", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.5", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 86.241, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.241", + "raw_stderr": "0.903", + "latency": "13.522", + "cost_per_test": "0.001249", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.903, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.02, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.02", + "raw_stderr": "0.204", + "latency": "13.91", + "cost_per_test": "0.001159", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.204, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 84.169, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.169", + "raw_stderr": "0.662", + "latency": "14.901", + "cost_per_test": "0.001305", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.662, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.963, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.963", + "raw_stderr": "0.473", + "latency": "15.819", + "cost_per_test": "0.001287", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.473, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.835, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.835", + "raw_stderr": "1.025", + "latency": "8.925", + "cost_per_test": "0.000914", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.025, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 72.55, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "72.55", + "raw_stderr": "1.804", + "latency": "27.838", + "cost_per_test": "0.001806", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.804, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.5/42bf65c1-fdef-4386-b41a-eb0798ed5032.json b/data/vals_ai/minimax/MiniMax-M2.5/42bf65c1-fdef-4386-b41a-eb0798ed5032.json new file mode 100644 index 0000000000000000000000000000000000000000..87e2b6c29f8f011404ca39e2ba2ce2af5e31374b --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.5/42bf65c1-fdef-4386-b41a-eb0798ed5032.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/minimax_MiniMax-M2.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.5", + "id": "minimax/MiniMax-M2.5", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.5", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 4.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "4", + "raw_stderr": "1.96", + "latency": "4514.143", + "cost_per_test": "0.057497", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.96, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.5/4b56dde8-94f2-4491-bbba-5fdf10fa2a8e.json b/data/vals_ai/minimax/MiniMax-M2.5/4b56dde8-94f2-4491-bbba-5fdf10fa2a8e.json new file mode 100644 index 0000000000000000000000000000000000000000..4f41c070d3b7465f7aa4dbfaf10ad5ab74a8eb96 --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.5/4b56dde8-94f2-4491-bbba-5fdf10fa2a8e.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/minimax_MiniMax-M2.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.5", + "id": "minimax/MiniMax-M2.5", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.5", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 84.343, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.343", + "raw_stderr": "3.375", + "latency": "163.291", + "cost_per_test": "0.017314", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 3.375, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 79.798, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.798", + "raw_stderr": "3.001", + "latency": "180.634", + "cost_per_test": "0.018424", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 3.001, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 82.07, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.07", + "raw_stderr": "2.258", + "latency": "171.962", + "cost_per_test": "0.017869", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.258, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.5/6498661f-6d7c-400b-af21-33b901d820ce.json b/data/vals_ai/minimax/MiniMax-M2.5/6498661f-6d7c-400b-af21-33b901d820ce.json new file mode 100644 index 0000000000000000000000000000000000000000..00cb37f9a2eeaaf3b567cdd0bbb1444c454518ed --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.5/6498661f-6d7c-400b-af21-33b901d820ce.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/minimax_MiniMax-M2.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.5", + "id": "minimax/MiniMax-M2.5", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.5", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.655, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.655", + "raw_stderr": "1.66", + "latency": "10.304", + "cost_per_test": "0.002217", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.66, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 56.76, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.76", + "raw_stderr": "1.691", + "latency": "21.903", + "cost_per_test": "0.045262", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.691, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 59.596, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.596", + "raw_stderr": "0.966", + "latency": "16.599", + "cost_per_test": "0.024284", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.966, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.373, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.373", + "raw_stderr": "1.67", + "latency": "17.59", + "cost_per_test": "0.025373", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.67, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.5/86e510e9-a5c2-4554-815d-657a007e9524.json b/data/vals_ai/minimax/MiniMax-M2.5/86e510e9-a5c2-4554-815d-657a007e9524.json new file mode 100644 index 0000000000000000000000000000000000000000..ff9097031ff00d0a27206f5bb2985bd593f1b967 --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.5/86e510e9-a5c2-4554-815d-657a007e9524.json @@ -0,0 +1,371 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/minimax_MiniMax-M2.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.5", + "id": "minimax/MiniMax-M2.5", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.5", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 38.095, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "38.095", + "raw_stderr": "7.584", + "latency": "620.254", + "cost_per_test": "0.877132", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 7.584, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 70.881, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "70.881", + "raw_stderr": "2.818", + "latency": "433.408", + "cost_per_test": "0.49433", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.818, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 87.113, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "87.113", + "raw_stderr": "2.412", + "latency": "312.652", + "cost_per_test": "0.312736", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.412, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "548.617", + "cost_per_test": "0.644463", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 74.2, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "74.2", + "raw_stderr": "1.959", + "latency": "402.941", + "cost_per_test": "0.456928", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.959, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.5/c0629fc8-b871-4950-8a33-812843a946f9.json b/data/vals_ai/minimax/MiniMax-M2.5/c0629fc8-b871-4950-8a33-812843a946f9.json new file mode 100644 index 0000000000000000000000000000000000000000..895f02f36378e3eade96848579f08f511070fb59 --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.5/c0629fc8-b871-4950-8a33-812843a946f9.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/minimax_MiniMax-M2.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.5", + "id": "minimax/MiniMax-M2.5", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.5", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.703, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.703", + "raw_stderr": "1.189", + "latency": "28.855", + "cost_per_test": "0.003528", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.189, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.565, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.565", + "raw_stderr": "1.219", + "latency": "45.357", + "cost_per_test": "0.005087", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.219, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.721, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.721", + "raw_stderr": "0.978", + "latency": "60.449", + "cost_per_test": "0.006569", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.978, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.854, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.854", + "raw_stderr": "1.721", + "latency": "40.754", + "cost_per_test": "0.005007", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.721, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.308, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.308", + "raw_stderr": "1.223", + "latency": "29.01", + "cost_per_test": "0.003273", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.223, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.439, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.439", + "raw_stderr": "1.392", + "latency": "142.331", + "cost_per_test": "0.015702", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.392, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.506, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.506", + "raw_stderr": "1.46", + "latency": "36.157", + "cost_per_test": "0.003324", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.46, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 67.454, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "67.454", + "raw_stderr": "2.4", + "latency": "34.015", + "cost_per_test": "0.004019", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.4, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 59.219, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "59.219", + "raw_stderr": "1.482", + "latency": "66.126", + "cost_per_test": "0.006817", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.482, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.598, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.598", + "raw_stderr": "0.719", + "latency": "42.988", + "cost_per_test": "0.005288", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.719, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.788, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.788", + "raw_stderr": "1.345", + "latency": "30.649", + "cost_per_test": "0.003394", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.345, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.093, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.093", + "raw_stderr": "0.39", + "latency": "48.271", + "cost_per_test": "0.005362", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.39, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 69.539, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.539", + "raw_stderr": "2.064", + "latency": "38.873", + "cost_per_test": "0.00379", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.064, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.529, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.529", + "raw_stderr": "0.922", + "latency": "55.012", + "cost_per_test": "0.006352", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.922, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.073, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.073", + "raw_stderr": "1.44", + "latency": "25.216", + "cost_per_test": "0.002921", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.44, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.5/e26bc9de-fe8f-445f-a529-3f58b029ab3d.json b/data/vals_ai/minimax/MiniMax-M2.5/e26bc9de-fe8f-445f-a529-3f58b029ab3d.json new file mode 100644 index 0000000000000000000000000000000000000000..5444c4f9be0ca0c3a426f54f5465158e11405699 --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.5/e26bc9de-fe8f-445f-a529-3f58b029ab3d.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/minimax_MiniMax-M2.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.5", + "id": "minimax/MiniMax-M2.5", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.5", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 55.437, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "55.437", + "raw_stderr": "1.421", + "latency": "73.243", + "cost_per_test": "0.008629", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.421, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 68.152, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.152", + "raw_stderr": "0.906", + "latency": "73.243", + "cost_per_test": "0.008629", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.906, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 80.867, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "80.867", + "raw_stderr": "1.125", + "latency": "73.243", + "cost_per_test": "0.008629", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.125, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.5/f13169ea-3926-403e-b7e5-050a09906d7b.json b/data/vals_ai/minimax/MiniMax-M2.5/f13169ea-3926-403e-b7e5-050a09906d7b.json new file mode 100644 index 0000000000000000000000000000000000000000..d3a578ad62cecc08a7bdbabcf180c1ae1be7281a --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.5/f13169ea-3926-403e-b7e5-050a09906d7b.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/minimax_MiniMax-M2.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.5", + "id": "minimax/MiniMax-M2.5", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.5", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 7.5, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "7.5", + "raw_stderr": "4.458", + "latency": "17357.397", + "cost_per_test": "1.508215", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 4.458, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 5.833, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "5.833", + "raw_stderr": "2.833", + "latency": "28418.598", + "cost_per_test": "1.494618", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.833, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 6.666, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "6.666", + "raw_stderr": "2.641", + "latency": "22887.998", + "cost_per_test": "1.501416", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.641, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.5/f7839d64-8616-4820-ae02-f616fe3b4ac4.json b/data/vals_ai/minimax/MiniMax-M2.5/f7839d64-8616-4820-ae02-f616fe3b4ac4.json new file mode 100644 index 0000000000000000000000000000000000000000..355a5212a0d8c70b3f8ab9d83d9648ced546c6c9 --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.5/f7839d64-8616-4820-ae02-f616fe3b4ac4.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/minimax_MiniMax-M2.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.5", + "id": "minimax/MiniMax-M2.5", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.5", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:minimax/MiniMax-M2.5:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 14.853, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "14.853", + "raw_stderr": "2.953", + "latency": "3065.622", + "cost_per_test": "2.197729", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.953, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.7/279c1345-e8da-40f2-9334-cbae592b89df.json b/data/vals_ai/minimax/MiniMax-M2.7/279c1345-e8da-40f2-9334-cbae592b89df.json new file mode 100644 index 0000000000000000000000000000000000000000..66e227f7eae5f78aa4311ae1ef1f5f347bd27841 --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.7/279c1345-e8da-40f2-9334-cbae592b89df.json @@ -0,0 +1,740 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/minimax_MiniMax-M2.7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.7", + "id": "minimax/MiniMax-M2.7", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.7", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 20.972, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "20.972", + "raw_stderr": "3.111", + "latency": "931.225", + "cost_per_test": "0.213775", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 3.111, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 17.135, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "17.135", + "raw_stderr": "0.187", + "latency": "803.528", + "cost_per_test": "0.18348", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.187, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 28.371, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "28.371", + "raw_stderr": "0.702", + "latency": "994.669", + "cost_per_test": "0.220717", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.702, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 34.96, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "34.96", + "raw_stderr": "1.681", + "latency": "680.199", + "cost_per_test": "0.189858", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.681, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 31.848, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "31.848", + "raw_stderr": "1.171", + "latency": "499.749", + "cost_per_test": "0.097748", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.171, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 4.073, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "4.073", + "raw_stderr": "0.778", + "latency": "1482.909", + "cost_per_test": "0.308759", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.778, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 42.376, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "42.376", + "raw_stderr": "0.992", + "latency": "438.133", + "cost_per_test": "0.102708", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.992, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 43.049, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "43.049", + "raw_stderr": "1.88", + "latency": "669.802", + "cost_per_test": "0.187197", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.88, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 33.392, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "33.392", + "raw_stderr": "1.419", + "latency": "404.99", + "cost_per_test": "0.119445", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.419, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:overall:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 27.887, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "27.887", + "raw_stderr": "0.508", + "latency": "803.528", + "cost_per_test": "0.18348", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.508, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 11.943, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "11.943", + "raw_stderr": "0.943", + "latency": "1168.843", + "cost_per_test": "0.209794", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.943, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.7/368ddac2-649d-4b3d-b49c-e2af2051d5bf.json b/data/vals_ai/minimax/MiniMax-M2.7/368ddac2-649d-4b3d-b49c-e2af2051d5bf.json new file mode 100644 index 0000000000000000000000000000000000000000..cf50a279296e46d3f1b37374e1cdefad737ee76a --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.7/368ddac2-649d-4b3d-b49c-e2af2051d5bf.json @@ -0,0 +1,408 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/minimax_MiniMax-M2.7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.7", + "id": "minimax/MiniMax-M2.7", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.7", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 63.753, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "63.753", + "raw_stderr": "1.641", + "latency": "11.623", + "cost_per_test": "0.023635", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.641, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 27.943, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "27.943", + "raw_stderr": "1.058", + "latency": "750.712", + "cost_per_test": "0.169353", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.058, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:overall:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 41.56, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "41.56", + "raw_stderr": "1.178", + "latency": "782.769", + "cost_per_test": "0.720814", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.178, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 69.608, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "69.608", + "raw_stderr": "4.554", + "latency": "961.544", + "cost_per_test": "0.525961" + }, + "uncertainty": { + "standard_error": { + "value": 4.554, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 48.689, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "48.689", + "raw_stderr": "3.066", + "latency": "812.638", + "cost_per_test": "0.066012", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 3.066, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 11.719, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "11.719", + "raw_stderr": "4.172", + "latency": "1377.326", + "cost_per_test": "2.819109", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 4.172, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.7/7b1786da-8a55-4daf-a585-99a1aacce87b.json b/data/vals_ai/minimax/MiniMax-M2.7/7b1786da-8a55-4daf-a585-99a1aacce87b.json new file mode 100644 index 0000000000000000000000000000000000000000..e1a1e712e73ecc10e8733836ffc62db93add7a58 --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.7/7b1786da-8a55-4daf-a585-99a1aacce87b.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/programbench/minimax_MiniMax-M2.7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProgramBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.7", + "id": "minimax/MiniMax-M2.7", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.7", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "programbench:almost:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.programbench.almost", + "source_data": { + "dataset_name": "ProgramBench - Almost Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "almost", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Almost Resolved).", + "metric_id": "vals_ai.programbench.almost.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "almost", + "task_name": "Almost Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1460.176", + "cost_per_test": "0.428451", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "programbench:overall:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.programbench.overall", + "source_data": { + "dataset_name": "ProgramBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Overall).", + "metric_id": "vals_ai.programbench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1460.176", + "cost_per_test": "0.428451", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "programbench:partial:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.programbench.partial", + "source_data": { + "dataset_name": "ProgramBench - Raw Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "partial", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Raw Pass Rate).", + "metric_id": "vals_ai.programbench.partial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 13.536, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "partial", + "task_name": "Raw Pass Rate", + "dataset_type": "public", + "industry": "coding", + "raw_score": "13.536", + "raw_stderr": "1.394", + "latency": "1460.176", + "cost_per_test": "0.428451", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.394, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "programbench:strict:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.programbench.strict", + "source_data": { + "dataset_name": "ProgramBench - Fully Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "strict", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Fully Resolved).", + "metric_id": "vals_ai.programbench.strict.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "strict", + "task_name": "Fully Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1460.176", + "cost_per_test": "0.428451", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.7/81a6702d-16a3-4418-8254-46272c6752f8.json b/data/vals_ai/minimax/MiniMax-M2.7/81a6702d-16a3-4418-8254-46272c6752f8.json new file mode 100644 index 0000000000000000000000000000000000000000..2af334225009ba1f61c183048122750bfdf169c6 --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.7/81a6702d-16a3-4418-8254-46272c6752f8.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/code-migration/minimax_MiniMax-M2.7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Code Migration", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.7", + "id": "minimax/MiniMax-M2.7", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.7", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "code-migration:cli:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.code-migration.cli", + "source_data": { + "dataset_name": "Code Migration - CLI", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cli", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (CLI).", + "metric_id": "vals_ai.code-migration.cli.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 10.152, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cli", + "task_name": "CLI", + "dataset_type": "private", + "industry": "coding", + "raw_score": "10.152", + "raw_stderr": "1.905", + "latency": "1377.278", + "cost_per_test": "0.646598", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.905, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "code-migration:cobol:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.code-migration.cobol", + "source_data": { + "dataset_name": "Code Migration - COBOL", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cobol", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (COBOL).", + "metric_id": "vals_ai.code-migration.cobol.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 4.286, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cobol", + "task_name": "COBOL", + "dataset_type": "private", + "industry": "coding", + "raw_score": "4.286", + "raw_stderr": "4.286", + "latency": "905.439", + "cost_per_test": "0.410702", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 4.286, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "code-migration:code_quality:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.code-migration.code_quality", + "source_data": { + "dataset_name": "Code Migration - Code Quality", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "code_quality", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Code Quality).", + "metric_id": "vals_ai.code-migration.code_quality.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 2.966, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "code_quality", + "task_name": "Code Quality", + "dataset_type": "private", + "industry": "coding", + "raw_score": "2.966", + "raw_stderr": "0.037", + "latency": "1259.318", + "cost_per_test": "0.587624", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.037, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "code-migration:overall:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.code-migration.overall", + "source_data": { + "dataset_name": "Code Migration - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Overall).", + "metric_id": "vals_ai.code-migration.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 8.685, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "8.685", + "raw_stderr": "1.786", + "latency": "1259.318", + "cost_per_test": "0.587624", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.786, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.7/85c70d9a-faac-49d0-bcfc-f0dfb8bbca56.json b/data/vals_ai/minimax/MiniMax-M2.7/85c70d9a-faac-49d0-bcfc-f0dfb8bbca56.json new file mode 100644 index 0000000000000000000000000000000000000000..6d134d4daca2b83cc0cf0cf077f3c699bf13101b --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.7/85c70d9a-faac-49d0-bcfc-f0dfb8bbca56.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/minimax_MiniMax-M2.7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.7", + "id": "minimax/MiniMax-M2.7", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.7", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 75.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "75", + "raw_stderr": "0", + "latency": "18081.196", + "cost_per_test": "1.468763", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 18.889, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "18.889", + "raw_stderr": "5.879", + "latency": "2410.826", + "cost_per_test": "0.195835", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 5.879, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 63.03, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "63.03", + "raw_stderr": "2.642", + "latency": "1314.996", + "cost_per_test": "0.106819", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.642, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 48.689, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "48.689", + "raw_stderr": "3.066", + "latency": "812.638", + "cost_per_test": "0.066012", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 3.066, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.7/9b9f4be5-e9d7-4f5b-ae3c-bf3ce00bf6b0.json b/data/vals_ai/minimax/MiniMax-M2.7/9b9f4be5-e9d7-4f5b-ae3c-bf3ce00bf6b0.json new file mode 100644 index 0000000000000000000000000000000000000000..6cdb5335f65a8e7621fee99417470862d53c0c43 --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.7/9b9f4be5-e9d7-4f5b-ae3c-bf3ce00bf6b0.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/minimax_MiniMax-M2.7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.7", + "id": "minimax/MiniMax-M2.7", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.7", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 54.129, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "54.129", + "raw_stderr": "1.425", + "latency": "28.694", + "cost_per_test": "0.003474", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.425, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 66.558, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.558", + "raw_stderr": "0.92", + "latency": "28.694", + "cost_per_test": "0.003474", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.92, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 78.986, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "78.986", + "raw_stderr": "1.165", + "latency": "28.694", + "cost_per_test": "0.003474", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.165, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.7/9e088ad4-7a4b-4c17-8bcd-e3d696fc8053.json b/data/vals_ai/minimax/MiniMax-M2.7/9e088ad4-7a4b-4c17-8bcd-e3d696fc8053.json new file mode 100644 index 0000000000000000000000000000000000000000..18e363a20e18ce0099530b5e6adae4c6f4072c9a --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.7/9e088ad4-7a4b-4c17-8bcd-e3d696fc8053.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/minimax_MiniMax-M2.7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.7", + "id": "minimax/MiniMax-M2.7", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.7", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 11.926, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "11.926", + "raw_stderr": "2.857", + "latency": "1377.326", + "cost_per_test": "2.819109", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.857, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.7/ab188e29-da2e-40da-8ec3-7622d1f7414a.json b/data/vals_ai/minimax/MiniMax-M2.7/ab188e29-da2e-40da-8ec3-7622d1f7414a.json new file mode 100644 index 0000000000000000000000000000000000000000..0a840d910f6c456f1e4569d0913998725189c9ec --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.7/ab188e29-da2e-40da-8ec3-7622d1f7414a.json @@ -0,0 +1,371 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/minimax_MiniMax-M2.7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.7", + "id": "minimax/MiniMax-M2.7", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.7", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 30.952, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "30.952", + "raw_stderr": "7.22", + "latency": "1154.13", + "cost_per_test": "0.768583", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 7.22, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 72.031, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "72.031", + "raw_stderr": "2.784", + "latency": "967.181", + "cost_per_test": "0.527187", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.784, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 86.082, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "86.082", + "raw_stderr": "2.491", + "latency": "717.426", + "cost_per_test": "0.317675", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.491, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "918.967", + "cost_per_test": "0.506961", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 73.8, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "73.8", + "raw_stderr": "1.968", + "latency": "885.69", + "cost_per_test": "0.466052", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.968, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.7/c2199eaf-46a0-4cbb-8171-23f24f9f9d05.json b/data/vals_ai/minimax/MiniMax-M2.7/c2199eaf-46a0-4cbb-8171-23f24f9f9d05.json new file mode 100644 index 0000000000000000000000000000000000000000..c83710c809762a7e222786fe41f24aaaf9c3d692 --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.7/c2199eaf-46a0-4cbb-8171-23f24f9f9d05.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/minimax_MiniMax-M2.7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.7", + "id": "minimax/MiniMax-M2.7", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.7", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 85.354, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.354", + "raw_stderr": "2.93", + "latency": "145.346", + "cost_per_test": "0.01282", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.93, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 87.879, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.879", + "raw_stderr": "2.583", + "latency": "198.108", + "cost_per_test": "0.015282", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.583, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 86.616, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.616", + "raw_stderr": "1.953", + "latency": "171.727", + "cost_per_test": "0.014051", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.953, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.7/ca5f8309-cb1d-4b68-b330-a25b42321de5.json b/data/vals_ai/minimax/MiniMax-M2.7/ca5f8309-cb1d-4b68-b330-a25b42321de5.json new file mode 100644 index 0000000000000000000000000000000000000000..729e71a91e738dec7c260800905600b2972e14ef --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.7/ca5f8309-cb1d-4b68-b330-a25b42321de5.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/minimax_MiniMax-M2.7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.7", + "id": "minimax/MiniMax-M2.7", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.7", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.261, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.261", + "raw_stderr": "1.169", + "latency": "30.377", + "cost_per_test": "0.002294", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.169, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.847, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.847", + "raw_stderr": "1.169", + "latency": "43.941", + "cost_per_test": "0.003205", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.169, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.869, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.869", + "raw_stderr": "0.938", + "latency": "66.867", + "cost_per_test": "0.005376", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.938, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.098, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.098", + "raw_stderr": "1.709", + "latency": "43.448", + "cost_per_test": "0.00347", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.709, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.953, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.953", + "raw_stderr": "1.25", + "latency": "26.258", + "cost_per_test": "0.002168", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.25, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.613, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.613", + "raw_stderr": "1.415", + "latency": "146.95", + "cost_per_test": "0.012338", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.415, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.384, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.384", + "raw_stderr": "1.476", + "latency": "29.718", + "cost_per_test": "0.002225", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.476, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 65.617, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "65.617", + "raw_stderr": "2.438", + "latency": "29.914", + "cost_per_test": "0.002667", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.438, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 60.49, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "60.49", + "raw_stderr": "1.476", + "latency": "64.611", + "cost_per_test": "0.004826", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.476, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.672, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.672", + "raw_stderr": "0.712", + "latency": "49.349", + "cost_per_test": "0.004255", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.712, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.489, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.489", + "raw_stderr": "1.386", + "latency": "36.909", + "cost_per_test": "0.002513", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.386, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.427, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.427", + "raw_stderr": "0.39", + "latency": "50.005", + "cost_per_test": "0.003947", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.39, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.747, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.747", + "raw_stderr": "1.989", + "latency": "40.819", + "cost_per_test": "0.002607", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.989, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.991, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.991", + "raw_stderr": "0.909", + "latency": "67.562", + "cost_per_test": "0.005466", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.909, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.945, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.945", + "raw_stderr": "1.485", + "latency": "23.345", + "cost_per_test": "0.001847", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.485, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.7/cc937701-62ed-45bf-983d-5e6fc6663277.json b/data/vals_ai/minimax/MiniMax-M2.7/cc937701-62ed-45bf-983d-5e6fc6663277.json new file mode 100644 index 0000000000000000000000000000000000000000..b4325322076c5350e52684f8e53179bcefd84f3c --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.7/cc937701-62ed-45bf-983d-5e6fc6663277.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/minimax_MiniMax-M2.7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.7", + "id": "minimax/MiniMax-M2.7", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.7", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.256, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.256", + "raw_stderr": "1.671", + "latency": "7.227", + "cost_per_test": "0.00105", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.671, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 59.557, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.557", + "raw_stderr": "1.675", + "latency": "14.069", + "cost_per_test": "0.043303", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.675, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.189, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.189", + "raw_stderr": "0.96", + "latency": "10.973", + "cost_per_test": "0.022663", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.96, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 63.753, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.753", + "raw_stderr": "1.641", + "latency": "11.623", + "cost_per_test": "0.023635", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.641, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.7/d8d7699b-64f7-4318-ae23-e2fc8b1668a7.json b/data/vals_ai/minimax/MiniMax-M2.7/d8d7699b-64f7-4318-ae23-e2fc8b1668a7.json new file mode 100644 index 0000000000000000000000000000000000000000..b5c833ed0ccac00c72688276b5cd38e375f1a2bf --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.7/d8d7699b-64f7-4318-ae23-e2fc8b1668a7.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/minimax_MiniMax-M2.7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.7", + "id": "minimax/MiniMax-M2.7", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.7", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.137, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.137", + "raw_stderr": "0.754", + "latency": "43.179", + "cost_per_test": "0.003387", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.754, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 56.0, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "56", + "raw_stderr": "2.653", + "latency": "664.796", + "cost_per_test": "0.044094", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.653, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 85.64, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "85.64", + "raw_stderr": "1.792", + "latency": "248.827", + "cost_per_test": "0.016778", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.792, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 79.926, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "79.926", + "raw_stderr": "1.096", + "latency": "318.934", + "cost_per_test": "0.02142", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.096, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.7/dd5283f7-4e67-42f8-b9a7-2b39c7ef9650.json b/data/vals_ai/minimax/MiniMax-M2.7/dd5283f7-4e67-42f8-b9a7-2b39c7ef9650.json new file mode 100644 index 0000000000000000000000000000000000000000..f42ec6fc99ab0c8d71de0c04ef478c92757341a8 --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.7/dd5283f7-4e67-42f8-b9a7-2b39c7ef9650.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/minimax_MiniMax-M2.7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.7", + "id": "minimax/MiniMax-M2.7", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.7", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 0.5, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0.5", + "raw_stderr": "0.456", + "latency": "3472.218", + "cost_per_test": "1.262533", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.456, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 9.333, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "9.333", + "raw_stderr": "4.736", + "latency": "3013.297", + "cost_per_test": "1.04722", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 4.736, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 4.916, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "4.916", + "raw_stderr": "2.379", + "latency": "3242.758", + "cost_per_test": "1.154876", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.379, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.7/e1030a1d-e744-48bd-bf46-976c123a5b95.json b/data/vals_ai/minimax/MiniMax-M2.7/e1030a1d-e744-48bd-bf46-976c123a5b95.json new file mode 100644 index 0000000000000000000000000000000000000000..378744b379c5a2c6b3706396855178b5b79c4d28 --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.7/e1030a1d-e744-48bd-bf46-976c123a5b95.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/minimax_MiniMax-M2.7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.7", + "id": "minimax/MiniMax-M2.7", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.7", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 79.867, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "79.867", + "raw_stderr": "1.86", + "latency": "27.25", + "cost_per_test": "0.005124", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.86, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.7/e17b5629-1430-458f-a080-8b62a7d92fba.json b/data/vals_ai/minimax/MiniMax-M2.7/e17b5629-1430-458f-a080-8b62a7d92fba.json new file mode 100644 index 0000000000000000000000000000000000000000..baacb27d55e305ff8fab44efc44ce4eb08a9b270 --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.7/e17b5629-1430-458f-a080-8b62a7d92fba.json @@ -0,0 +1,438 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/minimax_MiniMax-M2.7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.7", + "id": "minimax/MiniMax-M2.7", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.7", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 89.339, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.339", + "raw_stderr": "0.689", + "latency": "8.912", + "cost_per_test": "0.000458", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.689, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.653, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.653", + "raw_stderr": "0.203", + "latency": "8.579", + "cost_per_test": "0.000625", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.203, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.118, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.118", + "raw_stderr": "0.64", + "latency": "8.837", + "cost_per_test": "0.000927", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.64, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.98, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.98", + "raw_stderr": "0.411", + "latency": "8.916", + "cost_per_test": "0.000626", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.411, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.804, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.804", + "raw_stderr": "0.933", + "latency": "6.991", + "cost_per_test": "0.000552", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.933, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.988, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.988", + "raw_stderr": "1.559", + "latency": "11.261", + "cost_per_test": "0.000568", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.559, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M2.7/f2a4fd1e-3ba9-4e4d-ac7b-d79c6254ad85.json b/data/vals_ai/minimax/MiniMax-M2.7/f2a4fd1e-3ba9-4e4d-ac7b-d79c6254ad85.json new file mode 100644 index 0000000000000000000000000000000000000000..326090918ddb890353f01727bcb30911a21a9c5a --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M2.7/f2a4fd1e-3ba9-4e4d-ac7b-d79c6254ad85.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/minimax_MiniMax-M2.7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M2.7", + "id": "minimax/MiniMax-M2.7", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M2.7", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:minimax/MiniMax-M2.7:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 3.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "3", + "raw_stderr": "1.706", + "latency": "1634.068", + "cost_per_test": "0.04293", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "196608", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.706, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 196608, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M3/00625946-62c2-4259-af1d-8a05fcf8dc6e.json b/data/vals_ai/minimax/MiniMax-M3/00625946-62c2-4259-af1d-8a05fcf8dc6e.json new file mode 100644 index 0000000000000000000000000000000000000000..4df0914e803476e3d71514974578f3b88e6a873b --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M3/00625946-62c2-4259-af1d-8a05fcf8dc6e.json @@ -0,0 +1,676 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/emb/minimax_MiniMax-M3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - EMB", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M3", + "id": "minimax/MiniMax-M3", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M3", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "emb:comps:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.emb.comps", + "source_data": { + "dataset_name": "EMB - Comps", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "comps", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Comps).", + "metric_id": "vals_ai.emb.comps.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 52.763, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "comps", + "task_name": "Comps", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.763", + "raw_stderr": "7.575", + "latency": "1421.599", + "cost_per_test": "1.442667", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 7.575, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:dataroom:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.emb.dataroom", + "source_data": { + "dataset_name": "EMB - Dataroom", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dataroom", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Dataroom).", + "metric_id": "vals_ai.emb.dataroom.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 52.743, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dataroom", + "task_name": "Dataroom", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.743", + "raw_stderr": "10.633", + "latency": "1882.98", + "cost_per_test": "2.081819", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 10.633, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:dcf:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.emb.dcf", + "source_data": { + "dataset_name": "EMB - DCF", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dcf", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (DCF).", + "metric_id": "vals_ai.emb.dcf.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 36.629, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dcf", + "task_name": "DCF", + "dataset_type": "private", + "industry": "finance", + "raw_score": "36.629", + "raw_stderr": "8.66", + "latency": "1752.148", + "cost_per_test": "2.435709", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 8.66, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:lbo:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.emb.lbo", + "source_data": { + "dataset_name": "EMB - LBO", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "lbo", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (LBO).", + "metric_id": "vals_ai.emb.lbo.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 44.93, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "lbo", + "task_name": "LBO", + "dataset_type": "private", + "industry": "finance", + "raw_score": "44.93", + "raw_stderr": "5.701", + "latency": "2429.627", + "cost_per_test": "3.032953", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 5.701, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:ma:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.emb.ma", + "source_data": { + "dataset_name": "EMB - M&A", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "ma", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (M&A).", + "metric_id": "vals_ai.emb.ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 48.366, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "ma", + "task_name": "M&A", + "dataset_type": "private", + "industry": "finance", + "raw_score": "48.366", + "raw_stderr": "9.642", + "latency": "1834.694", + "cost_per_test": "1.463202", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 9.642, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:misc:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.emb.misc", + "source_data": { + "dataset_name": "EMB - Misc", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "misc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Misc).", + "metric_id": "vals_ai.emb.misc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 50.022, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "misc", + "task_name": "Misc", + "dataset_type": "private", + "industry": "finance", + "raw_score": "50.022", + "raw_stderr": "5.091", + "latency": "2177.177", + "cost_per_test": "2.184482", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 5.091, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:operating:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.emb.operating", + "source_data": { + "dataset_name": "EMB - Operating", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "operating", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Operating).", + "metric_id": "vals_ai.emb.operating.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 48.863, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "operating", + "task_name": "Operating", + "dataset_type": "private", + "industry": "finance", + "raw_score": "48.863", + "raw_stderr": "7.366", + "latency": "1780.736", + "cost_per_test": "1.98786", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 7.366, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:overall:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.emb.overall", + "source_data": { + "dataset_name": "EMB - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Overall).", + "metric_id": "vals_ai.emb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 47.759, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "47.759", + "raw_stderr": "2.946", + "latency": "1890.343", + "cost_per_test": "2.091762", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.946, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:scratch:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.emb.scratch", + "source_data": { + "dataset_name": "EMB - Scratch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "scratch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Scratch).", + "metric_id": "vals_ai.emb.scratch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 61.012, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "scratch", + "task_name": "Scratch", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.012", + "raw_stderr": "2.558", + "latency": "1976.511", + "cost_per_test": "1.854875", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.558, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:template:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.emb.template", + "source_data": { + "dataset_name": "EMB - Template", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "template", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Template).", + "metric_id": "vals_ai.emb.template.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 34.546, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "template", + "task_name": "Template", + "dataset_type": "private", + "industry": "finance", + "raw_score": "34.546", + "raw_stderr": "4.741", + "latency": "1804.175", + "cost_per_test": "2.328648", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 4.741, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M3/1dc3a9dd-90cf-4d77-a597-e1619b87289e.json b/data/vals_ai/minimax/MiniMax-M3/1dc3a9dd-90cf-4d77-a597-e1619b87289e.json new file mode 100644 index 0000000000000000000000000000000000000000..f9956308f3f3d67ad9b6027cabc599da831ed8bb --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M3/1dc3a9dd-90cf-4d77-a597-e1619b87289e.json @@ -0,0 +1,103 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/minimax_MiniMax-M3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M3", + "id": "minimax/MiniMax-M3", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M3", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 81.156, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.156", + "raw_stderr": "0.94", + "latency": "68.354", + "cost_per_test": "0.009431", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.94, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M3/1ecdc9fe-468d-4e11-9d8f-bf826264171a.json b/data/vals_ai/minimax/MiniMax-M3/1ecdc9fe-468d-4e11-9d8f-bf826264171a.json new file mode 100644 index 0000000000000000000000000000000000000000..e3c49ae97b156e43b726594ec9c02da902362313 --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M3/1ecdc9fe-468d-4e11-9d8f-bf826264171a.json @@ -0,0 +1,3386 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/hlab/minimax_MiniMax-M3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Harvey's Legal Agent Benchmark", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M3", + "id": "minimax/MiniMax-M3", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M3", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "hlab:antitrust_competition:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1344.043", + "cost_per_test": "1.499283", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:arbitration_international_dispute_resolution:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1866.312", + "cost_per_test": "2.296246", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:banking_finance:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1180.706", + "cost_per_test": "1.170507", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:bankruptcy_restructuring:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1343.818", + "cost_per_test": "1.508844", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:capital_markets:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "931.198", + "cost_per_test": "0.941373", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:corporate_governance:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1401.48", + "cost_per_test": "1.37422", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:corporate_ma:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1139.463", + "cost_per_test": "1.100807", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Criteria Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Criteria Pass Rate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 86.279, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate", + "task_name": "Criteria Pass Rate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.279", + "raw_stderr": "0.414", + "latency": "1351.039", + "cost_per_test": "1.460248", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.414, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__antitrust_competition:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 83.333, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.333", + "raw_stderr": "2.247", + "latency": "1344.043", + "cost_per_test": "1.499283", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.247, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__arbitration_international_dispute_resolution:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 86.738, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.738", + "raw_stderr": "2.034", + "latency": "1866.312", + "cost_per_test": "2.296246", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.034, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__banking_finance:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 86.617, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.617", + "raw_stderr": "2.08", + "latency": "1180.706", + "cost_per_test": "1.170507", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.08, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__bankruptcy_restructuring:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 80.277, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.277", + "raw_stderr": "2.345", + "latency": "1343.818", + "cost_per_test": "1.508844", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.345, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__capital_markets:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 89.02, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.02", + "raw_stderr": "1.962", + "latency": "931.198", + "cost_per_test": "0.941373", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.962, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_governance:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 88.014, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.014", + "raw_stderr": "1.904", + "latency": "1401.48", + "cost_per_test": "1.37422", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.904, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_ma:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 90.064, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.064", + "raw_stderr": "1.696", + "latency": "1139.463", + "cost_per_test": "1.100807", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.696, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__data_privacy_cybersecurity:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 92.226, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "92.226", + "raw_stderr": "1.594", + "latency": "1542.717", + "cost_per_test": "2.157888", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.594, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__emerging_companies_venture_capital:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 83.276, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.276", + "raw_stderr": "2.184", + "latency": "1607.584", + "cost_per_test": "1.729985", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.184, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__employment_labor:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 77.903, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.903", + "raw_stderr": "2.544", + "latency": "1162.314", + "cost_per_test": "1.29091", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.544, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__energy_natural_resources:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 93.399, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "93.399", + "raw_stderr": "1.429", + "latency": "1700.964", + "cost_per_test": "1.80131", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.429, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__environmental_esg:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 83.512, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.512", + "raw_stderr": "2.226", + "latency": "1732.445", + "cost_per_test": "1.768355", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.226, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__funds_asset_management:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 82.507, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.507", + "raw_stderr": "2.054", + "latency": "1624.19", + "cost_per_test": "1.845597", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.054, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__healthcare_life_sciences:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 89.769, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.769", + "raw_stderr": "1.744", + "latency": "1435.346", + "cost_per_test": "1.4604", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.744, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__immigration:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 82.182, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.182", + "raw_stderr": "2.312", + "latency": "1061.727", + "cost_per_test": "0.987413", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.312, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__insurance:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 87.18, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.18", + "raw_stderr": "2.027", + "latency": "1205.072", + "cost_per_test": "1.05308", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.027, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__intellectual_property:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 90.459, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.459", + "raw_stderr": "1.749", + "latency": "1246.857", + "cost_per_test": "1.349078", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.749, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__international_trade_sanctions:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 86.508, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.508", + "raw_stderr": "2.156", + "latency": "1095.196", + "cost_per_test": "0.888266", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.156, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__litigation_dispute_resolution:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 88.764, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.764", + "raw_stderr": "1.936", + "latency": "1402.267", + "cost_per_test": "1.540985", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.936, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__real_estate:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 88.338, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.338", + "raw_stderr": "1.736", + "latency": "1101.194", + "cost_per_test": "0.893435", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.736, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__structured_finance_securitization:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 81.27, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.27", + "raw_stderr": "2.202", + "latency": "925.639", + "cost_per_test": "1.250112", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.202, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__tax:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 84.95, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.95", + "raw_stderr": "2.071", + "latency": "1470.603", + "cost_per_test": "1.553621", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.071, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__trusts_estates_private_client:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 85.496, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.496", + "raw_stderr": "2.18", + "latency": "1796.997", + "cost_per_test": "2.467191", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.18, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__white_collar_defense_investigations:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 88.552, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.552", + "raw_stderr": "1.851", + "latency": "1106.8", + "cost_per_test": "1.117058", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.851, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:data_privacy_cybersecurity:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1542.717", + "cost_per_test": "2.157888", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:emerging_companies_venture_capital:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "1607.584", + "cost_per_test": "1.729985", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:employment_labor:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1162.314", + "cost_per_test": "1.29091", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:energy_natural_resources:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "1700.964", + "cost_per_test": "1.80131", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:environmental_esg:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1732.445", + "cost_per_test": "1.768355", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:funds_asset_management:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1624.19", + "cost_per_test": "1.845597", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:healthcare_life_sciences:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "1435.346", + "cost_per_test": "1.4604", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:immigration:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1061.727", + "cost_per_test": "0.987413", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:insurance:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1205.072", + "cost_per_test": "1.05308", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:intellectual_property:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1246.857", + "cost_per_test": "1.349078", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:international_trade_sanctions:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1095.196", + "cost_per_test": "0.888266", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:litigation_dispute_resolution:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1402.267", + "cost_per_test": "1.540985", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:overall:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.overall", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Overall).", + "metric_id": "vals_ai.hlab.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 4.167, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "4.167", + "raw_stderr": "1.646", + "latency": "1351.039", + "cost_per_test": "1.460248", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.646, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:real_estate:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1101.194", + "cost_per_test": "0.893435", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:structured_finance_securitization:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "925.639", + "cost_per_test": "1.250112", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:tax:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1470.603", + "cost_per_test": "1.553621", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:trusts_estates_private_client:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1796.997", + "cost_per_test": "2.467191", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:white_collar_defense_investigations:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.hlab.white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1106.8", + "cost_per_test": "1.117058", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M3/235c32f1-3588-4d8c-8c51-b86b3a869a47.json b/data/vals_ai/minimax/MiniMax-M3/235c32f1-3588-4d8c-8c51-b86b3a869a47.json new file mode 100644 index 0000000000000000000000000000000000000000..faeddfab1839f4de6e0690a6787b77cb39346ded --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M3/235c32f1-3588-4d8c-8c51-b86b3a869a47.json @@ -0,0 +1,1188 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_research/minimax_MiniMax-M3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Legal Research Bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M3", + "id": "minimax/MiniMax-M3", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M3", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_research:administrative_regulatory:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.legal_research.administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 36.364, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "36.364", + "raw_stderr": "5.967", + "latency": "755.036", + "cost_per_test": "0.352926", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 5.967, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:business_commercial:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.legal_research.business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 25.581, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "25.581", + "raw_stderr": "6.733", + "latency": "867.939", + "cost_per_test": "0.34538", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 6.733, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:civil_litigation:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.legal_research.civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 25.926, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "25.926", + "raw_stderr": "8.594", + "latency": "823.116", + "cost_per_test": "0.385069", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 8.594, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:constitutional_civil_rights:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.legal_research.constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 34.211, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "34.211", + "raw_stderr": "7.799", + "latency": "733.352", + "cost_per_test": "0.291222", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 7.799, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:criminal:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.legal_research.criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 26.316, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "26.316", + "raw_stderr": "5.884", + "latency": "855.631", + "cost_per_test": "0.325282", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 5.884, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:family:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.legal_research.family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 13.636, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "13.636", + "raw_stderr": "7.489", + "latency": "951.192", + "cost_per_test": "0.370092", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 7.489, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:health:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.legal_research.health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 46.667, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "46.667", + "raw_stderr": "9.264", + "latency": "800.16", + "cost_per_test": "0.394232", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 9.264, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:immigration:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.legal_research.immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 27.273, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "27.273", + "raw_stderr": "14.084", + "latency": "882.27", + "cost_per_test": "0.482243", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 14.084, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:overall:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.legal_research.overall", + "source_data": { + "dataset_name": "Legal Research Bench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Overall).", + "metric_id": "vals_ai.legal_research.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 29.808, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "legal", + "raw_score": "29.808", + "raw_stderr": "3.179", + "latency": "813.838", + "cost_per_test": "0.342955", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 3.179, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass", + "source_data": { + "dataset_name": "Legal Research Bench - Weighted Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Weighted Pass).", + "metric_id": "vals_ai.legal_research.weighted_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 81.146, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass", + "task_name": "Weighted Pass", + "dataset_type": "private", + "industry": "legal", + "raw_score": "81.146", + "raw_stderr": "1.532", + "latency": "813.838", + "cost_per_test": "0.342955", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.532, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__administrative_regulatory:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.weighted_pass__administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 86.153, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "86.153", + "raw_stderr": "2.391", + "latency": "755.036", + "cost_per_test": "0.352926", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.391, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__business_commercial:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.weighted_pass__business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 82.694, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "82.694", + "raw_stderr": "3.326", + "latency": "867.939", + "cost_per_test": "0.34538", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 3.326, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__civil_litigation:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.weighted_pass__civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 84.854, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "84.854", + "raw_stderr": "2.956", + "latency": "823.116", + "cost_per_test": "0.385069", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.956, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__constitutional_civil_rights:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 85.08, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "85.08", + "raw_stderr": "2.727", + "latency": "733.352", + "cost_per_test": "0.291222", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.727, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__criminal:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.weighted_pass__criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 78.11, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "78.11", + "raw_stderr": "2.891", + "latency": "855.631", + "cost_per_test": "0.325282", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.891, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__family:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.weighted_pass__family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 68.039, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "68.039", + "raw_stderr": "5.235", + "latency": "951.192", + "cost_per_test": "0.370092", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 5.235, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__health:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.weighted_pass__health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 92.161, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "92.161", + "raw_stderr": "2.489", + "latency": "800.16", + "cost_per_test": "0.394232", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.489, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__immigration:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.weighted_pass__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 78.593, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "78.593", + "raw_stderr": "8.748", + "latency": "882.27", + "cost_per_test": "0.482243", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 8.748, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M3/2466322a-ec20-4bda-bbf9-0d0121cda3fd.json b/data/vals_ai/minimax/MiniMax-M3/2466322a-ec20-4bda-bbf9-0d0121cda3fd.json new file mode 100644 index 0000000000000000000000000000000000000000..99a42c79fbe0d1cf53f3fe7754c5de1a72c7cc41 --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M3/2466322a-ec20-4bda-bbf9-0d0121cda3fd.json @@ -0,0 +1,371 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/minimax_MiniMax-M3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M3", + "id": "minimax/MiniMax-M3", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M3", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 47.619, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "47.619", + "raw_stderr": "7.8", + "latency": "1117.332", + "cost_per_test": "0.747003", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 7.8, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 72.797, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "72.797", + "raw_stderr": "2.76", + "latency": "801.959", + "cost_per_test": "0.463694", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.76, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 84.536, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "84.536", + "raw_stderr": "2.603", + "latency": "530.992", + "cost_per_test": "0.274669", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.603, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "1256.214", + "cost_per_test": "0.83109", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 75.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "75", + "raw_stderr": "1.938", + "latency": "726.041", + "cost_per_test": "0.416355", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.938, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M3/29ed3b2b-4dc3-45cc-96a2-7f7a6d4d9190.json b/data/vals_ai/minimax/MiniMax-M3/29ed3b2b-4dc3-45cc-96a2-7f7a6d4d9190.json new file mode 100644 index 0000000000000000000000000000000000000000..05d1eb8430a8d21565c349da5906cb50184cc8e3 --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M3/29ed3b2b-4dc3-45cc-96a2-7f7a6d4d9190.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/minimax_MiniMax-M3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M3", + "id": "minimax/MiniMax-M3", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M3", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 64.918, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.918", + "raw_stderr": "1.629", + "latency": "18.56", + "cost_per_test": "0.002351", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.629, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 69.697, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "69.697", + "raw_stderr": "1.569", + "latency": "31.268", + "cost_per_test": "0.105284", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.569, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 68.104, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.104", + "raw_stderr": "0.918", + "latency": "24.459", + "cost_per_test": "0.049736", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.918, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 69.697, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "69.697", + "raw_stderr": "1.569", + "latency": "23.548", + "cost_per_test": "0.041574", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.569, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M3/2c7d2abc-e253-425d-a3a5-51e4c2eb2c7d.json b/data/vals_ai/minimax/MiniMax-M3/2c7d2abc-e253-425d-a3a5-51e4c2eb2c7d.json new file mode 100644 index 0000000000000000000000000000000000000000..d1419caf8c98543ad399c1f7790926b381bbff5c --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M3/2c7d2abc-e253-425d-a3a5-51e4c2eb2c7d.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/minimax_MiniMax-M3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M3", + "id": "minimax/MiniMax-M3", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M3", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 77.266, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "77.266", + "raw_stderr": "1.182", + "latency": "39.014", + "cost_per_test": "0.009618", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.182, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 68.362, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.362", + "raw_stderr": "0.91", + "latency": "26.756", + "cost_per_test": "0.008326", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.91, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 59.459, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.459", + "raw_stderr": "1.384", + "latency": "14.499", + "cost_per_test": "0.007033", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.384, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M3/40623280-36c0-4ff7-8566-5136c83f0833.json b/data/vals_ai/minimax/MiniMax-M3/40623280-36c0-4ff7-8566-5136c83f0833.json new file mode 100644 index 0000000000000000000000000000000000000000..f8252ae0119597fe8420cd661c56a50148135b1d --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M3/40623280-36c0-4ff7-8566-5136c83f0833.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/minimax_MiniMax-M3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M3", + "id": "minimax/MiniMax-M3", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M3", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.748, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.748", + "raw_stderr": "1.01", + "latency": "33.196", + "cost_per_test": "0.002851", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.01, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.34, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.34", + "raw_stderr": "1.164", + "latency": "30.784", + "cost_per_test": "0.003849", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.164, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.548, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.548", + "raw_stderr": "0.912", + "latency": "50.811", + "cost_per_test": "0.007038", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.912, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.293, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.293", + "raw_stderr": "1.683", + "latency": "38.531", + "cost_per_test": "0.004813", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.683, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.915, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.915", + "raw_stderr": "1.136", + "latency": "24.193", + "cost_per_test": "0.002542", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.136, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.115, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.115", + "raw_stderr": "1.29", + "latency": "136.739", + "cost_per_test": "0.017722", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.29, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.196, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.196", + "raw_stderr": "1.406", + "latency": "29.591", + "cost_per_test": "0.002572", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.406, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.228, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.228", + "raw_stderr": "2.268", + "latency": "25.596", + "cost_per_test": "0.00265", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.268, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 66.485, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.485", + "raw_stderr": "1.429", + "latency": "57.733", + "cost_per_test": "0.005721", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.429, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.597, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.597", + "raw_stderr": "0.675", + "latency": "33.812", + "cost_per_test": "0.004697", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.675, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.277, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.277", + "raw_stderr": "1.289", + "latency": "18.963", + "cost_per_test": "0.002008", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.289, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.22, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.22", + "raw_stderr": "0.364", + "latency": "41.356", + "cost_per_test": "0.004806", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.364, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.16, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.16", + "raw_stderr": "1.792", + "latency": "23.038", + "cost_per_test": "0.002277", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.792, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.839, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.839", + "raw_stderr": "0.847", + "latency": "50.009", + "cost_per_test": "0.006439", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.847, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.333, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.333", + "raw_stderr": "1.323", + "latency": "25.992", + "cost_per_test": "0.002098", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.323, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M3/43896f76-d2db-4770-ba52-1be93a8ed6a6.json b/data/vals_ai/minimax/MiniMax-M3/43896f76-d2db-4770-ba52-1be93a8ed6a6.json new file mode 100644 index 0000000000000000000000000000000000000000..515dd9b5e835d1e8bdb1ef1fd5efac15608f3c80 --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M3/43896f76-d2db-4770-ba52-1be93a8ed6a6.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/public-benefits-bench/minimax_MiniMax-M3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Public Benefits Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "social mobility", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M3", + "id": "minimax/MiniMax-M3", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M3", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "public-benefits-bench:overall:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.public-benefits-bench.overall", + "source_data": { + "dataset_name": "Public Benefits Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Public Benefits Bench v1.1 (Overall).", + "metric_id": "vals_ai.public-benefits-bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "score_details": { + "score": 64.141, + "details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "social mobility", + "raw_score": "64.141", + "raw_stderr": "1.247", + "latency": "841.824", + "cost_per_test": "0.2543", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.247, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M3/459c5289-caf5-40a7-a660-99ca9b028c9e.json b/data/vals_ai/minimax/MiniMax-M3/459c5289-caf5-40a7-a660-99ca9b028c9e.json new file mode 100644 index 0000000000000000000000000000000000000000..0db57df5a772ffa7428c0e5b21a237e4d38a2dcf --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M3/459c5289-caf5-40a7-a660-99ca9b028c9e.json @@ -0,0 +1,740 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/minimax_MiniMax-M3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M3", + "id": "minimax/MiniMax-M3", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M3", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 42.321, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "42.321", + "raw_stderr": "2.357", + "latency": "436.409", + "cost_per_test": "0.280596", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.357, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 36.693, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "36.693", + "raw_stderr": "0.157", + "latency": "496.894", + "cost_per_test": "0.32345", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.157, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 41.669, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "41.669", + "raw_stderr": "3.912", + "latency": "577.327", + "cost_per_test": "0.388805", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 3.912, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 59.831, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.831", + "raw_stderr": "1.976", + "latency": "471.873", + "cost_per_test": "0.346007", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.976, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 66.191, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.191", + "raw_stderr": "1.409", + "latency": "303.41", + "cost_per_test": "0.130797", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.409, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 13.99, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "13.99", + "raw_stderr": "2.011", + "latency": "674.781", + "cost_per_test": "0.358739", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.011, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 61.691, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.691", + "raw_stderr": "2.969", + "latency": "498.741", + "cost_per_test": "0.352382", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.969, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 66.483, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.483", + "raw_stderr": "0.823", + "latency": "376.513", + "cost_per_test": "0.308361", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.823, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 59.007, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.007", + "raw_stderr": "0.331", + "latency": "369.203", + "cost_per_test": "0.194569", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.331, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:overall:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 48.269, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "48.269", + "raw_stderr": "0.437", + "latency": "496.894", + "cost_per_test": "0.32345", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.437, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 23.24, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "23.24", + "raw_stderr": "1.574", + "latency": "804.141", + "cost_per_test": "0.571079", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.574, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M3/4debdded-95df-43ca-8261-28e4928cf267.json b/data/vals_ai/minimax/MiniMax-M3/4debdded-95df-43ca-8261-28e4928cf267.json new file mode 100644 index 0000000000000000000000000000000000000000..9801146d81c073821f270f191d02feef50a66b73 --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M3/4debdded-95df-43ca-8261-28e4928cf267.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/minimax_MiniMax-M3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M3", + "id": "minimax/MiniMax-M3", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M3", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 83.333, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "83.333", + "raw_stderr": "8.333", + "latency": "22771.001", + "cost_per_test": "4.582903", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 8.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 30.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "30", + "raw_stderr": "1.925", + "latency": "3036.133", + "cost_per_test": "0.611054", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.925, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 64.242, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "64.242", + "raw_stderr": "1.603", + "latency": "1656.073", + "cost_per_test": "0.333302", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.603, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 53.558, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "53.558", + "raw_stderr": "0.749", + "latency": "1023.416", + "cost_per_test": "0.205973", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.749, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M3/9abea29c-2f23-4718-be56-7069fbdfaf66.json b/data/vals_ai/minimax/MiniMax-M3/9abea29c-2f23-4718-be56-7069fbdfaf66.json new file mode 100644 index 0000000000000000000000000000000000000000..b959ff5aeb83c83a881afe454f7904ec1c42e200 --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M3/9abea29c-2f23-4718-be56-7069fbdfaf66.json @@ -0,0 +1,536 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_multimodal_index/minimax_MiniMax-M3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Multimodal Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M3", + "id": "minimax/MiniMax-M3", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M3", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_multimodal_index:corp_fin_v2:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.vals_multimodal_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Multimodal Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (CorpFin v2).", + "metric_id": "vals_ai.vals_multimodal_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 69.697, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "69.697", + "raw_stderr": "1.569", + "latency": "23.548", + "cost_per_test": "0.041574", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.569, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:finance_agent:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.vals_multimodal_index.finance_agent", + "source_data": { + "dataset_name": "Vals Multimodal Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_multimodal_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 48.008, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "48.008", + "raw_stderr": "1.421", + "latency": "502.942", + "cost_per_test": "0.291233", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.421, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:mortgage_tax:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.vals_multimodal_index.mortgage_tax", + "source_data": { + "dataset_name": "Vals Multimodal Index - Mortgage Tax", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "mortgage_tax", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Mortgage Tax).", + "metric_id": "vals_ai.vals_multimodal_index.mortgage_tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 68.362, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "mortgage_tax", + "task_name": "Mortgage Tax", + "dataset_type": "private", + "industry": "index", + "raw_score": "68.362", + "raw_stderr": "0.91", + "latency": "26.756", + "cost_per_test": "0.008326", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.91, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:overall:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.vals_multimodal_index.overall", + "source_data": { + "dataset_name": "Vals Multimodal Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Overall).", + "metric_id": "vals_ai.vals_multimodal_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 59.974, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "59.974", + "raw_stderr": "1.646", + "latency": "1076.237", + "cost_per_test": "1.076447", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.646, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:sage:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.vals_multimodal_index.sage", + "source_data": { + "dataset_name": "Vals Multimodal Index - SAGE", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "sage", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SAGE).", + "metric_id": "vals_ai.vals_multimodal_index.sage.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 50.574, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "sage", + "task_name": "SAGE", + "dataset_type": "private", + "industry": "index", + "raw_score": "50.574", + "raw_stderr": "3.436", + "latency": "118.538", + "cost_per_test": "0.025844", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 3.436, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:swebench:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.vals_multimodal_index.swebench", + "source_data": { + "dataset_name": "Vals Multimodal Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SWE-bench).", + "metric_id": "vals_ai.vals_multimodal_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 69.608, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "69.608", + "raw_stderr": "4.554", + "latency": "765.232", + "cost_per_test": "0.508278" + }, + "uncertainty": { + "standard_error": { + "value": 4.554, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:terminal_bench_2_1:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.vals_multimodal_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Multimodal Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_multimodal_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 53.558, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "53.558", + "raw_stderr": "0.749", + "latency": "1023.416", + "cost_per_test": "0.205973", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.749, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:vibe_code_bench:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.vals_multimodal_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Multimodal Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_multimodal_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 56.54, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "56.54", + "raw_stderr": "7.963", + "latency": "5073.23", + "cost_per_test": "6.453899", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 7.963, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M3/9d1bb9a4-37a7-4fe7-855f-c037794075ec.json b/data/vals_ai/minimax/MiniMax-M3/9d1bb9a4-37a7-4fe7-855f-c037794075ec.json new file mode 100644 index 0000000000000000000000000000000000000000..cdb8d22047177c67b2ac8f34c11f0b8ca0e1ceb6 --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M3/9d1bb9a4-37a7-4fe7-855f-c037794075ec.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/minimax_MiniMax-M3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M3", + "id": "minimax/MiniMax-M3", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M3", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.758, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.758", + "raw_stderr": "0.617", + "latency": "67.765", + "cost_per_test": "0.005632", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.617, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 59.714, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "59.714", + "raw_stderr": "2.622", + "latency": "723.191", + "cost_per_test": "0.076845", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.622, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 87.99, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "87.99", + "raw_stderr": "1.661", + "latency": "308.563", + "cost_per_test": "0.029814", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.661, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 82.154, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "82.154", + "raw_stderr": "1.055", + "latency": "366.506", + "cost_per_test": "0.03743", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.055, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M3/9d25c364-85d5-4b68-8a46-d380248a05ba.json b/data/vals_ai/minimax/MiniMax-M3/9d25c364-85d5-4b68-8a46-d380248a05ba.json new file mode 100644 index 0000000000000000000000000000000000000000..52bb30e7aebfa810cac706610cedf92e299940de --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M3/9d25c364-85d5-4b68-8a46-d380248a05ba.json @@ -0,0 +1,408 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/minimax_MiniMax-M3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M3", + "id": "minimax/MiniMax-M3", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M3", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 69.697, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "69.697", + "raw_stderr": "1.569", + "latency": "23.548", + "cost_per_test": "0.041574", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.569, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 48.008, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "48.008", + "raw_stderr": "1.421", + "latency": "502.942", + "cost_per_test": "0.291233", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.421, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:overall:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 58.939, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "58.939", + "raw_stderr": "1.817", + "latency": "1477.674", + "cost_per_test": "1.500191", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.817, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 69.608, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "69.608", + "raw_stderr": "4.554", + "latency": "765.232", + "cost_per_test": "0.508278" + }, + "uncertainty": { + "standard_error": { + "value": 4.554, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 53.558, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "53.558", + "raw_stderr": "0.749", + "latency": "1023.416", + "cost_per_test": "0.205973", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.749, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 56.54, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "56.54", + "raw_stderr": "7.963", + "latency": "5073.23", + "cost_per_test": "6.453899", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 7.963, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M3/a6e524a5-06c6-4483-ac04-cc245df88fd9.json b/data/vals_ai/minimax/MiniMax-M3/a6e524a5-06c6-4483-ac04-cc245df88fd9.json new file mode 100644 index 0000000000000000000000000000000000000000..6d3ea153e29839208151a985f0837010cd6b84af --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M3/a6e524a5-06c6-4483-ac04-cc245df88fd9.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/minimax_MiniMax-M3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M3", + "id": "minimax/MiniMax-M3", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M3", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 87.252, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "87.252", + "raw_stderr": "1.957", + "latency": "123.967", + "cost_per_test": "0.013748", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.957, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M3/cbb01c77-b3ee-432e-9198-a821c68826f3.json b/data/vals_ai/minimax/MiniMax-M3/cbb01c77-b3ee-432e-9198-a821c68826f3.json new file mode 100644 index 0000000000000000000000000000000000000000..e81822f61a61b8b6d0acc72a7c559865e129ee4d --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M3/cbb01c77-b3ee-432e-9198-a821c68826f3.json @@ -0,0 +1,438 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/minimax_MiniMax-M3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M3", + "id": "minimax/MiniMax-M3", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M3", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 89.276, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.276", + "raw_stderr": "0.742", + "latency": "9.224", + "cost_per_test": "0.001085", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.742, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.511, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.511", + "raw_stderr": "0.199", + "latency": "6.584", + "cost_per_test": "0.00083", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.199, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 90.862, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.862", + "raw_stderr": "0.553", + "latency": "6.976", + "cost_per_test": "0.0009", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.553, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 85.415, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.415", + "raw_stderr": "0.436", + "latency": "8.205", + "cost_per_test": "0.00091", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.436, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.455, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.455", + "raw_stderr": "0.866", + "latency": "6.409", + "cost_per_test": "0.000697", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.866, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.972, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.972", + "raw_stderr": "1.76", + "latency": "11.832", + "cost_per_test": "0.001039", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.76, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M3/cd40355a-fcb3-4b09-94f6-f866dc312db5.json b/data/vals_ai/minimax/MiniMax-M3/cd40355a-fcb3-4b09-94f6-f866dc312db5.json new file mode 100644 index 0000000000000000000000000000000000000000..c65bcc06f1c328172eacc6a9139aa8afa3fb8037 --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M3/cd40355a-fcb3-4b09-94f6-f866dc312db5.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/minimax_MiniMax-M3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M3", + "id": "minimax/MiniMax-M3", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M3", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 19.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "19", + "raw_stderr": "3.943", + "latency": "1394.151", + "cost_per_test": "0.371186", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 3.943, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M3/d0e61303-cdf3-47be-96a2-3db0ca761523.json b/data/vals_ai/minimax/MiniMax-M3/d0e61303-cdf3-47be-96a2-3db0ca761523.json new file mode 100644 index 0000000000000000000000000000000000000000..4fd0c2f950d5ac4f32c6d8337be99e4e84a60641 --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M3/d0e61303-cdf3-47be-96a2-3db0ca761523.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/code-migration/minimax_MiniMax-M3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Code Migration", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M3", + "id": "minimax/MiniMax-M3", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M3", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "code-migration:cli:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.code-migration.cli", + "source_data": { + "dataset_name": "Code Migration - CLI", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cli", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (CLI).", + "metric_id": "vals_ai.code-migration.cli.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 11.934, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cli", + "task_name": "CLI", + "dataset_type": "private", + "industry": "coding", + "raw_score": "11.934", + "raw_stderr": "1.453", + "latency": "4906.501", + "cost_per_test": "8.42404", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.453, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "code-migration:cobol:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.code-migration.cobol", + "source_data": { + "dataset_name": "Code Migration - COBOL", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cobol", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (COBOL).", + "metric_id": "vals_ai.code-migration.cobol.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 43.896, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cobol", + "task_name": "COBOL", + "dataset_type": "private", + "industry": "coding", + "raw_score": "43.896", + "raw_stderr": "15.162", + "latency": "2993.546", + "cost_per_test": "3.018516", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 15.162, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "code-migration:code_quality:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.code-migration.code_quality", + "source_data": { + "dataset_name": "Code Migration - Code Quality", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "code_quality", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Code Quality).", + "metric_id": "vals_ai.code-migration.code_quality.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 3.718, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "code_quality", + "task_name": "Code Quality", + "dataset_type": "private", + "industry": "coding", + "raw_score": "3.718", + "raw_stderr": "0.083", + "latency": "4428.262", + "cost_per_test": "7.072659", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.083, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "code-migration:overall:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.code-migration.overall", + "source_data": { + "dataset_name": "Code Migration - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Overall).", + "metric_id": "vals_ai.code-migration.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 19.925, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "19.925", + "raw_stderr": "3.944", + "latency": "4428.262", + "cost_per_test": "7.072659", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 3.944, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M3/e7b9fe54-df0b-4b2d-a92b-728785531111.json b/data/vals_ai/minimax/MiniMax-M3/e7b9fe54-df0b-4b2d-a92b-728785531111.json new file mode 100644 index 0000000000000000000000000000000000000000..9bd5760888d6787a79a484a0e459553d402e301a --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M3/e7b9fe54-df0b-4b2d-a92b-728785531111.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/minimax_MiniMax-M3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M3", + "id": "minimax/MiniMax-M3", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M3", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 47.566, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "47.566", + "raw_stderr": "5.442", + "latency": "5073.23", + "cost_per_test": "6.453899", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 5.442, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M3/eaa498a8-321c-4166-b22f-acd839ceb016.json b/data/vals_ai/minimax/MiniMax-M3/eaa498a8-321c-4166-b22f-acd839ceb016.json new file mode 100644 index 0000000000000000000000000000000000000000..919c2351534041c25483541b2fbcc3a031d99c5a --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M3/eaa498a8-321c-4166-b22f-acd839ceb016.json @@ -0,0 +1,195 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/cyber/minimax_MiniMax-M3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CyberBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "beta", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M3", + "id": "minimax/MiniMax-M3", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M3", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "cyber:overall:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.cyber.overall", + "source_data": { + "dataset_name": "CyberBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (Overall).", + "metric_id": "vals_ai.cyber.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 76.271, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "beta", + "raw_score": "76.271", + "raw_stderr": "5.532", + "latency": "1433.642", + "cost_per_test": "1.738908", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 5.532, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "cyber:patch:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.cyber.patch", + "source_data": { + "dataset_name": "CyberBench - Patch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "patch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (Patch).", + "metric_id": "vals_ai.cyber.patch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 77.966, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "patch", + "task_name": "Patch", + "dataset_type": "private", + "industry": "beta", + "raw_score": "77.966", + "raw_stderr": "5.396", + "latency": "1401.021", + "cost_per_test": "0.818236", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 5.396, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "cyber:poc:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.cyber.poc", + "source_data": { + "dataset_name": "CyberBench - PoC", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "poc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (PoC).", + "metric_id": "vals_ai.cyber.poc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 74.576, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "poc", + "task_name": "PoC", + "dataset_type": "private", + "industry": "beta", + "raw_score": "74.576", + "raw_stderr": "5.669", + "latency": "1466.263", + "cost_per_test": "2.65958", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 5.669, + "method": "vals_reported" + } + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M3/edb32705-1205-4ed9-ab6d-f988f3fc4d80.json b/data/vals_ai/minimax/MiniMax-M3/edb32705-1205-4ed9-ab6d-f988f3fc4d80.json new file mode 100644 index 0000000000000000000000000000000000000000..7cb37b3c5ddd1a87d18274622392f73ff31be6dc --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M3/edb32705-1205-4ed9-ab6d-f988f3fc4d80.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/minimax_MiniMax-M3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M3", + "id": "minimax/MiniMax-M3", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M3", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 47.619, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "47.619", + "raw_stderr": "6.27", + "latency": "143.824", + "cost_per_test": "0.022988", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 6.27, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 37.302, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "37.302", + "raw_stderr": "6.222", + "latency": "115.085", + "cost_per_test": "0.020054", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 6.222, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 66.802, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "66.802", + "raw_stderr": "5.314", + "latency": "96.704", + "cost_per_test": "0.034489", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 5.314, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 50.574, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "50.574", + "raw_stderr": "3.436", + "latency": "118.538", + "cost_per_test": "0.025844", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 3.436, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M3/f527f3ef-892c-47bd-bd71-230c8c0a3e51.json b/data/vals_ai/minimax/MiniMax-M3/f527f3ef-892c-47bd-bd71-230c8c0a3e51.json new file mode 100644 index 0000000000000000000000000000000000000000..0522068623895d9a9fea39df076c242a50907651 --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M3/f527f3ef-892c-47bd-bd71-230c8c0a3e51.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/minimax_MiniMax-M3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M3", + "id": "minimax/MiniMax-M3", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M3", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 92.424, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.424", + "raw_stderr": "1.881", + "latency": "244.029", + "cost_per_test": "0.023", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.881, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 92.929, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.929", + "raw_stderr": "2.188", + "latency": "313.788", + "cost_per_test": "0.028001", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 2.188, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 92.676, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.676", + "raw_stderr": "1.443", + "latency": "278.908", + "cost_per_test": "0.0255", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.443, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/minimax/MiniMax-M3/fd7b33d9-a497-43bf-babf-1c6886cf5fd1.json b/data/vals_ai/minimax/MiniMax-M3/fd7b33d9-a497-43bf-babf-1c6886cf5fd1.json new file mode 100644 index 0000000000000000000000000000000000000000..929b6a68a1a0b5b0142c5d2c8f51185a3a317d8d --- /dev/null +++ b/data/vals_ai/minimax/MiniMax-M3/fd7b33d9-a497-43bf-babf-1c6886cf5fd1.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/minimax_MiniMax-M3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "MiniMax-M3", + "id": "minimax/MiniMax-M3", + "developer": "minimax", + "additional_details": { + "vals_model_id": "minimax/MiniMax-M3", + "vals_provider": "MiniMax" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 60.016, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.016", + "raw_stderr": "1.401", + "latency": "94.036", + "cost_per_test": "0.00728", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.401, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 72.731, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "72.731", + "raw_stderr": "0.863", + "latency": "94.036", + "cost_per_test": "0.00728", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 0.863, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:minimax/MiniMax-M3:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 85.446, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "85.446", + "raw_stderr": "1.008", + "latency": "94.036", + "cost_per_test": "0.00728", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "512000", + "provider": "MiniMax" + }, + "uncertainty": { + "standard_error": { + "value": 1.008, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 512000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/Mistral-7B-v0.1/3fd59607-9ad6-4a39-8b1f-91874707af0b.json b/data/vals_ai/mistralai/Mistral-7B-v0.1/3fd59607-9ad6-4a39-8b1f-91874707af0b.json new file mode 100644 index 0000000000000000000000000000000000000000..1f94c59988e26d23023bbf98a69c326395781ca5 --- /dev/null +++ b/data/vals_ai/mistralai/Mistral-7B-v0.1/3fd59607-9ad6-4a39-8b1f-91874707af0b.json @@ -0,0 +1,408 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/mistralai_Mistral-7B-v0.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Mistral-7B-v0.1", + "id": "mistralai/Mistral-7B-v0.1", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "together/mistralai/Mistral-7B-v0.1", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:together/mistralai/Mistral-7B-v0.1:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 61.629, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "61.629", + "raw_stderr": "1.08", + "latency": "0.527", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.08, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:together/mistralai/Mistral-7B-v0.1:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 49.631, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "49.631", + "raw_stderr": "0.109", + "latency": "1.919", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.109, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:together/mistralai/Mistral-7B-v0.1:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 51.677, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "51.677", + "raw_stderr": "0.176", + "latency": "3.546", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.176, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:together/mistralai/Mistral-7B-v0.1:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 53.77, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "53.77", + "raw_stderr": "0.779", + "latency": "1.715", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.779, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:together/mistralai/Mistral-7B-v0.1:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 51.087, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "51.087", + "raw_stderr": "1.364", + "latency": "1.699", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.364, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:together/mistralai/Mistral-7B-v0.1:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 54.826, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "54.826", + "raw_stderr": "3.481", + "latency": "0.882", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.481, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/Mixtral-8x7B-v0.1/f6623335-57b0-4b2d-bbf5-fd1dc3edd9d4.json b/data/vals_ai/mistralai/Mixtral-8x7B-v0.1/f6623335-57b0-4b2d-bbf5-fd1dc3edd9d4.json new file mode 100644 index 0000000000000000000000000000000000000000..8d89c74c44ee963cfcecbd102be796bb8dbac728 --- /dev/null +++ b/data/vals_ai/mistralai/Mixtral-8x7B-v0.1/f6623335-57b0-4b2d-bbf5-fd1dc3edd9d4.json @@ -0,0 +1,408 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/mistralai_Mixtral-8x7B-v0.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Mixtral-8x7B-v0.1", + "id": "mistralai/Mixtral-8x7B-v0.1", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "together/mistralai/Mixtral-8x7B-v0.1", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:together/mistralai/Mixtral-8x7B-v0.1:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 64.321, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "64.321", + "raw_stderr": "1.239", + "latency": "10.413", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.239, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:together/mistralai/Mixtral-8x7B-v0.1:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 52.164, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "52.164", + "raw_stderr": "0.12", + "latency": "12.31", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.12, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:together/mistralai/Mixtral-8x7B-v0.1:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 53.75, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "53.75", + "raw_stderr": "0.338", + "latency": "28.063", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.338, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:together/mistralai/Mixtral-8x7B-v0.1:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 55.836, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "55.836", + "raw_stderr": "0.742", + "latency": "14.142", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.742, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:together/mistralai/Mixtral-8x7B-v0.1:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 46.708, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "46.708", + "raw_stderr": "1.126", + "latency": "13.914", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.126, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:together/mistralai/Mixtral-8x7B-v0.1:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 62.236, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "62.236", + "raw_stderr": "3.293", + "latency": "6.012", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.293, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/devstral-2512/9c190984-aa9b-4711-b96f-a6673850f212.json b/data/vals_ai/mistralai/devstral-2512/9c190984-aa9b-4711-b96f-a6673850f212.json new file mode 100644 index 0000000000000000000000000000000000000000..58376a1511a9869e2de1a16039e710fb7fcebcdc --- /dev/null +++ b/data/vals_ai/mistralai/devstral-2512/9c190984-aa9b-4711-b96f-a6673850f212.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/mistralai_devstral-2512/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "devstral-2512", + "id": "mistralai/devstral-2512", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/devstral-2512", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:mistralai/devstral-2512:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "683.82", + "cost_per_test": "1.817543", + "temperature": "0.2", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.2, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:mistralai/devstral-2512:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 1.333, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "1.333", + "raw_stderr": "1.217", + "latency": "15405.47", + "cost_per_test": "1.689963", + "temperature": "0.2", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.217, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.2, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:mistralai/devstral-2512:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 0.666, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0.666", + "raw_stderr": "0.608", + "latency": "8044.645", + "cost_per_test": "1.753753", + "temperature": "0.2", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.608, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.2, + "max_tokens": 256000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/devstral-2512/b646df64-b98e-4f55-b6f1-7d69c39bb4eb.json b/data/vals_ai/mistralai/devstral-2512/b646df64-b98e-4f55-b6f1-7d69c39bb4eb.json new file mode 100644 index 0000000000000000000000000000000000000000..b5d2afa60a60631a03ad24a8eba136ad08f1bc1d --- /dev/null +++ b/data/vals_ai/mistralai/devstral-2512/b646df64-b98e-4f55-b6f1-7d69c39bb4eb.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/mistralai_devstral-2512/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "devstral-2512", + "id": "mistralai/devstral-2512", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/devstral-2512", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:mistralai/devstral-2512:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 90.683, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "90.683", + "raw_stderr": "1.62", + "latency": "19.735", + "cost_per_test": "0.004242", + "temperature": "0.2", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.62, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.2, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:mistralai/devstral-2512:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 13.143, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "13.143", + "raw_stderr": "1.806", + "latency": "159.536", + "cost_per_test": "0.111846", + "temperature": "0.2", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.806, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.2, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:mistralai/devstral-2512:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 51.697, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "51.697", + "raw_stderr": "2.553", + "latency": "83.438", + "cost_per_test": "0.053806", + "temperature": "0.2", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.553, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.2, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:mistralai/devstral-2512:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 51.841, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "51.841", + "raw_stderr": "1.174", + "latency": "87.57", + "cost_per_test": "0.056631", + "temperature": "0.2", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.174, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.2, + "max_tokens": 256000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/devstral-2512/d0c2d5eb-074b-4e24-9af9-0f1e66d811f1.json b/data/vals_ai/mistralai/devstral-2512/d0c2d5eb-074b-4e24-9af9-0f1e66d811f1.json new file mode 100644 index 0000000000000000000000000000000000000000..c401b6a6e22c172eb6746eea6ab8978e6e9a7720 --- /dev/null +++ b/data/vals_ai/mistralai/devstral-2512/d0c2d5eb-074b-4e24-9af9-0f1e66d811f1.json @@ -0,0 +1,361 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/mistralai_devstral-2512/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "devstral-2512", + "id": "mistralai/devstral-2512", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/devstral-2512", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:mistralai/devstral-2512:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 26.19, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "26.19", + "raw_stderr": "6.867", + "latency": "288.881", + "cost_per_test": "0.985003", + "temperature": "0.2", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 6.867, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.2, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:mistralai/devstral-2512:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 56.322, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "56.322", + "raw_stderr": "3.076", + "latency": "264.845", + "cost_per_test": "0.819155", + "temperature": "0.2", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.076, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.2, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:mistralai/devstral-2512:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 80.412, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "80.412", + "raw_stderr": "2.857", + "latency": "202.284", + "cost_per_test": "0.589267", + "temperature": "0.2", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.857, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.2, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:mistralai/devstral-2512:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "419.145", + "cost_per_test": "1.378311", + "temperature": "0.2", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.2, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:mistralai/devstral-2512:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 62.8, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "62.8", + "raw_stderr": "2.164", + "latency": "243.516", + "cost_per_test": "0.747244", + "temperature": "0.2", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.164, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.2, + "max_tokens": 256000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/labs-devstral-small-2512/6072e2f1-3383-4d90-af5d-0d5dd2d36270.json b/data/vals_ai/mistralai/labs-devstral-small-2512/6072e2f1-3383-4d90-af5d-0d5dd2d36270.json new file mode 100644 index 0000000000000000000000000000000000000000..210729b659ee2c62e13767c1c49caa84ad53ff21 --- /dev/null +++ b/data/vals_ai/mistralai/labs-devstral-small-2512/6072e2f1-3383-4d90-af5d-0d5dd2d36270.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/mistralai_labs-devstral-small-2512/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "labs-devstral-small-2512", + "id": "mistralai/labs-devstral-small-2512", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/labs-devstral-small-2512", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:mistralai/labs-devstral-small-2512:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 85.714, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "85.714", + "raw_stderr": "1.95", + "latency": "3.373", + "cost_per_test": "0.000106", + "temperature": "0.2", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.95, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.2, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:mistralai/labs-devstral-small-2512:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 8.571, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "8.571", + "raw_stderr": "1.496", + "latency": "38.334", + "cost_per_test": "0.000622", + "temperature": "0.2", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.496, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.2, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:mistralai/labs-devstral-small-2512:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 35.248, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "35.248", + "raw_stderr": "2.441", + "latency": "14.756", + "cost_per_test": "0.000311", + "temperature": "0.2", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.441, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.2, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:mistralai/labs-devstral-small-2512:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 43.178, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "43.178", + "raw_stderr": "1.155", + "latency": "18.821", + "cost_per_test": "0.000346", + "temperature": "0.2", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.155, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.2, + "max_tokens": 256000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/labs-devstral-small-2512/f2e21a39-d1f1-4d72-8fc9-54899df286a9.json b/data/vals_ai/mistralai/labs-devstral-small-2512/f2e21a39-d1f1-4d72-8fc9-54899df286a9.json new file mode 100644 index 0000000000000000000000000000000000000000..6596050553c1fe33d2e29c5dc93d1f2967eef122 --- /dev/null +++ b/data/vals_ai/mistralai/labs-devstral-small-2512/f2e21a39-d1f1-4d72-8fc9-54899df286a9.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/mistralai_labs-devstral-small-2512/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "labs-devstral-small-2512", + "id": "mistralai/labs-devstral-small-2512", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/labs-devstral-small-2512", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:mistralai/labs-devstral-small-2512:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 3.667, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "3.667", + "raw_stderr": "3.347", + "latency": "858.154", + "cost_per_test": "0.513867", + "temperature": "0.2", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.347, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.2, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:mistralai/labs-devstral-small-2512:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 1.333, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "1.333", + "raw_stderr": "1.217", + "latency": "14075.052", + "cost_per_test": "0.497513", + "temperature": "0.2", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.217, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.2, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:mistralai/labs-devstral-small-2512:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 2.5, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "2.5", + "raw_stderr": "1.781", + "latency": "7466.603", + "cost_per_test": "0.50569", + "temperature": "0.2", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.781, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.2, + "max_tokens": 256000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/magistral-medium-2509/0da160e9-d2b8-4890-9ec5-537231f29d9b.json b/data/vals_ai/mistralai/magistral-medium-2509/0da160e9-d2b8-4890-9ec5-537231f29d9b.json new file mode 100644 index 0000000000000000000000000000000000000000..bf9a983d1b24e8ba552dd73a77f8030b77134ed7 --- /dev/null +++ b/data/vals_ai/mistralai/magistral-medium-2509/0da160e9-d2b8-4890-9ec5-537231f29d9b.json @@ -0,0 +1,101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/mistralai_magistral-medium-2509/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "magistral-medium-2509", + "id": "mistralai/magistral-medium-2509", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/magistral-medium-2509", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 64.566, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "64.566", + "raw_stderr": "1.15", + "latency": "42.144", + "cost_per_test": "0.021457", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.15, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/magistral-medium-2509/2f10d429-d535-47e0-80f2-034d68a73128.json b/data/vals_ai/mistralai/magistral-medium-2509/2f10d429-d535-47e0-80f2-034d68a73128.json new file mode 100644 index 0000000000000000000000000000000000000000..ca2a986d96d20153336c5cc075c8036c7b73a719 --- /dev/null +++ b/data/vals_ai/mistralai/magistral-medium-2509/2f10d429-d535-47e0-80f2-034d68a73128.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/mistralai_magistral-medium-2509/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "magistral-medium-2509", + "id": "mistralai/magistral-medium-2509", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/magistral-medium-2509", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.946, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.946", + "raw_stderr": "1.013", + "latency": "11.895", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.013, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 52.125, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "52.125", + "raw_stderr": "0.309", + "latency": "16.249", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.309, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 40.342, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "40.342", + "raw_stderr": "0.895", + "latency": "12.951", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.895, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 54.767, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "54.767", + "raw_stderr": "0.561", + "latency": "11.734", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.561, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 43.556, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "43.556", + "raw_stderr": "1.499", + "latency": "8.779", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.499, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 55.866, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "55.866", + "raw_stderr": "1.921", + "latency": "8.795", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.921, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/magistral-medium-2509/3069be58-e3a7-4f88-adb3-5be79da93e1e.json b/data/vals_ai/mistralai/magistral-medium-2509/3069be58-e3a7-4f88-adb3-5be79da93e1e.json new file mode 100644 index 0000000000000000000000000000000000000000..9b1a6c24c536ed5842b800376a5f38fa7e291436 --- /dev/null +++ b/data/vals_ai/mistralai/magistral-medium-2509/3069be58-e3a7-4f88-adb3-5be79da93e1e.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/mistralai_magistral-medium-2509/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "magistral-medium-2509", + "id": "mistralai/magistral-medium-2509", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/magistral-medium-2509", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 55.485, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "55.485", + "raw_stderr": "1.401", + "latency": "33.584", + "cost_per_test": "0.022751", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.401, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 51.868, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "51.868", + "raw_stderr": "0.993", + "latency": "25.948", + "cost_per_test": "0.019966", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.993, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 48.251, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "48.251", + "raw_stderr": "1.409", + "latency": "18.311", + "cost_per_test": "0.017181", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.409, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/magistral-medium-2509/b5ddb7ef-c151-4607-9489-5448866dec74.json b/data/vals_ai/mistralai/magistral-medium-2509/b5ddb7ef-c151-4607-9489-5448866dec74.json new file mode 100644 index 0000000000000000000000000000000000000000..3e3da5f7e3977f956aa03630f3aa2d7982bb04c2 --- /dev/null +++ b/data/vals_ai/mistralai/magistral-medium-2509/b5ddb7ef-c151-4607-9489-5448866dec74.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/mistralai_magistral-medium-2509/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "magistral-medium-2509", + "id": "mistralai/magistral-medium-2509", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/magistral-medium-2509", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.758, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.758", + "raw_stderr": "0.617", + "latency": "54.06", + "cost_per_test": "0.020032", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.617, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 45.143, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "45.143", + "raw_stderr": "2.66", + "latency": "374.337", + "cost_per_test": "0.081503", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.66, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 80.679, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "80.679", + "raw_stderr": "2.017", + "latency": "201.514", + "cost_per_test": "0.048182", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.017, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 74.86, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "74.86", + "raw_stderr": "1.132", + "latency": "209.97", + "cost_per_test": "0.049906", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.132, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/magistral-medium-2509/bd54b042-45df-4823-9b49-896cd9129ec8.json b/data/vals_ai/mistralai/magistral-medium-2509/bd54b042-45df-4823-9b49-896cd9129ec8.json new file mode 100644 index 0000000000000000000000000000000000000000..cd3ae9948e0ebbec9887975666c4accae77db069 --- /dev/null +++ b/data/vals_ai/mistralai/magistral-medium-2509/bd54b042-45df-4823-9b49-896cd9129ec8.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/mistralai_magistral-medium-2509/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "magistral-medium-2509", + "id": "mistralai/magistral-medium-2509", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/magistral-medium-2509", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 51.981, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "51.981", + "raw_stderr": "1.706", + "latency": "16.96", + "cost_per_test": "0.011202", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.706, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 43.007, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "43.007", + "raw_stderr": "1.69", + "latency": "96.844", + "cost_per_test": "0.216228", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.69, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 47.397, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "47.397", + "raw_stderr": "0.982", + "latency": "55.265", + "cost_per_test": "0.125244", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.982, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 47.203, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "47.203", + "raw_stderr": "1.704", + "latency": "51.99", + "cost_per_test": "0.148302", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.704, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/magistral-medium-2509/be9b2e82-674e-490f-a5ff-04bd16d71946.json b/data/vals_ai/mistralai/magistral-medium-2509/be9b2e82-674e-490f-a5ff-04bd16d71946.json new file mode 100644 index 0000000000000000000000000000000000000000..7ea6b6d5fa4bc6b6f6e85ce25c1aa9fac6644431 --- /dev/null +++ b/data/vals_ai/mistralai/magistral-medium-2509/be9b2e82-674e-490f-a5ff-04bd16d71946.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/mistralai_magistral-medium-2509/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "magistral-medium-2509", + "id": "mistralai/magistral-medium-2509", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/magistral-medium-2509", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 52.821, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.821", + "raw_stderr": "1.427", + "latency": "46.254", + "cost_per_test": "0.025214", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.427, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 61.938, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.938", + "raw_stderr": "0.964", + "latency": "46.254", + "cost_per_test": "0.025214", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.964, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 71.055, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.055", + "raw_stderr": "1.297", + "latency": "46.254", + "cost_per_test": "0.025214", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.297, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/magistral-medium-2509/cfe9dedc-b3c0-4e8e-909b-98497f4074a6.json b/data/vals_ai/mistralai/magistral-medium-2509/cfe9dedc-b3c0-4e8e-909b-98497f4074a6.json new file mode 100644 index 0000000000000000000000000000000000000000..49a339d43794ce565d3fd7fe4a8e4e6bb34d54df --- /dev/null +++ b/data/vals_ai/mistralai/magistral-medium-2509/cfe9dedc-b3c0-4e8e-909b-98497f4074a6.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/mistralai_magistral-medium-2509/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "magistral-medium-2509", + "id": "mistralai/magistral-medium-2509", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/magistral-medium-2509", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 0.5, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0.5", + "raw_stderr": "0.456", + "latency": "2735.47", + "cost_per_test": "8.012136", + "temperature": "0.7", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.456, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 0.833, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0.833", + "raw_stderr": "0.761", + "latency": "1688.096", + "cost_per_test": "4.676308", + "temperature": "0.7", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.761, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 0.666, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0.666", + "raw_stderr": "0.444", + "latency": "2211.783", + "cost_per_test": "6.344222", + "temperature": "0.7", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.444, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/magistral-medium-2509/e9a6013a-9477-4cb7-9c6a-d6ba609a0476.json b/data/vals_ai/mistralai/magistral-medium-2509/e9a6013a-9477-4cb7-9c6a-d6ba609a0476.json new file mode 100644 index 0000000000000000000000000000000000000000..684da5f2298593e75cba4c1803ed69ef12e04685 --- /dev/null +++ b/data/vals_ai/mistralai/magistral-medium-2509/e9a6013a-9477-4cb7-9c6a-d6ba609a0476.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/mistralai_magistral-medium-2509/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "magistral-medium-2509", + "id": "mistralai/magistral-medium-2509", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/magistral-medium-2509", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 62.626, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "62.626", + "raw_stderr": "3.523", + "latency": "83.642", + "cost_per_test": "0.033256", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.523, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 62.121, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "62.121", + "raw_stderr": "3.363", + "latency": "104.648", + "cost_per_test": "0.038764", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.363, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 62.374, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "62.374", + "raw_stderr": "2.435", + "latency": "94.145", + "cost_per_test": "0.03601", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.435, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/magistral-medium-2509/f2000b3e-872a-48ec-8587-1dd72429063c.json b/data/vals_ai/mistralai/magistral-medium-2509/f2000b3e-872a-48ec-8587-1dd72429063c.json new file mode 100644 index 0000000000000000000000000000000000000000..82a4a58cd769b920b943dad47c343cd244d641dd --- /dev/null +++ b/data/vals_ai/mistralai/magistral-medium-2509/f2000b3e-872a-48ec-8587-1dd72429063c.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/mistralai_magistral-medium-2509/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "magistral-medium-2509", + "id": "mistralai/magistral-medium-2509", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/magistral-medium-2509", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.24, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.24", + "raw_stderr": "1.366", + "latency": "18.163", + "cost_per_test": "0.010747", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.366, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.709, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.709", + "raw_stderr": "1.213", + "latency": "44.91", + "cost_per_test": "0.01941", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.213, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 40.283, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "40.283", + "raw_stderr": "1.458", + "latency": "36.238", + "cost_per_test": "0.018854", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.458, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 70.0, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70", + "raw_stderr": "2.263", + "latency": "27.638", + "cost_per_test": "0.014657", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.263, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.398, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.398", + "raw_stderr": "1.343", + "latency": "43.071", + "cost_per_test": "0.012351", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.343, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 56.45, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "56.45", + "raw_stderr": "1.593", + "latency": "64.176", + "cost_per_test": "0.030606", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.593, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.817, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.817", + "raw_stderr": "1.518", + "latency": "16.969", + "cost_per_test": "0.008671", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.518, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 65.092, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "65.092", + "raw_stderr": "2.442", + "latency": "18.999", + "cost_per_test": "0.011342", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.442, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 56.494, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "56.494", + "raw_stderr": "1.494", + "latency": "37.236", + "cost_per_test": "0.018636", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.494, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 49.741, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "49.741", + "raw_stderr": "1.36", + "latency": "33.17", + "cost_per_test": "0.015989", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.36, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.082, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.082", + "raw_stderr": "1.403", + "latency": "29.004", + "cost_per_test": "0.009138", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.403, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 68.659, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "68.659", + "raw_stderr": "0.433", + "latency": "31.988", + "cost_per_test": "0.014664", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.433, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 70.14, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.14", + "raw_stderr": "2.049", + "latency": "22.523", + "cost_per_test": "0.009835", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.049, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.828, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.828", + "raw_stderr": "1.188", + "latency": "38.255", + "cost_per_test": "0.016465", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.188, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:mistralai/magistral-medium-2509:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.952, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.952", + "raw_stderr": "1.397", + "latency": "17.485", + "cost_per_test": "0.008589", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.397, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/magistral-small-2509/066d8e98-ac1d-430a-94e6-c591192d2084.json b/data/vals_ai/mistralai/magistral-small-2509/066d8e98-ac1d-430a-94e6-c591192d2084.json new file mode 100644 index 0000000000000000000000000000000000000000..7f9b06025ee934c7f677917f46bdd3ecd1ee9bf4 --- /dev/null +++ b/data/vals_ai/mistralai/magistral-small-2509/066d8e98-ac1d-430a-94e6-c591192d2084.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/mistralai_magistral-small-2509/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "magistral-small-2509", + "id": "mistralai/magistral-small-2509", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/magistral-small-2509", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 76.791, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "76.791", + "raw_stderr": "0.911", + "latency": "5.94", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.911, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 31.417, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "31.417", + "raw_stderr": "0.196", + "latency": "6.782", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.196, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 19.942, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "19.942", + "raw_stderr": "0.681", + "latency": "6.31", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.681, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 40.023, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "40.023", + "raw_stderr": "0.472", + "latency": "6.67", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.472, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 27.28, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "27.28", + "raw_stderr": "0.755", + "latency": "3.618", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.755, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 44.684, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "44.684", + "raw_stderr": "1.913", + "latency": "10.701", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.913, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/magistral-small-2509/38924c88-8841-4740-b1cf-52d256e0df4f.json b/data/vals_ai/mistralai/magistral-small-2509/38924c88-8841-4740-b1cf-52d256e0df4f.json new file mode 100644 index 0000000000000000000000000000000000000000..cb38dee9a06f56bdbf22878cb0ff18ab4106c5d4 --- /dev/null +++ b/data/vals_ai/mistralai/magistral-small-2509/38924c88-8841-4740-b1cf-52d256e0df4f.json @@ -0,0 +1,101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/mistralai_magistral-small-2509/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "magistral-small-2509", + "id": "mistralai/magistral-small-2509", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/magistral-small-2509", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 65.202, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "65.202", + "raw_stderr": "1.145", + "latency": "19.095", + "cost_per_test": "0.004413", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.145, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/magistral-small-2509/3cc2b5b1-a7d6-48cd-9745-6e862dc554d1.json b/data/vals_ai/mistralai/magistral-small-2509/3cc2b5b1-a7d6-48cd-9745-6e862dc554d1.json new file mode 100644 index 0000000000000000000000000000000000000000..79a10e71192f56ec5de86db3d5d225ef0cb0816c --- /dev/null +++ b/data/vals_ai/mistralai/magistral-small-2509/3cc2b5b1-a7d6-48cd-9745-6e862dc554d1.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/mistralai_magistral-small-2509/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "magistral-small-2509", + "id": "mistralai/magistral-small-2509", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/magistral-small-2509", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 70.588, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "70.588", + "raw_stderr": "1.285", + "latency": "14.218", + "cost_per_test": "0.005141", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.285, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 62.122, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.122", + "raw_stderr": "0.952", + "latency": "10.648", + "cost_per_test": "0.004484", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.952, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 53.657, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "53.657", + "raw_stderr": "1.406", + "latency": "7.079", + "cost_per_test": "0.003826", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.406, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/magistral-small-2509/8bc14d36-8225-4ce5-8c7a-d9aaf6347b50.json b/data/vals_ai/mistralai/magistral-small-2509/8bc14d36-8225-4ce5-8c7a-d9aaf6347b50.json new file mode 100644 index 0000000000000000000000000000000000000000..de554222b34db76205fd335df6bc82232701a82b --- /dev/null +++ b/data/vals_ai/mistralai/magistral-small-2509/8bc14d36-8225-4ce5-8c7a-d9aaf6347b50.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/mistralai_magistral-small-2509/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "magistral-small-2509", + "id": "mistralai/magistral-small-2509", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/magistral-small-2509", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 50.041, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "50.041", + "raw_stderr": "1.43", + "latency": "16.441", + "cost_per_test": "0.003909", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.43, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 60.302, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.302", + "raw_stderr": "0.967", + "latency": "16.441", + "cost_per_test": "0.003909", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.967, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 70.564, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "70.564", + "raw_stderr": "1.303", + "latency": "16.441", + "cost_per_test": "0.003909", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.303, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/magistral-small-2509/92d13ab8-1a83-47bf-86c6-d5b3f7e4c646.json b/data/vals_ai/mistralai/magistral-small-2509/92d13ab8-1a83-47bf-86c6-d5b3f7e4c646.json new file mode 100644 index 0000000000000000000000000000000000000000..1b70bdc9b7eabb2bb4afdbf987cefbe0b9e92398 --- /dev/null +++ b/data/vals_ai/mistralai/magistral-small-2509/92d13ab8-1a83-47bf-86c6-d5b3f7e4c646.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/mistralai_magistral-small-2509/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "magistral-small-2509", + "id": "mistralai/magistral-small-2509", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/magistral-small-2509", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 55.478, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "55.478", + "raw_stderr": "1.697", + "latency": "8.521", + "cost_per_test": "0.002287", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.697, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 36.597, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "36.597", + "raw_stderr": "1.644", + "latency": "18.965", + "cost_per_test": "0.053106", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.644, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 44.017, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "44.017", + "raw_stderr": "0.965", + "latency": "14.053", + "cost_per_test": "0.030437", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.965, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 39.977, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "39.977", + "raw_stderr": "1.672", + "latency": "14.674", + "cost_per_test": "0.035917", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.672, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/magistral-small-2509/94b2d939-17d1-4d1f-81b9-d39bfdaf7bc3.json b/data/vals_ai/mistralai/magistral-small-2509/94b2d939-17d1-4d1f-81b9-d39bfdaf7bc3.json new file mode 100644 index 0000000000000000000000000000000000000000..e4140e351c93f043781d0743ffbe1f828590f9e9 --- /dev/null +++ b/data/vals_ai/mistralai/magistral-small-2509/94b2d939-17d1-4d1f-81b9-d39bfdaf7bc3.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/mistralai_magistral-small-2509/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "magistral-small-2509", + "id": "mistralai/magistral-small-2509", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/magistral-small-2509", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.658, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.658", + "raw_stderr": "1.346", + "latency": "13.848", + "cost_per_test": "0.00293", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.346, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.539, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.539", + "raw_stderr": "1.53", + "latency": "11.181", + "cost_per_test": "0.003029", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.53, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 66.608, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.608", + "raw_stderr": "1.402", + "latency": "22.217", + "cost_per_test": "0.005379", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.402, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.39, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.39", + "raw_stderr": "2.156", + "latency": "13.435", + "cost_per_test": "0.003485", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.156, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.066, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.066", + "raw_stderr": "1.469", + "latency": "11.336", + "cost_per_test": "0.00282", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.469, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 52.735, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "52.735", + "raw_stderr": "1.604", + "latency": "31.202", + "cost_per_test": "0.00718", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.604, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 54.279, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "54.279", + "raw_stderr": "1.742", + "latency": "8.939", + "cost_per_test": "0.002333", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.742, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 58.005, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "58.005", + "raw_stderr": "2.529", + "latency": "8.725", + "cost_per_test": "0.002683", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.529, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 28.247, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "28.247", + "raw_stderr": "1.357", + "latency": "17.029", + "cost_per_test": "0.003685", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.357, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 39.97, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "39.97", + "raw_stderr": "1.333", + "latency": "33.945", + "cost_per_test": "0.00496", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 54.545, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "54.545", + "raw_stderr": "1.638", + "latency": "9.55", + "cost_per_test": "0.002112", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.638, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 62.135, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "62.135", + "raw_stderr": "0.452", + "latency": "16.124", + "cost_per_test": "0.00355", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.452, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 57.515, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "57.515", + "raw_stderr": "2.213", + "latency": "10.587", + "cost_per_test": "0.002458", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.213, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.905, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.905", + "raw_stderr": "1.187", + "latency": "25.587", + "cost_per_test": "0.004659", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.187, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.429, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.429", + "raw_stderr": "1.599", + "latency": "8.153", + "cost_per_test": "0.001986", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.599, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/magistral-small-2509/b062e004-d479-4ebf-a0d8-5adb09a7dcb8.json b/data/vals_ai/mistralai/magistral-small-2509/b062e004-d479-4ebf-a0d8-5adb09a7dcb8.json new file mode 100644 index 0000000000000000000000000000000000000000..73d62b83764fd030b63af06c044de3549bacb549 --- /dev/null +++ b/data/vals_ai/mistralai/magistral-small-2509/b062e004-d479-4ebf-a0d8-5adb09a7dcb8.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/mistralai_magistral-small-2509/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "magistral-small-2509", + "id": "mistralai/magistral-small-2509", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/magistral-small-2509", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 60.101, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "60.101", + "raw_stderr": "3.494", + "latency": "45.93", + "cost_per_test": "0.006825", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.494, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 56.566, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "56.566", + "raw_stderr": "3.539", + "latency": "46.573", + "cost_per_test": "0.007291", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.539, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 58.334, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "58.334", + "raw_stderr": "2.487", + "latency": "46.252", + "cost_per_test": "0.007058", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.487, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/magistral-small-2509/eb8e3aa4-0ef8-4a37-8c8c-635e60b7bb11.json b/data/vals_ai/mistralai/magistral-small-2509/eb8e3aa4-0ef8-4a37-8c8c-635e60b7bb11.json new file mode 100644 index 0000000000000000000000000000000000000000..f4364da59883e5c0b0f9a0768f7af0275541cce9 --- /dev/null +++ b/data/vals_ai/mistralai/magistral-small-2509/eb8e3aa4-0ef8-4a37-8c8c-635e60b7bb11.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/mistralai_magistral-small-2509/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "magistral-small-2509", + "id": "mistralai/magistral-small-2509", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/magistral-small-2509", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.758, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.758", + "raw_stderr": "0.617", + "latency": "29.279", + "cost_per_test": "0.005942", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.617, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 38.0, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "38", + "raw_stderr": "2.594", + "latency": "172.409", + "cost_per_test": "0.024321", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.594, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 79.634, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "79.634", + "raw_stderr": "2.058", + "latency": "100.657", + "cost_per_test": "0.015391", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.058, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:mistralai/magistral-small-2509:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 72.131, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "72.131", + "raw_stderr": "1.123", + "latency": "100.782", + "cost_per_test": "0.015218", + "temperature": "0.7", + "max_output_tokens": "128000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.123, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-large-2411/25a44af0-4d9f-43c0-97c2-ff7d069aa5d7.json b/data/vals_ai/mistralai/mistral-large-2411/25a44af0-4d9f-43c0-97c2-ff7d069aa5d7.json new file mode 100644 index 0000000000000000000000000000000000000000..0bff8caa90fdc1969098ce9cfe0f07a221332899 --- /dev/null +++ b/data/vals_ai/mistralai/mistral-large-2411/25a44af0-4d9f-43c0-97c2-ff7d069aa5d7.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/mistralai_mistral-large-2411/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-large-2411", + "id": "mistralai/mistral-large-2411", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-large-2411", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:mistralai/mistral-large-2411:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 81.677, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "81.677", + "raw_stderr": "2.156", + "latency": "3.616", + "cost_per_test": "0.002239", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.156, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:mistralai/mistral-large-2411:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 4.0, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "4", + "raw_stderr": "1.047", + "latency": "6.776", + "cost_per_test": "0.004168", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.047, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:mistralai/mistral-large-2411:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 25.587, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "25.587", + "raw_stderr": "2.23", + "latency": "5.654", + "cost_per_test": "0.003466", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.23, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:mistralai/mistral-large-2411:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 37.088, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "37.088", + "raw_stderr": "1.091", + "latency": "5.349", + "cost_per_test": "0.003291", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.091, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-large-2411/9d52ce9d-4d21-4823-9111-3fed2c3b5c73.json b/data/vals_ai/mistralai/mistral-large-2411/9d52ce9d-4d21-4823-9111-3fed2c3b5c73.json new file mode 100644 index 0000000000000000000000000000000000000000..b2c670cbba63afae742f04a35d7b1c9aa8b26bc7 --- /dev/null +++ b/data/vals_ai/mistralai/mistral-large-2411/9d52ce9d-4d21-4823-9111-3fed2c3b5c73.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/mistralai_mistral-large-2411/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-large-2411", + "id": "mistralai/mistral-large-2411", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-large-2411", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:mistralai/mistral-large-2411:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 46.97, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "46.97", + "raw_stderr": "3.547", + "latency": "16.517", + "cost_per_test": "0.008783", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.547, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:mistralai/mistral-large-2411:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 48.485, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "48.485", + "raw_stderr": "3.523", + "latency": "15.087", + "cost_per_test": "0.004644", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.523, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:mistralai/mistral-large-2411:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 47.728, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "47.728", + "raw_stderr": "2.5", + "latency": "15.802", + "cost_per_test": "0.006714", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.5, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-large-2411/c1375fd8-c3dc-4003-a48a-55710f7eefd5.json b/data/vals_ai/mistralai/mistral-large-2411/c1375fd8-c3dc-4003-a48a-55710f7eefd5.json new file mode 100644 index 0000000000000000000000000000000000000000..5faf63e75be903c55b2c2880957fa4b73b233a64 --- /dev/null +++ b/data/vals_ai/mistralai/mistral-large-2411/c1375fd8-c3dc-4003-a48a-55710f7eefd5.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/mistralai_mistral-large-2411/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-large-2411", + "id": "mistralai/mistral-large-2411", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-large-2411", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:mistralai/mistral-large-2411:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 53.639, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "53.639", + "raw_stderr": "1.426", + "latency": "12.019", + "cost_per_test": "0.004079", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.426, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:mistralai/mistral-large-2411:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 63.778, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.778", + "raw_stderr": "0.95", + "latency": "12.019", + "cost_per_test": "0.004079", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.95, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:mistralai/mistral-large-2411:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 73.917, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "73.917", + "raw_stderr": "1.256", + "latency": "12.019", + "cost_per_test": "0.004079", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.256, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-large-2411/dc92baa0-1501-4f9d-99ee-3ba0f066ebc9.json b/data/vals_ai/mistralai/mistral-large-2411/dc92baa0-1501-4f9d-99ee-3ba0f066ebc9.json new file mode 100644 index 0000000000000000000000000000000000000000..7eb5a29605ffe77db62919d93c198d2afaed129a --- /dev/null +++ b/data/vals_ai/mistralai/mistral-large-2411/dc92baa0-1501-4f9d-99ee-3ba0f066ebc9.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/mistralai_mistral-large-2411/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-large-2411", + "id": "mistralai/mistral-large-2411", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-large-2411", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:mistralai/mistral-large-2411:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.821, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.821", + "raw_stderr": "1.375", + "latency": "5.347", + "cost_per_test": "0.004702", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.375, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:mistralai/mistral-large-2411:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.736, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.736", + "raw_stderr": "1.603", + "latency": "7.912", + "cost_per_test": "0.004557", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.603, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:mistralai/mistral-large-2411:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.113, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.113", + "raw_stderr": "1.347", + "latency": "12.738", + "cost_per_test": "0.007141", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.347, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:mistralai/mistral-large-2411:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.61, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.61", + "raw_stderr": "2.121", + "latency": "7.355", + "cost_per_test": "0.005473", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.121, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:mistralai/mistral-large-2411:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.066, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.066", + "raw_stderr": "1.469", + "latency": "5.326", + "cost_per_test": "0.004166", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.469, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:mistralai/mistral-large-2411:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 52.219, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "52.219", + "raw_stderr": "1.605", + "latency": "16.566", + "cost_per_test": "0.007721", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.605, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:mistralai/mistral-large-2411:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 72.861, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.861", + "raw_stderr": "1.555", + "latency": "4.163", + "cost_per_test": "0.003581", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.555, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:mistralai/mistral-large-2411:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 64.042, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "64.042", + "raw_stderr": "2.458", + "latency": "3.039", + "cost_per_test": "0.005579", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.458, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:mistralai/mistral-large-2411:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 49.046, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "49.046", + "raw_stderr": "1.507", + "latency": "4.608", + "cost_per_test": "0.004517", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.507, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:mistralai/mistral-large-2411:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.722, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.722", + "raw_stderr": "1.167", + "latency": "10.179", + "cost_per_test": "0.006698", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.167, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:mistralai/mistral-large-2411:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 70.671, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.671", + "raw_stderr": "1.498", + "latency": "3.589", + "cost_per_test": "0.003046", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.498, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:mistralai/mistral-large-2411:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 69.706, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.706", + "raw_stderr": "0.442", + "latency": "7.192", + "cost_per_test": "0.004997", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.442, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:mistralai/mistral-large-2411:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 63.727, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "63.727", + "raw_stderr": "2.152", + "latency": "4.929", + "cost_per_test": "0.003555", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.152, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:mistralai/mistral-large-2411:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 72.055, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.055", + "raw_stderr": "1.245", + "latency": "10.883", + "cost_per_test": "0.005755", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.245, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:mistralai/mistral-large-2411:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.193, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.193", + "raw_stderr": "1.485", + "latency": "4.051", + "cost_per_test": "0.003469", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.485, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-large-2512/01a4b6e2-ade8-4c54-a7e9-1cc1c03499c7.json b/data/vals_ai/mistralai/mistral-large-2512/01a4b6e2-ade8-4c54-a7e9-1cc1c03499c7.json new file mode 100644 index 0000000000000000000000000000000000000000..418f3069ea3031fef2b4af02d10ba7fcfd488d24 --- /dev/null +++ b/data/vals_ai/mistralai/mistral-large-2512/01a4b6e2-ade8-4c54-a7e9-1cc1c03499c7.json @@ -0,0 +1,426 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/mistralai_mistral-large-2512/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-large-2512", + "id": "mistralai/mistral-large-2512", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-large-2512", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.032, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.032", + "raw_stderr": "1.001", + "latency": "0.793", + "cost_per_test": "0.000225", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.001, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 76.594, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "76.594", + "raw_stderr": "0.207", + "latency": "0.889", + "cost_per_test": "0.000401", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.207, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.006, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.006", + "raw_stderr": "0.694", + "latency": "0.938", + "cost_per_test": "0.0008", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.694, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.138, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.138", + "raw_stderr": "0.46", + "latency": "1.731", + "cost_per_test": "0.000439", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.46, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.926, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.926", + "raw_stderr": "1.238", + "latency": "0.886", + "cost_per_test": "0.000427", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.238, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.13, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.13", + "raw_stderr": "1.496", + "latency": "5.148", + "cost_per_test": "0.000344", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.496, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-large-2512/107399f1-7905-4007-b7ae-99590afaa9cd.json b/data/vals_ai/mistralai/mistral-large-2512/107399f1-7905-4007-b7ae-99590afaa9cd.json new file mode 100644 index 0000000000000000000000000000000000000000..11f1a80fe643d995d6bdc9e4253fd02124be35f8 --- /dev/null +++ b/data/vals_ai/mistralai/mistral-large-2512/107399f1-7905-4007-b7ae-99590afaa9cd.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/mistralai_mistral-large-2512/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-large-2512", + "id": "mistralai/mistral-large-2512", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-large-2512", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 50.0, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "50", + "raw_stderr": "1.41", + "latency": "9.865", + "cost_per_test": "0.003256", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.41, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 52.106, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.106", + "raw_stderr": "0.995", + "latency": "9.292", + "cost_per_test": "0.003226", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.995, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 54.213, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "54.213", + "raw_stderr": "1.405", + "latency": "8.719", + "cost_per_test": "0.003196", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.405, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-large-2512/118c916c-9995-4a4b-97a1-a2e89e29eee2.json b/data/vals_ai/mistralai/mistral-large-2512/118c916c-9995-4a4b-97a1-a2e89e29eee2.json new file mode 100644 index 0000000000000000000000000000000000000000..e167101e6634c1f6cc6f0946546b3c91a0a01aa2 --- /dev/null +++ b/data/vals_ai/mistralai/mistral-large-2512/118c916c-9995-4a4b-97a1-a2e89e29eee2.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/mistralai_mistral-large-2512/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-large-2512", + "id": "mistralai/mistral-large-2512", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-large-2512", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.563, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.563", + "raw_stderr": "1.865", + "latency": "12.468", + "cost_per_test": "0.001387", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.865, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.664, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.664", + "raw_stderr": "1.779", + "latency": "11.164", + "cost_per_test": "0.001253", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.779, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.982, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.982", + "raw_stderr": "1.375", + "latency": "15.68", + "cost_per_test": "0.001859", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.375, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.171, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.171", + "raw_stderr": "2.247", + "latency": "11.872", + "cost_per_test": "0.00154", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.247, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.782, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.782", + "raw_stderr": "1.662", + "latency": "11.545", + "cost_per_test": "0.001255", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.662, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.923, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.923", + "raw_stderr": "1.595", + "latency": "26.843", + "cost_per_test": "0.002489", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.595, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.262, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.262", + "raw_stderr": "1.738", + "latency": "11.318", + "cost_per_test": "0.00113", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.738, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 70.866, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.866", + "raw_stderr": "2.555", + "latency": "11.778", + "cost_per_test": "0.001663", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.555, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 56.403, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "56.403", + "raw_stderr": "1.306", + "latency": "17.014", + "cost_per_test": "0.001604", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.306, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.379, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.379", + "raw_stderr": "1.232", + "latency": "15.539", + "cost_per_test": "0.001912", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.232, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.221, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.221", + "raw_stderr": "1.632", + "latency": "8.956", + "cost_per_test": "0.000929", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.632, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.823, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.823", + "raw_stderr": "0.472", + "latency": "13.554", + "cost_per_test": "0.001481", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.472, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.353, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.353", + "raw_stderr": "2.169", + "latency": "9.977", + "cost_per_test": "0.001026", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.169, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.374, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.374", + "raw_stderr": "1.213", + "latency": "14.859", + "cost_per_test": "0.001605", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.213, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.576, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.576", + "raw_stderr": "1.754", + "latency": "10.746", + "cost_per_test": "0.001079", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.754, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-large-2512/1f2ea130-1db9-4236-8819-71f0303596c4.json b/data/vals_ai/mistralai/mistral-large-2512/1f2ea130-1db9-4236-8819-71f0303596c4.json new file mode 100644 index 0000000000000000000000000000000000000000..8d08f7b57f3143195ddf9cfa4beef52ce5cde60d --- /dev/null +++ b/data/vals_ai/mistralai/mistral-large-2512/1f2ea130-1db9-4236-8819-71f0303596c4.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/mistralai_mistral-large-2512/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-large-2512", + "id": "mistralai/mistral-large-2512", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-large-2512", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 61.57, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.57", + "raw_stderr": "1.391", + "latency": "29.531", + "cost_per_test": "0.001601", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.391, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 73.058, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "73.058", + "raw_stderr": "0.867", + "latency": "29.531", + "cost_per_test": "0.001601", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.867, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 84.546, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "84.546", + "raw_stderr": "1.034", + "latency": "29.531", + "cost_per_test": "0.001601", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.034, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-large-2512/33498f56-0608-483b-91ac-9e86933fb2b8.json b/data/vals_ai/mistralai/mistral-large-2512/33498f56-0608-483b-91ac-9e86933fb2b8.json new file mode 100644 index 0000000000000000000000000000000000000000..923cc27eb27fed7e77139b4f696aa0beeec6dd88 --- /dev/null +++ b/data/vals_ai/mistralai/mistral-large-2512/33498f56-0608-483b-91ac-9e86933fb2b8.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/mistralai_mistral-large-2512/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-large-2512", + "id": "mistralai/mistral-large-2512", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-large-2512", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 95.342, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "95.342", + "raw_stderr": "1.174", + "latency": "8.113", + "cost_per_test": "0.00075", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.174, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 17.143, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "17.143", + "raw_stderr": "2.015", + "latency": "53.158", + "cost_per_test": "0.003963", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.015, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 53.525, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "53.525", + "raw_stderr": "2.549", + "latency": "25.394", + "cost_per_test": "0.001931", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.549, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 55.337, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "55.337", + "raw_stderr": "1.152", + "latency": "28.888", + "cost_per_test": "0.002215", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.152, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-large-2512/37620406-4eb9-4da7-bb30-826bb71ecb6c.json b/data/vals_ai/mistralai/mistral-large-2512/37620406-4eb9-4da7-bb30-826bb71ecb6c.json new file mode 100644 index 0000000000000000000000000000000000000000..66caf95f875136de1892a52facbf6ea0ec99cbbd --- /dev/null +++ b/data/vals_ai/mistralai/mistral-large-2512/37620406-4eb9-4da7-bb30-826bb71ecb6c.json @@ -0,0 +1,361 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/mistralai_mistral-large-2512/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-large-2512", + "id": "mistralai/mistral-large-2512", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-large-2512", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 4.762, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "4.762", + "raw_stderr": "3.326", + "latency": "216.75", + "cost_per_test": "0.33932", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.326, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 36.398, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "36.398", + "raw_stderr": "2.984", + "latency": "208.783", + "cost_per_test": "0.385336", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.984, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 56.186, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "56.186", + "raw_stderr": "3.571", + "latency": "150.597", + "cost_per_test": "0.229311", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.571, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "221.861", + "cost_per_test": "0.507432", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 41.4, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "41.4", + "raw_stderr": "2.205", + "latency": "186.955", + "cost_per_test": "0.321666", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.205, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-large-2512/71085131-86de-4f79-9899-5e8b60cd950c.json b/data/vals_ai/mistralai/mistral-large-2512/71085131-86de-4f79-9899-5e8b60cd950c.json new file mode 100644 index 0000000000000000000000000000000000000000..251ea2c7f50a18193358655a8a548deb4e3a7540 --- /dev/null +++ b/data/vals_ai/mistralai/mistral-large-2512/71085131-86de-4f79-9899-5e8b60cd950c.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/mistralai_mistral-large-2512/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-large-2512", + "id": "mistralai/mistral-large-2512", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-large-2512", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 64.141, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "64.141", + "raw_stderr": "3.324", + "latency": "21.802", + "cost_per_test": "0.002625", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.324, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 72.727, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.727", + "raw_stderr": "3.386", + "latency": "20.631", + "cost_per_test": "0.001675", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.386, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 68.434, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "68.434", + "raw_stderr": "2.372", + "latency": "21.216", + "cost_per_test": "0.00215", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.372, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-large-2512/73de82bd-cc61-4961-9a2e-2185a89b05c0.json b/data/vals_ai/mistralai/mistral-large-2512/73de82bd-cc61-4961-9a2e-2185a89b05c0.json new file mode 100644 index 0000000000000000000000000000000000000000..b5a255f4800ca2665484a67dbac9d7827dd7454f --- /dev/null +++ b/data/vals_ai/mistralai/mistral-large-2512/73de82bd-cc61-4961-9a2e-2185a89b05c0.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/mistralai_mistral-large-2512/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-large-2512", + "id": "mistralai/mistral-large-2512", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-large-2512", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 6.5, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "6.5", + "raw_stderr": "2.786", + "latency": "7790.385", + "cost_per_test": "3.946165", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.786, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 1.5, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "1.5", + "raw_stderr": "0.874", + "latency": "5786.766", + "cost_per_test": "2.20173", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.874, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 4.0, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "4", + "raw_stderr": "1.46", + "latency": "6788.576", + "cost_per_test": "3.073948", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.46, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-large-2512/7d8f6f65-3ea0-47a8-a18a-68e98f195e38.json b/data/vals_ai/mistralai/mistral-large-2512/7d8f6f65-3ea0-47a8-a18a-68e98f195e38.json new file mode 100644 index 0000000000000000000000000000000000000000..ee70d875215c0e5e6fc7f75934656d4a78e26f27 --- /dev/null +++ b/data/vals_ai/mistralai/mistral-large-2512/7d8f6f65-3ea0-47a8-a18a-68e98f195e38.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/mistralai_mistral-large-2512/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-large-2512", + "id": "mistralai/mistral-large-2512", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-large-2512", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 59.79, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.79", + "raw_stderr": "1.674", + "latency": "20.072", + "cost_per_test": "0.035945", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.674, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 63.753, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.753", + "raw_stderr": "1.641", + "latency": "53.318", + "cost_per_test": "0.001232", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.641, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.033, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.033", + "raw_stderr": "0.96", + "latency": "48.842", + "cost_per_test": "0.024166", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.96, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 59.557, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.557", + "raw_stderr": "1.675", + "latency": "73.135", + "cost_per_test": "0.035322", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.675, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-large-2512/97d19a5d-86d5-4581-b020-5a4b6c43d108.json b/data/vals_ai/mistralai/mistral-large-2512/97d19a5d-86d5-4581-b020-5a4b6c43d108.json new file mode 100644 index 0000000000000000000000000000000000000000..913bd2e4e71cfea4f3c32d2f6bc8634a740c7b97 --- /dev/null +++ b/data/vals_ai/mistralai/mistral-large-2512/97d19a5d-86d5-4581-b020-5a4b6c43d108.json @@ -0,0 +1,101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/mistralai_mistral-large-2512/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-large-2512", + "id": "mistralai/mistral-large-2512", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-large-2512", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 66.185, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.185", + "raw_stderr": "1.137", + "latency": "13.55", + "cost_per_test": "0.001477", + "temperature": "0.065", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.137, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.065, + "max_tokens": 256000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-large-2512/d71cb884-7962-4b52-b6cd-a8ed3292918b.json b/data/vals_ai/mistralai/mistral-large-2512/d71cb884-7962-4b52-b6cd-a8ed3292918b.json new file mode 100644 index 0000000000000000000000000000000000000000..1cf9603429746dee7fbb19b9fd9f8babbf4ee23e --- /dev/null +++ b/data/vals_ai/mistralai/mistral-large-2512/d71cb884-7962-4b52-b6cd-a8ed3292918b.json @@ -0,0 +1,346 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/mistralai_mistral-large-2512/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-large-2512", + "id": "mistralai/mistral-large-2512", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-large-2512", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 17.46, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "17.46", + "raw_stderr": "4.764", + "latency": "53.013", + "cost_per_test": "0.006272", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 4.764, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 11.905, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "11.905", + "raw_stderr": "4.152", + "latency": "62.961", + "cost_per_test": "0.00818", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 4.152, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "0", + "raw_stderr": "0", + "latency": "101.608", + "cost_per_test": "0.010832", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 44.419, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "44.419", + "raw_stderr": "5.581", + "latency": "49.23", + "cost_per_test": "0.009728", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 5.581, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:mistralai/mistral-large-2512:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 24.595, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "24.595", + "raw_stderr": "2.81", + "latency": "55.068", + "cost_per_test": "0.00806", + "temperature": "1", + "max_output_tokens": "30000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.81, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-medium-2505/29f943cf-66ad-4af7-80bb-3e9b73ce15cd.json b/data/vals_ai/mistralai/mistral-medium-2505/29f943cf-66ad-4af7-80bb-3e9b73ce15cd.json new file mode 100644 index 0000000000000000000000000000000000000000..25ce4d43348b5bda8d1f8c963f47897f1835d2ec --- /dev/null +++ b/data/vals_ai/mistralai/mistral-medium-2505/29f943cf-66ad-4af7-80bb-3e9b73ce15cd.json @@ -0,0 +1,101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/mistralai_mistral-medium-2505/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-medium-2505", + "id": "mistralai/mistral-medium-2505", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-medium-2505", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 62.969, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "62.969", + "raw_stderr": "1.161", + "latency": "12.56", + "cost_per_test": "0.001567", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.161, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-medium-2505/4c14cbaa-7cba-47a3-86c4-7a4e9fc8b0cd.json b/data/vals_ai/mistralai/mistral-medium-2505/4c14cbaa-7cba-47a3-86c4-7a4e9fc8b0cd.json new file mode 100644 index 0000000000000000000000000000000000000000..3a1f67b544629509ad1acfa15be3e3ac4dbebbce --- /dev/null +++ b/data/vals_ai/mistralai/mistral-medium-2505/4c14cbaa-7cba-47a3-86c4-7a4e9fc8b0cd.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/mistralai_mistral-medium-2505/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-medium-2505", + "id": "mistralai/mistral-medium-2505", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-medium-2505", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 66.057, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.057", + "raw_stderr": "1.335", + "latency": "6.77", + "cost_per_test": "0.002657", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.335, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 36.446, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "36.446", + "raw_stderr": "0.756", + "latency": "7.138", + "cost_per_test": "0.00269", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.756, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 6.836, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "6.836", + "raw_stderr": "0.712", + "latency": "7.507", + "cost_per_test": "0.002722", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.712, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-medium-2505/75c7e0f1-2e5f-4ff8-807a-82e8fc3f2804.json b/data/vals_ai/mistralai/mistral-medium-2505/75c7e0f1-2e5f-4ff8-807a-82e8fc3f2804.json new file mode 100644 index 0000000000000000000000000000000000000000..f9d22c52ceea62337c2ded20a8865d6a5a4a7ec6 --- /dev/null +++ b/data/vals_ai/mistralai/mistral-medium-2505/75c7e0f1-2e5f-4ff8-807a-82e8fc3f2804.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/mistralai_mistral-medium-2505/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-medium-2505", + "id": "mistralai/mistral-medium-2505", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-medium-2505", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.866, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.866", + "raw_stderr": "1.305", + "latency": "6.356", + "cost_per_test": "0.001278", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.305, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.735, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.735", + "raw_stderr": "1.435", + "latency": "8.464", + "cost_per_test": "0", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.435, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.594, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.594", + "raw_stderr": "1.204", + "latency": "15.685", + "cost_per_test": "0", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.204, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.293, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.293", + "raw_stderr": "2.044", + "latency": "8.798", + "cost_per_test": "0", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.044, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.161, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.161", + "raw_stderr": "1.374", + "latency": "6.655", + "cost_per_test": "0", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.374, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 59.649, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "59.649", + "raw_stderr": "1.581", + "latency": "20.497", + "cost_per_test": "0", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.581, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.939, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.939", + "raw_stderr": "1.57", + "latency": "4.505", + "cost_per_test": "0", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.57, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 67.979, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "67.979", + "raw_stderr": "2.405", + "latency": "3.949", + "cost_per_test": "0", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.405, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 52.77, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "52.77", + "raw_stderr": "1.505", + "latency": "6.141", + "cost_per_test": "0", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.505, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.902, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.902", + "raw_stderr": "1.026", + "latency": "14.004", + "cost_per_test": "0", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.026, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.082, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.082", + "raw_stderr": "1.41", + "latency": "3.419", + "cost_per_test": "0.00077", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.41, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.29, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.29", + "raw_stderr": "0.421", + "latency": "8.831", + "cost_per_test": "0.000209", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.421, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.543, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.543", + "raw_stderr": "2.028", + "latency": "5.458", + "cost_per_test": "0", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.028, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.216, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.216", + "raw_stderr": "1.109", + "latency": "14.915", + "cost_per_test": "0", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.109, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.326, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.326", + "raw_stderr": "1.459", + "latency": "4.791", + "cost_per_test": "0.000878", + "temperature": "0.7", + "max_output_tokens": "65536", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.459, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 65536, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-medium-2505/9cacd3b3-9d9a-40c6-a29e-2d94f74591b1.json b/data/vals_ai/mistralai/mistral-medium-2505/9cacd3b3-9d9a-40c6-a29e-2d94f74591b1.json new file mode 100644 index 0000000000000000000000000000000000000000..bb9e0e58387a965b5da3441920abf3519cc505c6 --- /dev/null +++ b/data/vals_ai/mistralai/mistral-medium-2505/9cacd3b3-9d9a-40c6-a29e-2d94f74591b1.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/mistralai_mistral-medium-2505/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-medium-2505", + "id": "mistralai/mistral-medium-2505", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-medium-2505", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.586, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.586", + "raw_stderr": "0.862", + "latency": "0.307", + "temperature": "0.7", + "max_output_tokens": "32768", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.862, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 62.799, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "62.799", + "raw_stderr": "0.257", + "latency": "0.385", + "temperature": "0.7", + "max_output_tokens": "32768", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.257, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 36.119, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "36.119", + "raw_stderr": "0.731", + "latency": "0.47", + "temperature": "0.7", + "max_output_tokens": "32768", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.731, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 61.979, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "61.979", + "raw_stderr": "0.473", + "latency": "0.676", + "temperature": "0.7", + "max_output_tokens": "32768", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.473, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 59.773, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "59.773", + "raw_stderr": "1.086", + "latency": "0.407", + "temperature": "0.7", + "max_output_tokens": "32768", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.086, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 70.616, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "70.616", + "raw_stderr": "1.751", + "latency": "1.81", + "temperature": "0.7", + "max_output_tokens": "32768", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.751, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-medium-2505/a2d50563-0e3f-4f8c-bfc2-489f40bc62b2.json b/data/vals_ai/mistralai/mistral-medium-2505/a2d50563-0e3f-4f8c-bfc2-489f40bc62b2.json new file mode 100644 index 0000000000000000000000000000000000000000..88f87366cf91f9bb392dc8d65d7f3eb84f9d2f45 --- /dev/null +++ b/data/vals_ai/mistralai/mistral-medium-2505/a2d50563-0e3f-4f8c-bfc2-489f40bc62b2.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/mistralai_mistral-medium-2505/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-medium-2505", + "id": "mistralai/mistral-medium-2505", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-medium-2505", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 58.79, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.79", + "raw_stderr": "1.407", + "latency": "11.052", + "cost_per_test": "0.001422", + "temperature": "0.7", + "max_output_tokens": "16384", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.407, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 70.319, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "70.319", + "raw_stderr": "0.894", + "latency": "11.052", + "cost_per_test": "0.001422", + "temperature": "0.7", + "max_output_tokens": "16384", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.894, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 81.848, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "81.848", + "raw_stderr": "1.102", + "latency": "11.052", + "cost_per_test": "0.001422", + "temperature": "0.7", + "max_output_tokens": "16384", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.102, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-medium-2505/d0e7483a-d15c-4eae-9f0e-4cacb4d895a6.json b/data/vals_ai/mistralai/mistral-medium-2505/d0e7483a-d15c-4eae-9f0e-4cacb4d895a6.json new file mode 100644 index 0000000000000000000000000000000000000000..c362a4dead7991ce1a5ed85cc8b02a34e516ba93 --- /dev/null +++ b/data/vals_ai/mistralai/mistral-medium-2505/d0e7483a-d15c-4eae-9f0e-4cacb4d895a6.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/mistralai_mistral-medium-2505/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-medium-2505", + "id": "mistralai/mistral-medium-2505", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-medium-2505", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.538, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.538", + "raw_stderr": "1.661", + "latency": "11.934", + "cost_per_test": "0.00108", + "temperature": "0.7", + "max_output_tokens": "32768", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.661, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 44.522, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "44.522", + "raw_stderr": "1.697", + "latency": "43.687", + "cost_per_test": "0.042618", + "temperature": "0.7", + "max_output_tokens": "32768", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.697, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 50.738, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "50.738", + "raw_stderr": "0.974", + "latency": "30.346", + "cost_per_test": "0.024082", + "temperature": "0.7", + "max_output_tokens": "32768", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.974, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 46.154, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "46.154", + "raw_stderr": "1.702", + "latency": "35.417", + "cost_per_test": "0.028549", + "temperature": "0.7", + "max_output_tokens": "32768", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.702, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-medium-2505/e252a4f9-79a1-4c56-841c-c5be1b551f92.json b/data/vals_ai/mistralai/mistral-medium-2505/e252a4f9-79a1-4c56-841c-c5be1b551f92.json new file mode 100644 index 0000000000000000000000000000000000000000..9bbeb536b1715f2d4eec6645aad943febe02850f --- /dev/null +++ b/data/vals_ai/mistralai/mistral-medium-2505/e252a4f9-79a1-4c56-841c-c5be1b551f92.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/mistralai_mistral-medium-2505/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-medium-2505", + "id": "mistralai/mistral-medium-2505", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-medium-2505", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 85.714, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "85.714", + "raw_stderr": "1.95", + "latency": "5.163", + "cost_per_test": "0.000698", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.95, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 6.0, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "6", + "raw_stderr": "1.269", + "latency": "18.796", + "cost_per_test": "0.002413", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.269, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 42.82, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "42.82", + "raw_stderr": "2.528", + "latency": "13.49", + "cost_per_test": "0.001719", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.528, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:mistralai/mistral-medium-2505:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 44.845, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "44.845", + "raw_stderr": "1.145", + "latency": "12.483", + "cost_per_test": "0.00161", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.145, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-medium-3.5/07ba7f68-ca23-4e8e-bd13-6c76b7cdc3ca.json b/data/vals_ai/mistralai/mistral-medium-3.5/07ba7f68-ca23-4e8e-bd13-6c76b7cdc3ca.json new file mode 100644 index 0000000000000000000000000000000000000000..ddfb373e754268182aea51340f96b6efcc31d6e3 --- /dev/null +++ b/data/vals_ai/mistralai/mistral-medium-3.5/07ba7f68-ca23-4e8e-bd13-6c76b7cdc3ca.json @@ -0,0 +1,249 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/mistralai_mistral-medium-3.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-medium-3.5", + "id": "mistralai/mistral-medium-3.5", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-medium-3.5", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 34.848, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "34.848", + "raw_stderr": "3.386", + "latency": "248.163", + "cost_per_test": "0.089386", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.386, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 34.848, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "34.848", + "raw_stderr": "3.375", + "latency": "264.176", + "cost_per_test": "0.101592", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.375, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 34.848, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "34.848", + "raw_stderr": "2.39", + "latency": "256.17", + "cost_per_test": "0.095489", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.39, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-medium-3.5/374d397f-8c78-4c34-ab26-693e7ae0a3bd.json b/data/vals_ai/mistralai/mistral-medium-3.5/374d397f-8c78-4c34-ab26-693e7ae0a3bd.json new file mode 100644 index 0000000000000000000000000000000000000000..fa3f4eb79433a735fb7e17d231288303ade2eabc --- /dev/null +++ b/data/vals_ai/mistralai/mistral-medium-3.5/374d397f-8c78-4c34-ab26-693e7ae0a3bd.json @@ -0,0 +1,784 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/mistralai_mistral-medium-3.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-medium-3.5", + "id": "mistralai/mistral-medium-3.5", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-medium-3.5", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 27.312, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "27.312", + "raw_stderr": "2.309", + "latency": "853.326", + "cost_per_test": "1.995722", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.309, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 21.632, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "21.632", + "raw_stderr": "0.963", + "latency": "768.405", + "cost_per_test": "1.908513", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.963, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 31.512, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "31.512", + "raw_stderr": "3.178", + "latency": "1007.855", + "cost_per_test": "2.097265", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.178, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 40.223, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "40.223", + "raw_stderr": "1.634", + "latency": "929.601", + "cost_per_test": "1.750126", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.634, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 42.125, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "42.125", + "raw_stderr": "2.287", + "latency": "548.402", + "cost_per_test": "1.725068", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.287, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 5.141, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "5.141", + "raw_stderr": "0.821", + "latency": "907.652", + "cost_per_test": "2.387633", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.821, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 47.935, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "47.935", + "raw_stderr": "2.346", + "latency": "618.117", + "cost_per_test": "1.416491", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.346, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 44.752, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "44.752", + "raw_stderr": "2.387", + "latency": "761.631", + "cost_per_test": "1.886966", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.387, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 37.023, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "37.023", + "raw_stderr": "1.316", + "latency": "560.062", + "cost_per_test": "1.648409", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.316, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:overall:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 32.103, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "32.103", + "raw_stderr": "0.967", + "latency": "768.405", + "cost_per_test": "1.908513", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.967, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 12.901, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "12.901", + "raw_stderr": "2.211", + "latency": "1118.384", + "cost_per_test": "2.287472", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.211, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-medium-3.5/5d3f156c-d6f7-4473-8283-80ec219c16d0.json b/data/vals_ai/mistralai/mistral-medium-3.5/5d3f156c-d6f7-4473-8283-80ec219c16d0.json new file mode 100644 index 0000000000000000000000000000000000000000..f8dd7243b36f46ad821ed4cba362b8e61ec4d8b0 --- /dev/null +++ b/data/vals_ai/mistralai/mistral-medium-3.5/5d3f156c-d6f7-4473-8283-80ec219c16d0.json @@ -0,0 +1,104 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/mistralai_mistral-medium-3.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-medium-3.5", + "id": "mistralai/mistral-medium-3.5", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-medium-3.5", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 10.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "10", + "raw_stderr": "3.015", + "latency": "402.484", + "cost_per_test": "1.501427", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.015, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-medium-3.5/7ec48bcc-88f4-4c2a-83cb-e6c1429f3acd.json b/data/vals_ai/mistralai/mistral-medium-3.5/7ec48bcc-88f4-4c2a-83cb-e6c1429f3acd.json new file mode 100644 index 0000000000000000000000000000000000000000..0cb5569874c5e46e2babdeb973e8da4414d3d477 --- /dev/null +++ b/data/vals_ai/mistralai/mistral-medium-3.5/7ec48bcc-88f4-4c2a-83cb-e6c1429f3acd.json @@ -0,0 +1,320 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/mistralai_mistral-medium-3.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-medium-3.5", + "id": "mistralai/mistral-medium-3.5", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-medium-3.5", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66.667", + "raw_stderr": "8.333", + "latency": "22009.013", + "cost_per_test": "26.499387", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 8.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 15.556, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "15.556", + "raw_stderr": "2.222", + "latency": "2934.535", + "cost_per_test": "3.533252", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.222, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 49.697, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "49.697", + "raw_stderr": "4.242", + "latency": "1600.656", + "cost_per_test": "1.927228", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 4.242, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 38.951, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "38.951", + "raw_stderr": "2.701", + "latency": "989.169", + "cost_per_test": "1.190984", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.701, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-medium-3.5/88cfb2cc-c9fc-4f39-9ade-4e329c9d421a.json b/data/vals_ai/mistralai/mistral-medium-3.5/88cfb2cc-c9fc-4f39-9ade-4e329c9d421a.json new file mode 100644 index 0000000000000000000000000000000000000000..0681d1147465295202b024ae2f2c9b566b0cb79d --- /dev/null +++ b/data/vals_ai/mistralai/mistral-medium-3.5/88cfb2cc-c9fc-4f39-9ade-4e329c9d421a.json @@ -0,0 +1,564 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_multimodal_index/mistralai_mistral-medium-3.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Multimodal Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-medium-3.5", + "id": "mistralai/mistral-medium-3.5", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-medium-3.5", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_multimodal_index:corp_fin_v2:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Multimodal Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (CorpFin v2).", + "metric_id": "vals_ai.vals_multimodal_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 59.907, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "59.907", + "raw_stderr": "1.673", + "latency": "42.14", + "cost_per_test": "0.127154", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.673, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:finance_agent:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.finance_agent", + "source_data": { + "dataset_name": "Vals Multimodal Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_multimodal_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 31.511, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "31.511", + "raw_stderr": "2.394", + "latency": "661.916", + "cost_per_test": "1.736097", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.394, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:mortgage_tax:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.mortgage_tax", + "source_data": { + "dataset_name": "Vals Multimodal Index - Mortgage Tax", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "mortgage_tax", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Mortgage Tax).", + "metric_id": "vals_ai.vals_multimodal_index.mortgage_tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 28.895, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "mortgage_tax", + "task_name": "Mortgage Tax", + "dataset_type": "private", + "industry": "index", + "raw_score": "28.895", + "raw_stderr": "0.838", + "latency": "68.43", + "cost_per_test": "0.024032", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.838, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:overall:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.overall", + "source_data": { + "dataset_name": "Vals Multimodal Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Overall).", + "metric_id": "vals_ai.vals_multimodal_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 34.768, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "34.768", + "raw_stderr": "0.833", + "latency": "734.703", + "cost_per_test": "3.453606", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.833, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:sage:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.sage", + "source_data": { + "dataset_name": "Vals Multimodal Index - SAGE", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "sage", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SAGE).", + "metric_id": "vals_ai.vals_multimodal_index.sage.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 37.613, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "sage", + "task_name": "SAGE", + "dataset_type": "private", + "industry": "index", + "raw_score": "37.613", + "raw_stderr": "3.356", + "latency": "55.844", + "cost_per_test": "0.072775", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.356, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:swebench:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.swebench", + "source_data": { + "dataset_name": "Vals Multimodal Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SWE-bench).", + "metric_id": "vals_ai.vals_multimodal_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 64.706, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "64.706", + "raw_stderr": "4.732", + "latency": "854.891", + "cost_per_test": "4.288697" + }, + "uncertainty": { + "standard_error": { + "value": 4.732, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:terminal_bench_2_1:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Multimodal Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_multimodal_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 38.951, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "38.951", + "raw_stderr": "2.701", + "latency": "989.169", + "cost_per_test": "1.190984", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.701, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:vibe_code_bench:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Multimodal Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_multimodal_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 1.24, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "1.24", + "raw_stderr": "1.24", + "latency": "2470.533", + "cost_per_test": "16.735503", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.24, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-medium-3.5/96a99de2-f543-4aba-bb70-5453cbf87343.json b/data/vals_ai/mistralai/mistral-medium-3.5/96a99de2-f543-4aba-bb70-5453cbf87343.json new file mode 100644 index 0000000000000000000000000000000000000000..2c9eb9655b82ec9dbd502d4c621b05ccec7f33f8 --- /dev/null +++ b/data/vals_ai/mistralai/mistral-medium-3.5/96a99de2-f543-4aba-bb70-5453cbf87343.json @@ -0,0 +1,391 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/mistralai_mistral-medium-3.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-medium-3.5", + "id": "mistralai/mistral-medium-3.5", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-medium-3.5", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "7.362", + "latency": "1391.329", + "cost_per_test": "5.432118", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 7.362, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 62.452, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "62.452", + "raw_stderr": "3.003", + "latency": "1267.104", + "cost_per_test": "4.395121", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.003, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 79.381, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "79.381", + "raw_stderr": "2.912", + "latency": "902.978", + "cost_per_test": "3.131586", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.912, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "2504.149", + "cost_per_test": "6.502794", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:overall:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 66.4, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66.4", + "raw_stderr": "2.114", + "latency": "1143.68", + "cost_per_test": "4.004623", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.114, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-medium-3.5/a9d6cf20-bdfe-4eda-b772-00bc47c92e15.json b/data/vals_ai/mistralai/mistral-medium-3.5/a9d6cf20-bdfe-4eda-b772-00bc47c92e15.json new file mode 100644 index 0000000000000000000000000000000000000000..69bd7ff6f062b6af1f9fcdd6b85a4fb8856cd70a --- /dev/null +++ b/data/vals_ai/mistralai/mistral-medium-3.5/a9d6cf20-bdfe-4eda-b772-00bc47c92e15.json @@ -0,0 +1,308 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/mistralai_mistral-medium-3.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-medium-3.5", + "id": "mistralai/mistral-medium-3.5", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-medium-3.5", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 55.361, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "55.361", + "raw_stderr": "1.697", + "latency": "30.593", + "cost_per_test": "0.018845", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.697, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.072, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.072", + "raw_stderr": "1.665", + "latency": "58.363", + "cost_per_test": "0.291221", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.665, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 58.78, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.78", + "raw_stderr": "0.969", + "latency": "43.699", + "cost_per_test": "0.14574", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.969, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 59.907, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.907", + "raw_stderr": "1.673", + "latency": "42.14", + "cost_per_test": "0.127154", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.673, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-medium-3.5/b236202b-60a7-46c7-a30b-740d8f7a7537.json b/data/vals_ai/mistralai/mistral-medium-3.5/b236202b-60a7-46c7-a30b-740d8f7a7537.json new file mode 100644 index 0000000000000000000000000000000000000000..3749c9436a51ee2914cd6e5889128a4b1a97052a --- /dev/null +++ b/data/vals_ai/mistralai/mistral-medium-3.5/b236202b-60a7-46c7-a30b-740d8f7a7537.json @@ -0,0 +1,240 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/mistralai_mistral-medium-3.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-medium-3.5", + "id": "mistralai/mistral-medium-3.5", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-medium-3.5", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 56.746, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.746", + "raw_stderr": "1.417", + "latency": "102.392", + "cost_per_test": "0.079101", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.417, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 67.988, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.988", + "raw_stderr": "0.916", + "latency": "102.392", + "cost_per_test": "0.079101", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.916, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 79.231, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "79.231", + "raw_stderr": "1.16", + "latency": "102.392", + "cost_per_test": "0.079101", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.16, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-medium-3.5/b8026df0-a245-4ff2-8b04-852222f8b6ce.json b/data/vals_ai/mistralai/mistral-medium-3.5/b8026df0-a245-4ff2-8b04-852222f8b6ce.json new file mode 100644 index 0000000000000000000000000000000000000000..0ba7a6bb0230d42356d96da163870e16063e9212 --- /dev/null +++ b/data/vals_ai/mistralai/mistral-medium-3.5/b8026df0-a245-4ff2-8b04-852222f8b6ce.json @@ -0,0 +1,104 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/mistralai_mistral-medium-3.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-medium-3.5", + "id": "mistralai/mistral-medium-3.5", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-medium-3.5", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 2.888, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "2.888", + "raw_stderr": "1.713", + "latency": "2470.533", + "cost_per_test": "16.735503", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.713, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-medium-3.5/bd107313-7374-4855-b2c2-2d203f066f8a.json b/data/vals_ai/mistralai/mistral-medium-3.5/bd107313-7374-4855-b2c2-2d203f066f8a.json new file mode 100644 index 0000000000000000000000000000000000000000..564ebecfa73af11744f740788956fbfebc85d013 --- /dev/null +++ b/data/vals_ai/mistralai/mistral-medium-3.5/bd107313-7374-4855-b2c2-2d203f066f8a.json @@ -0,0 +1,240 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/mistralai_mistral-medium-3.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-medium-3.5", + "id": "mistralai/mistral-medium-3.5", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-medium-3.5", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 11.924, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "11.924", + "raw_stderr": "0.914", + "latency": "75.468", + "cost_per_test": "0.030861", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.914, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 28.895, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "28.895", + "raw_stderr": "0.838", + "latency": "68.43", + "cost_per_test": "0.024032", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.838, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 45.866, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "45.866", + "raw_stderr": "1.405", + "latency": "61.393", + "cost_per_test": "0.017204", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.405, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-medium-3.5/c03a918a-e4a7-42eb-8fc9-2011c8c9430f.json b/data/vals_ai/mistralai/mistral-medium-3.5/c03a918a-e4a7-42eb-8fc9-2011c8c9430f.json new file mode 100644 index 0000000000000000000000000000000000000000..7860f8b1ad0ef01a7bb7b90c16a4c8628b8531ae --- /dev/null +++ b/data/vals_ai/mistralai/mistral-medium-3.5/c03a918a-e4a7-42eb-8fc9-2011c8c9430f.json @@ -0,0 +1,376 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/mistralai_mistral-medium-3.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-medium-3.5", + "id": "mistralai/mistral-medium-3.5", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-medium-3.5", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 41.27, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "41.27", + "raw_stderr": "6.185", + "latency": "66.2", + "cost_per_test": "0.079624", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 6.185, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:automata:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 30.159, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "30.159", + "raw_stderr": "5.701", + "latency": "54.582", + "cost_per_test": "0.066288", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 5.701, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:calculus:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 2.778, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "2.778", + "raw_stderr": "2.66", + "latency": "73.75", + "cost_per_test": "0.073682", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.66, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 41.411, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "41.411", + "raw_stderr": "5.535", + "latency": "46.749", + "cost_per_test": "0.072413", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 5.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:overall:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 37.613, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "37.613", + "raw_stderr": "3.356", + "latency": "55.844", + "cost_per_test": "0.072775", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.356, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-medium-3.5/c11d7602-4174-481d-a623-91829b0416a0.json b/data/vals_ai/mistralai/mistral-medium-3.5/c11d7602-4174-481d-a623-91829b0416a0.json new file mode 100644 index 0000000000000000000000000000000000000000..b2d5952c7a3814bbbcfe65a99ad034c2fe63365e --- /dev/null +++ b/data/vals_ai/mistralai/mistral-medium-3.5/c11d7602-4174-481d-a623-91829b0416a0.json @@ -0,0 +1,428 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/mistralai_mistral-medium-3.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-medium-3.5", + "id": "mistralai/mistral-medium-3.5", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-medium-3.5", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 59.907, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "59.907", + "raw_stderr": "1.673", + "latency": "42.14", + "cost_per_test": "0.127154", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.673, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 31.511, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "31.511", + "raw_stderr": "2.394", + "latency": "661.916", + "cost_per_test": "1.736097", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.394, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:overall:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 37.814, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "37.814", + "raw_stderr": "1.057", + "latency": "1003.73", + "cost_per_test": "4.815687", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.057, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 64.706, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "64.706", + "raw_stderr": "4.732", + "latency": "854.891", + "cost_per_test": "4.288697" + }, + "uncertainty": { + "standard_error": { + "value": 4.732, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 38.951, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "38.951", + "raw_stderr": "2.701", + "latency": "989.169", + "cost_per_test": "1.190984", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.701, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 1.24, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "1.24", + "raw_stderr": "1.24", + "latency": "2470.533", + "cost_per_test": "16.735503", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.24, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-medium-3.5/c607d791-1b2e-43a5-8507-6c321eea0cb1.json b/data/vals_ai/mistralai/mistral-medium-3.5/c607d791-1b2e-43a5-8507-6c321eea0cb1.json new file mode 100644 index 0000000000000000000000000000000000000000..c727b42678c7160f09a39316c6f45db895aa991c --- /dev/null +++ b/data/vals_ai/mistralai/mistral-medium-3.5/c607d791-1b2e-43a5-8507-6c321eea0cb1.json @@ -0,0 +1,104 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/mistralai_mistral-medium-3.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-medium-3.5", + "id": "mistralai/mistral-medium-3.5", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-medium-3.5", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 67.728, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "67.728", + "raw_stderr": "2.011", + "latency": "104.053", + "cost_per_test": "0.157641", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.011, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-medium-3.5/cbeb6ba0-e8bd-4d29-89c8-cd9041100b12.json b/data/vals_ai/mistralai/mistral-medium-3.5/cbeb6ba0-e8bd-4d29-89c8-cd9041100b12.json new file mode 100644 index 0000000000000000000000000000000000000000..80ce3cd06ca2c952dd2c12ed67d022936d96b8a4 --- /dev/null +++ b/data/vals_ai/mistralai/mistral-medium-3.5/cbeb6ba0-e8bd-4d29-89c8-cd9041100b12.json @@ -0,0 +1,1101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/mistralai_mistral-medium-3.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-medium-3.5", + "id": "mistralai/mistral-medium-3.5", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-medium-3.5", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 64.435, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "64.435", + "raw_stderr": "1.788", + "latency": "90.757", + "cost_per_test": "0.03228", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.788, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 64.385, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "64.385", + "raw_stderr": "1.705", + "latency": "101.826", + "cost_per_test": "0.047867", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.705, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.057, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.057", + "raw_stderr": "1.319", + "latency": "97.072", + "cost_per_test": "0.0594", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.319, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.927, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.927", + "raw_stderr": "1.858", + "latency": "50.137", + "cost_per_test": "0.046075", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.858, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.199, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.199", + "raw_stderr": "1.421", + "latency": "60.82", + "cost_per_test": "0.032339", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.421, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.225, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.225", + "raw_stderr": "1.326", + "latency": "81.498", + "cost_per_test": "0.086574", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.326, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.306, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.306", + "raw_stderr": "1.508", + "latency": "57.996", + "cost_per_test": "0.023976", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.508, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 69.554, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.554", + "raw_stderr": "2.358", + "latency": "52.52", + "cost_per_test": "0.030174", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.358, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 56.857, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "56.857", + "raw_stderr": "1.493", + "latency": "72.657", + "cost_per_test": "0.053547", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.493, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.97, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.97", + "raw_stderr": "0.78", + "latency": "48.241", + "cost_per_test": "0.037224", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.78, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.139, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.139", + "raw_stderr": "1.36", + "latency": "39.948", + "cost_per_test": "0.026514", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.36, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.335, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.335", + "raw_stderr": "0.418", + "latency": "66.384", + "cost_per_test": "0.041763", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.418, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.146, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.146", + "raw_stderr": "1.984", + "latency": "52.443", + "cost_per_test": "0.029496", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.984, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.528, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.528", + "raw_stderr": "0.947", + "latency": "71.364", + "cost_per_test": "0.056352", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.947, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:mistralai/mistral-medium-3.5:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.957, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.957", + "raw_stderr": "1.331", + "latency": "52.096", + "cost_per_test": "0.022867", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "80000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.331, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 80000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-small-2402/04d5b34e-9cef-48ae-b5d7-26c5358247bc.json b/data/vals_ai/mistralai/mistral-small-2402/04d5b34e-9cef-48ae-b5d7-26c5358247bc.json new file mode 100644 index 0000000000000000000000000000000000000000..42f329cc9449d72ad12cb62f842e59d4d66e2eef --- /dev/null +++ b/data/vals_ai/mistralai/mistral-small-2402/04d5b34e-9cef-48ae-b5d7-26c5358247bc.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/mistralai_mistral-small-2402/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-small-2402", + "id": "mistralai/mistral-small-2402", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-small-2402", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:mistralai/mistral-small-2402:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.637, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.637", + "raw_stderr": "1.504", + "latency": "3.367", + "cost_per_test": "0.000439", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.504, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:mistralai/mistral-small-2402:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 70.596, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.596", + "raw_stderr": "1.622", + "latency": "5.154", + "cost_per_test": "0.000479", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.622, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:mistralai/mistral-small-2402:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 68.11, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "68.11", + "raw_stderr": "1.385", + "latency": "8.57", + "cost_per_test": "0.000727", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.385, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:mistralai/mistral-small-2402:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 67.561, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "67.561", + "raw_stderr": "2.312", + "latency": "4.094", + "cost_per_test": "0.000531", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.312, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:mistralai/mistral-small-2402:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.327, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.327", + "raw_stderr": "1.557", + "latency": "3.519", + "cost_per_test": "0.000429", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.557, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:mistralai/mistral-small-2402:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 46.027, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "46.027", + "raw_stderr": "1.601", + "latency": "9.576", + "cost_per_test": "0.000792", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.601, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:mistralai/mistral-small-2402:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 69.56, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.56", + "raw_stderr": "1.609", + "latency": "2.709", + "cost_per_test": "0.000342", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.609, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:mistralai/mistral-small-2402:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 54.593, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "54.593", + "raw_stderr": "2.551", + "latency": "3.661", + "cost_per_test": "0.000569", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.551, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:mistralai/mistral-small-2402:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 40.872, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "40.872", + "raw_stderr": "1.482", + "latency": "3.275", + "cost_per_test": "0.000464", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.482, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:mistralai/mistral-small-2402:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 72.835, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.835", + "raw_stderr": "1.21", + "latency": "6.125", + "cost_per_test": "0.000645", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.21, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:mistralai/mistral-small-2402:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 65.043, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "65.043", + "raw_stderr": "1.569", + "latency": "2.356", + "cost_per_test": "0.0003", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.569, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:mistralai/mistral-small-2402:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 64.443, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "64.443", + "raw_stderr": "0.461", + "latency": "4.665", + "cost_per_test": "0.000502", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.461, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:mistralai/mistral-small-2402:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 58.517, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "58.517", + "raw_stderr": "2.206", + "latency": "3.092", + "cost_per_test": "0.000351", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.206, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:mistralai/mistral-small-2402:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 68.976, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "68.976", + "raw_stderr": "1.283", + "latency": "7.145", + "cost_per_test": "0.00061", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.283, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:mistralai/mistral-small-2402:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 68.546, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "68.546", + "raw_stderr": "1.644", + "latency": "2.667", + "cost_per_test": "0.000347", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.644, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-small-2402/63ea9b91-1d8d-4560-963e-ceeda8542667.json b/data/vals_ai/mistralai/mistral-small-2402/63ea9b91-1d8d-4560-963e-ceeda8542667.json new file mode 100644 index 0000000000000000000000000000000000000000..27cb5d43dfdf9c782126d93de507f48b2e710d22 --- /dev/null +++ b/data/vals_ai/mistralai/mistral-small-2402/63ea9b91-1d8d-4560-963e-ceeda8542667.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/mistralai_mistral-small-2402/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-small-2402", + "id": "mistralai/mistral-small-2402", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-small-2402", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:mistralai/mistral-small-2402:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 42.857, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "42.857", + "raw_stderr": "2.758", + "latency": "3.336", + "cost_per_test": "0.000263", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.758, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:mistralai/mistral-small-2402:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 0.571, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0.571", + "raw_stderr": "0.403", + "latency": "6.101", + "cost_per_test": "0.00047", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.403, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:mistralai/mistral-small-2402:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 3.916, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "3.916", + "raw_stderr": "0.991", + "latency": "5.684", + "cost_per_test": "0.000414", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.991, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:mistralai/mistral-small-2402:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 15.781, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "15.781", + "raw_stderr": "0.986", + "latency": "5.04", + "cost_per_test": "0.000382", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.986, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-small-2402/aec1e783-09d7-427a-bcff-a6d6ac6c2186.json b/data/vals_ai/mistralai/mistral-small-2402/aec1e783-09d7-427a-bcff-a6d6ac6c2186.json new file mode 100644 index 0000000000000000000000000000000000000000..85ec31e260d081c8312372298cb18b0e264728f3 --- /dev/null +++ b/data/vals_ai/mistralai/mistral-small-2402/aec1e783-09d7-427a-bcff-a6d6ac6c2186.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/mistralai_mistral-small-2402/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-small-2402", + "id": "mistralai/mistral-small-2402", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-small-2402", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:mistralai/mistral-small-2402:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 44.949, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "44.949", + "raw_stderr": "3.535", + "latency": "8.344", + "cost_per_test": "0.000845", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:mistralai/mistral-small-2402:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 56.566, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "56.566", + "raw_stderr": "3.523", + "latency": "7.629", + "cost_per_test": "0.000326", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.523, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:mistralai/mistral-small-2402:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 50.758, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "50.758", + "raw_stderr": "2.495", + "latency": "7.986", + "cost_per_test": "0.000586", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.495, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-small-2402/b1e112ca-9dca-4267-a34b-a414762f76c7.json b/data/vals_ai/mistralai/mistral-small-2402/b1e112ca-9dca-4267-a34b-a414762f76c7.json new file mode 100644 index 0000000000000000000000000000000000000000..bb11d434d2a354d616f451d77d0cda7a27b1b968 --- /dev/null +++ b/data/vals_ai/mistralai/mistral-small-2402/b1e112ca-9dca-4267-a34b-a414762f76c7.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/mistralai_mistral-small-2402/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-small-2402", + "id": "mistralai/mistral-small-2402", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-small-2402", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:mistralai/mistral-small-2402:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 38.267, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "38.267", + "raw_stderr": "1.39", + "latency": "9.018", + "cost_per_test": "0.000355", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.39, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:mistralai/mistral-small-2402:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 49.142, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "49.142", + "raw_stderr": "0.987", + "latency": "9.018", + "cost_per_test": "0.000355", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.987, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:mistralai/mistral-small-2402:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 60.016, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.016", + "raw_stderr": "1.401", + "latency": "9.018", + "cost_per_test": "0.000355", + "temperature": "0.7", + "max_output_tokens": "256000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.401, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-small-2503/012ba2a5-0173-4f43-9f3b-d34875fb893d.json b/data/vals_ai/mistralai/mistral-small-2503/012ba2a5-0173-4f43-9f3b-d34875fb893d.json new file mode 100644 index 0000000000000000000000000000000000000000..2ea7751b701e9b2e94d7f1066102061408cf24ee --- /dev/null +++ b/data/vals_ai/mistralai/mistral-small-2503/012ba2a5-0173-4f43-9f3b-d34875fb893d.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/mistralai_mistral-small-2503/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-small-2503", + "id": "mistralai/mistral-small-2503", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-small-2503", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 72.981, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "72.981", + "raw_stderr": "2.475", + "latency": "3.225", + "cost_per_test": "0.000104", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.475, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 3.143, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "3.143", + "raw_stderr": "0.933", + "latency": "5.311", + "cost_per_test": "0.000186", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.933, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 19.321, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "19.321", + "raw_stderr": "2.017", + "latency": "3.485", + "cost_per_test": "0.000152", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.017, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 31.815, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "31.815", + "raw_stderr": "1.109", + "latency": "4.007", + "cost_per_test": "0.000147", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.109, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-small-2503/416c8814-eac6-4f87-ad90-853347a97640.json b/data/vals_ai/mistralai/mistral-small-2503/416c8814-eac6-4f87-ad90-853347a97640.json new file mode 100644 index 0000000000000000000000000000000000000000..a0f48e2b46f2c879f1d757113d269c528c97e660 --- /dev/null +++ b/data/vals_ai/mistralai/mistral-small-2503/416c8814-eac6-4f87-ad90-853347a97640.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/mistralai_mistral-small-2503/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-small-2503", + "id": "mistralai/mistral-small-2503", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-small-2503", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 43.939, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "43.939", + "raw_stderr": "3.523", + "latency": "10.003", + "cost_per_test": "0.000366", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.523, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 44.444, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "44.444", + "raw_stderr": "3.472", + "latency": "7.645", + "cost_per_test": "0.000191", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.472, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 44.192, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "44.192", + "raw_stderr": "2.473", + "latency": "8.824", + "cost_per_test": "0.000278", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.473, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-small-2503/681b66e9-dfbf-473a-82dc-8c903a4688fc.json b/data/vals_ai/mistralai/mistral-small-2503/681b66e9-dfbf-473a-82dc-8c903a4688fc.json new file mode 100644 index 0000000000000000000000000000000000000000..f95a9bc40ba912af91cc3b3f9c4a4a1ecc1d12e2 --- /dev/null +++ b/data/vals_ai/mistralai/mistral-small-2503/681b66e9-dfbf-473a-82dc-8c903a4688fc.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/mistralai_mistral-small-2503/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-small-2503", + "id": "mistralai/mistral-small-2503", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-small-2503", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 46.525, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "46.525", + "raw_stderr": "1.426", + "latency": "7.952", + "cost_per_test": "0.000184", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.426, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 58.3, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.3", + "raw_stderr": "0.968", + "latency": "7.952", + "cost_per_test": "0.000184", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.968, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 70.074, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "70.074", + "raw_stderr": "1.309", + "latency": "7.952", + "cost_per_test": "0.000184", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.309, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-small-2503/80791b07-79cb-4612-9cf4-ef9410a40754.json b/data/vals_ai/mistralai/mistral-small-2503/80791b07-79cb-4612-9cf4-ef9410a40754.json new file mode 100644 index 0000000000000000000000000000000000000000..3e5733504fba82ee01482ce2b96fc63548222964 --- /dev/null +++ b/data/vals_ai/mistralai/mistral-small-2503/80791b07-79cb-4612-9cf4-ef9410a40754.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/mistralai_mistral-small-2503/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-small-2503", + "id": "mistralai/mistral-small-2503", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-small-2503", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 61.857, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "61.857", + "raw_stderr": "1.154", + "latency": "0.568", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.154, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 71.413, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "71.413", + "raw_stderr": "0.244", + "latency": "0.369", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.244, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 76.453, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "76.453", + "raw_stderr": "0.775", + "latency": "0.329", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.775, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 69.157, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "69.157", + "raw_stderr": "0.512", + "latency": "0.583", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.512, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 69.778, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "69.778", + "raw_stderr": "1.004", + "latency": "0.443", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.004, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 66.283, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "66.283", + "raw_stderr": "1.887", + "latency": "1.206", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.887, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-small-2503/8855b9a5-f3be-4f3a-b9fa-47e7d8b627c0.json b/data/vals_ai/mistralai/mistral-small-2503/8855b9a5-f3be-4f3a-b9fa-47e7d8b627c0.json new file mode 100644 index 0000000000000000000000000000000000000000..8cee6b931b2df7658c1cbe97a28b8289f41710c7 --- /dev/null +++ b/data/vals_ai/mistralai/mistral-small-2503/8855b9a5-f3be-4f3a-b9fa-47e7d8b627c0.json @@ -0,0 +1,101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/mistralai_mistral-small-2503/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-small-2503", + "id": "mistralai/mistral-small-2503", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-small-2503", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 60.081, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "60.081", + "raw_stderr": "1.177", + "latency": "13.401", + "cost_per_test": "0.000265", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.177, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-small-2503/9090cece-29de-45d1-bd2f-68e7f3364d1c.json b/data/vals_ai/mistralai/mistral-small-2503/9090cece-29de-45d1-bd2f-68e7f3364d1c.json new file mode 100644 index 0000000000000000000000000000000000000000..1306848980256b66e69174e9c69476cc9d2c46e0 --- /dev/null +++ b/data/vals_ai/mistralai/mistral-small-2503/9090cece-29de-45d1-bd2f-68e7f3364d1c.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/mistralai_mistral-small-2503/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-small-2503", + "id": "mistralai/mistral-small-2503", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-small-2503", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 69.952, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "69.952", + "raw_stderr": "1.293", + "latency": "7.329", + "cost_per_test": "0.000477", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.293, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 61.208, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.208", + "raw_stderr": "0.956", + "latency": "8.313", + "cost_per_test": "0.000488", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.956, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 52.464, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.464", + "raw_stderr": "1.408", + "latency": "9.297", + "cost_per_test": "0.000499", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.408, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-small-2503/bbc36f11-803c-43bb-adbd-864092f92812.json b/data/vals_ai/mistralai/mistral-small-2503/bbc36f11-803c-43bb-adbd-864092f92812.json new file mode 100644 index 0000000000000000000000000000000000000000..28f2f258f9289da08601251311f33fcee70d06fd --- /dev/null +++ b/data/vals_ai/mistralai/mistral-small-2503/bbc36f11-803c-43bb-adbd-864092f92812.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/mistralai_mistral-small-2503/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-small-2503", + "id": "mistralai/mistral-small-2503", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-small-2503", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.172, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.172", + "raw_stderr": "1.46", + "latency": "2.868", + "cost_per_test": "0.000189", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.46, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 69.328, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.328", + "raw_stderr": "1.642", + "latency": "4.059", + "cost_per_test": "0.000198", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.642, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 66.608, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.608", + "raw_stderr": "1.402", + "latency": "5.882", + "cost_per_test": "0.000295", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.402, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.951, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.951", + "raw_stderr": "2.219", + "latency": "3.282", + "cost_per_test": "0.000219", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.219, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.578, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.578", + "raw_stderr": "1.518", + "latency": "2.73", + "cost_per_test": "0.000167", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.518, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 49.432, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "49.432", + "raw_stderr": "1.606", + "latency": "7.592", + "cost_per_test": "0.000322", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.606, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 70.782, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.782", + "raw_stderr": "1.59", + "latency": "2.018", + "cost_per_test": "0.000134", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.59, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 59.58, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "59.58", + "raw_stderr": "2.514", + "latency": "1.878", + "cost_per_test": "0.00021", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.514, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 45.504, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "45.504", + "raw_stderr": "1.501", + "latency": "2.897", + "cost_per_test": "0.000181", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.501, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.205, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.205", + "raw_stderr": "1.205", + "latency": "4.976", + "cost_per_test": "0.000274", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.205, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 66.45, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.45", + "raw_stderr": "1.553", + "latency": "1.757", + "cost_per_test": "0.00012", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.553, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 66.021, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.021", + "raw_stderr": "0.455", + "latency": "3.602", + "cost_per_test": "0.000203", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.455, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 55.311, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "55.311", + "raw_stderr": "2.227", + "latency": "2.479", + "cost_per_test": "0.000136", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.227, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 68.206, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "68.206", + "raw_stderr": "1.292", + "latency": "5.861", + "cost_per_test": "0.000262", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.292, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.183, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.183", + "raw_stderr": "1.568", + "latency": "2.151", + "cost_per_test": "0.000141", + "temperature": "0.7", + "max_output_tokens": "8192", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.568, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-small-2503/dec86c5f-c686-46ea-914e-a5a2538f9366.json b/data/vals_ai/mistralai/mistral-small-2503/dec86c5f-c686-46ea-914e-a5a2538f9366.json new file mode 100644 index 0000000000000000000000000000000000000000..8a2de4c3b85c3a40496ec2a12f57d635d2170d24 --- /dev/null +++ b/data/vals_ai/mistralai/mistral-small-2503/dec86c5f-c686-46ea-914e-a5a2538f9366.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/mistralai_mistral-small-2503/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-small-2503", + "id": "mistralai/mistral-small-2503", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-small-2503", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 52.448, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.448", + "raw_stderr": "1.705", + "latency": "8.023", + "cost_per_test": "0.000173", + "temperature": "0.7", + "max_output_tokens": "131000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.705, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 38.462, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "38.462", + "raw_stderr": "1.661", + "latency": "12.923", + "cost_per_test": "0.007811", + "temperature": "0.7", + "max_output_tokens": "131000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.661, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 44.173, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "44.173", + "raw_stderr": "0.972", + "latency": "11.291", + "cost_per_test": "0.004408", + "temperature": "0.7", + "max_output_tokens": "131000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.972, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:mistralai/mistral-small-2503:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 41.608, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "41.608", + "raw_stderr": "1.683", + "latency": "12.926", + "cost_per_test": "0.00524", + "temperature": "0.7", + "max_output_tokens": "131000", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.683, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 131000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/mistralai/mistral-small-2603/a21ac3d6-be38-4f3e-a400-557856ef7009.json b/data/vals_ai/mistralai/mistral-small-2603/a21ac3d6-be38-4f3e-a400-557856ef7009.json new file mode 100644 index 0000000000000000000000000000000000000000..5301ea372e4985074aca841591cd9e9a78869c7f --- /dev/null +++ b/data/vals_ai/mistralai/mistral-small-2603/a21ac3d6-be38-4f3e-a400-557856ef7009.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/mistralai_mistral-small-2603/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mistral-small-2603", + "id": "mistralai/mistral-small-2603", + "developer": "mistralai", + "additional_details": { + "vals_model_id": "mistralai/mistral-small-2603", + "vals_provider": "Mistral AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:mistralai/mistral-small-2603:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1814.681", + "cost_per_test": "6.473001", + "temperature": "0.7", + "max_output_tokens": "256000", + "reasoning_effort": "high", + "provider": "Mistral AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/Kimi-K2-Instruct/377e9e77-b550-410e-be13-f4da11e76e16.json b/data/vals_ai/moonshotai/Kimi-K2-Instruct/377e9e77-b550-410e-be13-f4da11e76e16.json new file mode 100644 index 0000000000000000000000000000000000000000..ca46611d9b032035b08ccef1fa6df6a9b4662944 --- /dev/null +++ b/data/vals_ai/moonshotai/Kimi-K2-Instruct/377e9e77-b550-410e-be13-f4da11e76e16.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/moonshotai_Kimi-K2-Instruct/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Kimi-K2-Instruct", + "id": "moonshotai/Kimi-K2-Instruct", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "together/moonshotai/Kimi-K2-Instruct", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 0.5, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0.5", + "raw_stderr": "0.456", + "latency": "1542.884", + "cost_per_test": "2.42593", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.456, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 2.0, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "2", + "raw_stderr": "1.179", + "latency": "1559.318", + "cost_per_test": "2.309216", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.179, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 1.25, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "1.25", + "raw_stderr": "0.632", + "latency": "1551.101", + "cost_per_test": "2.367573", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.632, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/Kimi-K2-Instruct/8c663c1c-41aa-4095-9b7a-28cc1a20dc18.json b/data/vals_ai/moonshotai/Kimi-K2-Instruct/8c663c1c-41aa-4095-9b7a-28cc1a20dc18.json new file mode 100644 index 0000000000000000000000000000000000000000..c34356d77dee826d670fb1adf23b62c95c8b48f7 --- /dev/null +++ b/data/vals_ai/moonshotai/Kimi-K2-Instruct/8c663c1c-41aa-4095-9b7a-28cc1a20dc18.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/moonshotai_Kimi-K2-Instruct/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Kimi-K2-Instruct", + "id": "moonshotai/Kimi-K2-Instruct", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "together/moonshotai/Kimi-K2-Instruct", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 97.205, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "97.205", + "raw_stderr": "0.919", + "latency": "17.242", + "cost_per_test": "0.002214", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.919, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 36.857, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "36.857", + "raw_stderr": "2.579", + "latency": "126.698", + "cost_per_test": "0.012321", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.579, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 77.285, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "77.285", + "raw_stderr": "2.141", + "latency": "56.013", + "cost_per_test": "0.006132", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.141, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 70.449, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "70.449", + "raw_stderr": "1.159", + "latency": "66.651", + "cost_per_test": "0.006889", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.159, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/Kimi-K2-Instruct/9518e733-b73b-45bd-a48c-6e338f340005.json b/data/vals_ai/moonshotai/Kimi-K2-Instruct/9518e733-b73b-45bd-a48c-6e338f340005.json new file mode 100644 index 0000000000000000000000000000000000000000..469d4cd0504b78225666430d274a5a5cb7dbc24d --- /dev/null +++ b/data/vals_ai/moonshotai/Kimi-K2-Instruct/9518e733-b73b-45bd-a48c-6e338f340005.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/moonshotai_Kimi-K2-Instruct/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Kimi-K2-Instruct", + "id": "moonshotai/Kimi-K2-Instruct", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "together/moonshotai/Kimi-K2-Instruct", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 49.417, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "49.417", + "raw_stderr": "1.707", + "latency": "46.214", + "cost_per_test": "0.111403", + "temperature": "0.3", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.707, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 53.03, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "53.03", + "raw_stderr": "1.704", + "latency": "33.801", + "cost_per_test": "0.001714", + "temperature": "0.3", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.704, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 50.388, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "50.388", + "raw_stderr": "0.985", + "latency": "40.894", + "cost_per_test": "0.059336", + "temperature": "0.3", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.985, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 48.718, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "48.718", + "raw_stderr": "1.706", + "latency": "42.666", + "cost_per_test": "0.064891", + "temperature": "0.3", + "max_output_tokens": "16384", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.706, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/Kimi-K2-Instruct/a7e771df-2de4-4498-90d3-8e39be72701b.json b/data/vals_ai/moonshotai/Kimi-K2-Instruct/a7e771df-2de4-4498-90d3-8e39be72701b.json new file mode 100644 index 0000000000000000000000000000000000000000..c8d163683d46d7e375b28685387a1da4e100cadc --- /dev/null +++ b/data/vals_ai/moonshotai/Kimi-K2-Instruct/a7e771df-2de4-4498-90d3-8e39be72701b.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/moonshotai_Kimi-K2-Instruct/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Kimi-K2-Instruct", + "id": "moonshotai/Kimi-K2-Instruct", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "together/moonshotai/Kimi-K2-Instruct", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 58.79, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.79", + "raw_stderr": "1.407", + "latency": "12.834", + "cost_per_test": "0.001543", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.407, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 70.196, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "70.196", + "raw_stderr": "0.895", + "latency": "12.834", + "cost_per_test": "0.001543", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.895, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 81.603, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "81.603", + "raw_stderr": "1.108", + "latency": "12.834", + "cost_per_test": "0.001543", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.108, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/Kimi-K2-Instruct/b3022af6-e120-40aa-a6e5-cc0f534cdcef.json b/data/vals_ai/moonshotai/Kimi-K2-Instruct/b3022af6-e120-40aa-a6e5-cc0f534cdcef.json new file mode 100644 index 0000000000000000000000000000000000000000..80b5f867c2d4cca793c20940a41c97f8445f6b90 --- /dev/null +++ b/data/vals_ai/moonshotai/Kimi-K2-Instruct/b3022af6-e120-40aa-a6e5-cc0f534cdcef.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/moonshotai_Kimi-K2-Instruct/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Kimi-K2-Instruct", + "id": "moonshotai/Kimi-K2-Instruct", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "together/moonshotai/Kimi-K2-Instruct", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 85.674, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.674", + "raw_stderr": "0.927", + "latency": "0.905", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.927, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.144, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.144", + "raw_stderr": "0.202", + "latency": "0.903", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.202, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.675, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.675", + "raw_stderr": "0.58", + "latency": "0.881", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.58, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.454, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.454", + "raw_stderr": "0.444", + "latency": "1.172", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.444, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.128, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.128", + "raw_stderr": "0.801", + "latency": "0.894", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.801, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 73.649, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "73.649", + "raw_stderr": "1.746", + "latency": "2.277", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.746, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/Kimi-K2-Instruct/c9c6485b-f7ac-4902-9349-3dbb710f174f.json b/data/vals_ai/moonshotai/Kimi-K2-Instruct/c9c6485b-f7ac-4902-9349-3dbb710f174f.json new file mode 100644 index 0000000000000000000000000000000000000000..ac67f57361c80eca2f703587f6081295adf7f97c --- /dev/null +++ b/data/vals_ai/moonshotai/Kimi-K2-Instruct/c9c6485b-f7ac-4902-9349-3dbb710f174f.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/moonshotai_Kimi-K2-Instruct/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Kimi-K2-Instruct", + "id": "moonshotai/Kimi-K2-Instruct", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "together/moonshotai/Kimi-K2-Instruct", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.098, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.098", + "raw_stderr": "1.115", + "latency": "23.45", + "cost_per_test": "0.003117", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.115, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.284, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.284", + "raw_stderr": "1.3", + "latency": "13.622", + "cost_per_test": "0.002074", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.3, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.067, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.067", + "raw_stderr": "1.14", + "latency": "27.216", + "cost_per_test": "0.003645", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.14, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.171, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.171", + "raw_stderr": "1.848", + "latency": "27.239", + "cost_per_test": "0.003545", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.848, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.308, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.308", + "raw_stderr": "1.219", + "latency": "15.862", + "cost_per_test": "0.002448", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.219, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 69.453, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.453", + "raw_stderr": "1.48", + "latency": "44.845", + "cost_per_test": "0.004529", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.48, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.506, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.506", + "raw_stderr": "1.46", + "latency": "15.73", + "cost_per_test": "0.002245", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.46, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 69.291, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.291", + "raw_stderr": "2.363", + "latency": "14.335", + "cost_per_test": "0.003214", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.363, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 62.67, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "62.67", + "raw_stderr": "1.458", + "latency": "21.29", + "cost_per_test": "0.00311", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.458, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.935, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.935", + "raw_stderr": "0.886", + "latency": "27.447", + "cost_per_test": "0.003777", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.886, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.922, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.922", + "raw_stderr": "1.364", + "latency": "10.042", + "cost_per_test": "0.001608", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.364, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.394, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.394", + "raw_stderr": "0.394", + "latency": "20.338", + "cost_per_test": "0.002836", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.394, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.747, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.747", + "raw_stderr": "1.97", + "latency": "9.06", + "cost_per_test": "0.001606", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.97, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.604, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.604", + "raw_stderr": "0.974", + "latency": "23.06", + "cost_per_test": "0.002917", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.974, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.456, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.456", + "raw_stderr": "1.346", + "latency": "11.527", + "cost_per_test": "0.001865", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.346, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/Kimi-K2-Instruct/dd50d7ee-e2c2-4f18-87e2-db0425d91f71.json b/data/vals_ai/moonshotai/Kimi-K2-Instruct/dd50d7ee-e2c2-4f18-87e2-db0425d91f71.json new file mode 100644 index 0000000000000000000000000000000000000000..b008a2fba0d1e5ae30042508cfb13dd384617d10 --- /dev/null +++ b/data/vals_ai/moonshotai/Kimi-K2-Instruct/dd50d7ee-e2c2-4f18-87e2-db0425d91f71.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/moonshotai_Kimi-K2-Instruct/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "Kimi-K2-Instruct", + "id": "moonshotai/Kimi-K2-Instruct", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "together/moonshotai/Kimi-K2-Instruct", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 69.192, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.192", + "raw_stderr": "3.281", + "latency": "30.413", + "cost_per_test": "0.004435", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.281, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 73.737, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.737", + "raw_stderr": "3.127", + "latency": "52.454", + "cost_per_test": "0.004527", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.127, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:together/moonshotai/Kimi-K2-Instruct:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 71.464, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.464", + "raw_stderr": "2.266", + "latency": "41.434", + "cost_per_test": "0.004481", + "temperature": "0.6", + "max_output_tokens": "32768", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.266, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2-thinking/3acc4d64-e805-4ecc-a000-3160b74963d9.json b/data/vals_ai/moonshotai/kimi-k2-thinking/3acc4d64-e805-4ecc-a000-3160b74963d9.json new file mode 100644 index 0000000000000000000000000000000000000000..9be84f706d14c87f93251db8e3c11f5dda7a2cc6 --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2-thinking/3acc4d64-e805-4ecc-a000-3160b74963d9.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/moonshotai_kimi-k2-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2-thinking", + "id": "moonshotai/kimi-k2-thinking", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2-thinking", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 78.788, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.788", + "raw_stderr": "3.127", + "latency": "546.998", + "cost_per_test": "0.022555", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.127, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 78.283, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.283", + "raw_stderr": "3.046", + "latency": "624.112", + "cost_per_test": "0.023588", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.046, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 78.536, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.536", + "raw_stderr": "2.183", + "latency": "585.555", + "cost_per_test": "0.023072", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.183, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2-thinking/3e5c24bf-f712-4e07-90f4-88e935bc9da0.json b/data/vals_ai/moonshotai/kimi-k2-thinking/3e5c24bf-f712-4e07-90f4-88e935bc9da0.json new file mode 100644 index 0000000000000000000000000000000000000000..f3d19c940ded9eebddfe583ee77ae7b86155deff --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2-thinking/3e5c24bf-f712-4e07-90f4-88e935bc9da0.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/moonshotai_kimi-k2-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2-thinking", + "id": "moonshotai/kimi-k2-thinking", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2-thinking", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 59.689, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.689", + "raw_stderr": "1.403", + "latency": "61.408", + "cost_per_test": "0.00913", + "temperature": "0.6", + "max_output_tokens": "32000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.403, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 71.709, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.709", + "raw_stderr": "0.878", + "latency": "61.408", + "cost_per_test": "0.00913", + "temperature": "0.6", + "max_output_tokens": "32000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.878, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 83.729, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "83.729", + "raw_stderr": "1.055", + "latency": "61.408", + "cost_per_test": "0.00913", + "temperature": "0.6", + "max_output_tokens": "32000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.055, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2-thinking/86e55dab-8135-4e3a-8512-fbd0074dc15d.json b/data/vals_ai/moonshotai/kimi-k2-thinking/86e55dab-8135-4e3a-8512-fbd0074dc15d.json new file mode 100644 index 0000000000000000000000000000000000000000..881e8a32dc65bc9c428f448e4e450d6cbc848f25 --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2-thinking/86e55dab-8135-4e3a-8512-fbd0074dc15d.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/moonshotai_kimi-k2-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2-thinking", + "id": "moonshotai/kimi-k2-thinking", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2-thinking", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.237, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.237", + "raw_stderr": "1.195", + "latency": "77.057", + "cost_per_test": "0.00166", + "temperature": "0.6", + "max_output_tokens": "32000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.195, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.565, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.565", + "raw_stderr": "1.349", + "latency": "89.584", + "cost_per_test": "0.001802", + "temperature": "0.6", + "max_output_tokens": "32000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.349, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.307, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.307", + "raw_stderr": "1.162", + "latency": "102.14", + "cost_per_test": "0.002417", + "temperature": "0.6", + "max_output_tokens": "32000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.162, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.902, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.902", + "raw_stderr": "1.919", + "latency": "112.665", + "cost_per_test": "0.002135", + "temperature": "0.6", + "max_output_tokens": "32000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.919, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.071, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.071", + "raw_stderr": "1.329", + "latency": "71.495", + "cost_per_test": "0.001543", + "temperature": "0.6", + "max_output_tokens": "32000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.329, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 69.35, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.35", + "raw_stderr": "1.516", + "latency": "169.443", + "cost_per_test": "0.003996", + "temperature": "0.6", + "max_output_tokens": "32000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.516, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.773, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.773", + "raw_stderr": "1.492", + "latency": "77.672", + "cost_per_test": "0.001618", + "temperature": "0.6", + "max_output_tokens": "32000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.492, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 72.178, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.178", + "raw_stderr": "2.302", + "latency": "87.832", + "cost_per_test": "0.002134", + "temperature": "0.6", + "max_output_tokens": "32000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.302, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 63.306, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "63.306", + "raw_stderr": "1.466", + "latency": "145.75", + "cost_per_test": "0.003132", + "temperature": "0.6", + "max_output_tokens": "32000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.466, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.636, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.636", + "raw_stderr": "0.92", + "latency": "76.794", + "cost_per_test": "0.002082", + "temperature": "0.6", + "max_output_tokens": "32000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.92, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.411, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.411", + "raw_stderr": "1.33", + "latency": "67.872", + "cost_per_test": "0.001378", + "temperature": "0.6", + "max_output_tokens": "32000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.33, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.068, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.068", + "raw_stderr": "0.396", + "latency": "94.771", + "cost_per_test": "0.002067", + "temperature": "0.6", + "max_output_tokens": "32000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.396, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.559, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.559", + "raw_stderr": "1.812", + "latency": "85.402", + "cost_per_test": "0.001517", + "temperature": "0.6", + "max_output_tokens": "32000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.812, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.451, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.451", + "raw_stderr": "1.014", + "latency": "98.509", + "cost_per_test": "0.0022", + "temperature": "0.6", + "max_output_tokens": "32000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.014, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.208, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.208", + "raw_stderr": "1.35", + "latency": "64.574", + "cost_per_test": "0.001324", + "temperature": "0.6", + "max_output_tokens": "32000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.35, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2-thinking/92de4cd2-51b8-4dfc-bbb0-4acbe35a9c6d.json b/data/vals_ai/moonshotai/kimi-k2-thinking/92de4cd2-51b8-4dfc-bbb0-4acbe35a9c6d.json new file mode 100644 index 0000000000000000000000000000000000000000..a6b448cc4060c9390c0879d563e430b3f334add5 --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2-thinking/92de4cd2-51b8-4dfc-bbb0-4acbe35a9c6d.json @@ -0,0 +1,426 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/moonshotai_kimi-k2-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2-thinking", + "id": "moonshotai/kimi-k2-thinking", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2-thinking", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 90.006, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.006", + "raw_stderr": "1.034", + "latency": "52.521", + "cost_per_test": "0.002334", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.034, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.98, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.98", + "raw_stderr": "0.207", + "latency": "42.595", + "cost_per_test": "0.002114", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.207, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.46, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.46", + "raw_stderr": "0.628", + "latency": "28.773", + "cost_per_test": "0.002043", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.628, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.201, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.201", + "raw_stderr": "0.459", + "latency": "43.991", + "cost_per_test": "0.002182", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.459, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 73.203, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "73.203", + "raw_stderr": "0.998", + "latency": "33.295", + "cost_per_test": "0.001787", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.998, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 76.355, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "76.355", + "raw_stderr": "1.666", + "latency": "62.773", + "cost_per_test": "0.002632", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.666, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2-thinking/955c0770-3e45-4645-9e1f-bad78eedfd0f.json b/data/vals_ai/moonshotai/kimi-k2-thinking/955c0770-3e45-4645-9e1f-bad78eedfd0f.json new file mode 100644 index 0000000000000000000000000000000000000000..bdd44a535e2f709f7dafcf78868e670814384164 --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2-thinking/955c0770-3e45-4645-9e1f-bad78eedfd0f.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/moonshotai_kimi-k2-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2-thinking", + "id": "moonshotai/kimi-k2-thinking", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2-thinking", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.137, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.137", + "raw_stderr": "0.754", + "latency": "297.043", + "cost_per_test": "0.008835", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.754, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 26.286, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "26.286", + "raw_stderr": "2.353", + "latency": "1894.45", + "cost_per_test": "0.055589", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.353, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 65.013, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "65.013", + "raw_stderr": "2.437", + "latency": "1260.379", + "cost_per_test": "0.036469", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.437, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 63.145, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "63.145", + "raw_stderr": "1.157", + "latency": "1150.624", + "cost_per_test": "0.033631", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.157, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2-thinking/ceaa20fa-cab0-4f19-92a9-26ccc9075aca.json b/data/vals_ai/moonshotai/kimi-k2-thinking/ceaa20fa-cab0-4f19-92a9-26ccc9075aca.json new file mode 100644 index 0000000000000000000000000000000000000000..50a558ea3eb86310142226158a24aed6e8bc002d --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2-thinking/ceaa20fa-cab0-4f19-92a9-26ccc9075aca.json @@ -0,0 +1,361 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/moonshotai_kimi-k2-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2-thinking", + "id": "moonshotai/kimi-k2-thinking", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2-thinking", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 19.048, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "19.048", + "raw_stderr": "6.133", + "latency": "2355.127", + "cost_per_test": "0.629091", + "temperature": "0.6", + "max_output_tokens": "32000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 6.133, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 53.64, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "53.64", + "raw_stderr": "3.093", + "latency": "1777.383", + "cost_per_test": "0.405479", + "temperature": "0.6", + "max_output_tokens": "32000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.093, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 78.351, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "78.351", + "raw_stderr": "2.965", + "latency": "1272.337", + "cost_per_test": "0.254182", + "temperature": "0.6", + "max_output_tokens": "32000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.965, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "2152.972", + "cost_per_test": "0.483343", + "temperature": "0.6", + "max_output_tokens": "32000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 60.2, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "60.2", + "raw_stderr": "2.191", + "latency": "1632.209", + "cost_per_test": "0.366026", + "temperature": "0.6", + "max_output_tokens": "32000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.191, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2-thinking/d4a45536-ee51-47ba-be89-bc26bca29795.json b/data/vals_ai/moonshotai/kimi-k2-thinking/d4a45536-ee51-47ba-be89-bc26bca29795.json new file mode 100644 index 0000000000000000000000000000000000000000..df5b1960f496d360f96d10449479450c02ea8851 --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2-thinking/d4a45536-ee51-47ba-be89-bc26bca29795.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/moonshotai_kimi-k2-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2-thinking", + "id": "moonshotai/kimi-k2-thinking", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2-thinking", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 59.091, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.091", + "raw_stderr": "1.679", + "latency": "67.814", + "cost_per_test": "0.003367", + "temperature": "0.6", + "max_output_tokens": "32000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.679, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 58.392, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.392", + "raw_stderr": "1.683", + "latency": "122.216", + "cost_per_test": "0.06934", + "temperature": "0.6", + "max_output_tokens": "32000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.683, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.567, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.567", + "raw_stderr": "0.962", + "latency": "93.297", + "cost_per_test": "0.038163", + "temperature": "0.6", + "max_output_tokens": "32000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.962, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:kimi/kimi-k2-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 64.219, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.219", + "raw_stderr": "1.636", + "latency": "89.86", + "cost_per_test": "0.041781", + "temperature": "0.6", + "max_output_tokens": "32000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.636, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "max_tokens": 32000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.5-thinking/0158d81f-a71f-4c19-a674-8fd314b771a4.json b/data/vals_ai/moonshotai/kimi-k2.5-thinking/0158d81f-a71f-4c19-a674-8fd314b771a4.json new file mode 100644 index 0000000000000000000000000000000000000000..4a9e20784f453334fe1e6cb68ae7cf08b11b7ee2 --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.5-thinking/0158d81f-a71f-4c19-a674-8fd314b771a4.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/moonshotai_kimi-k2.5-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.5-thinking", + "id": "moonshotai/kimi-k2.5-thinking", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.5-thinking", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 99.068, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "99.068", + "raw_stderr": "0.535", + "latency": "38.258", + "cost_per_test": "0.008048", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 64.286, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "64.286", + "raw_stderr": "2.561", + "latency": "415.345", + "cost_per_test": "0.079055", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.561, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 88.251, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "88.251", + "raw_stderr": "1.645", + "latency": "146.467", + "cost_per_test": "0.028865", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.645, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 83.868, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "83.868", + "raw_stderr": "1.03", + "latency": "200.023", + "cost_per_test": "0.038656", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.03, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.5-thinking/19e668ba-64f0-4cee-a481-eef24c7a5fc2.json b/data/vals_ai/moonshotai/kimi-k2.5-thinking/19e668ba-64f0-4cee-a481-eef24c7a5fc2.json new file mode 100644 index 0000000000000000000000000000000000000000..c650248dc0d9837df96506bd7831890d4ba4412c --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.5-thinking/19e668ba-64f0-4cee-a481-eef24c7a5fc2.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/moonshotai_kimi-k2.5-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.5-thinking", + "id": "moonshotai/kimi-k2.5-thinking", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.5-thinking", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 61.488, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.488", + "raw_stderr": "1.391", + "latency": "90.464", + "cost_per_test": "0.014221", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.391, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 74.202, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "74.202", + "raw_stderr": "0.846", + "latency": "90.464", + "cost_per_test": "0.014221", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.846, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 86.917, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "86.917", + "raw_stderr": "0.964", + "latency": "90.464", + "cost_per_test": "0.014221", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.964, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.5-thinking/34926fe8-9819-4199-8a1b-a6a53feebc5b.json b/data/vals_ai/moonshotai/kimi-k2.5-thinking/34926fe8-9819-4199-8a1b-a6a53feebc5b.json new file mode 100644 index 0000000000000000000000000000000000000000..afe4890f1b4faa8bc40ea0828a6a8121a79a9bb5 --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.5-thinking/34926fe8-9819-4199-8a1b-a6a53feebc5b.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/moonshotai_kimi-k2.5-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.5-thinking", + "id": "moonshotai/kimi-k2.5-thinking", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.5-thinking", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 76.442, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "76.442", + "raw_stderr": "1.986", + "latency": "147.155", + "cost_per_test": "0.02489", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.986, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.5-thinking/48c1f2ff-9a78-42a0-828f-72f5d1b8dc61.json b/data/vals_ai/moonshotai/kimi-k2.5-thinking/48c1f2ff-9a78-42a0-828f-72f5d1b8dc61.json new file mode 100644 index 0000000000000000000000000000000000000000..6f345182ea398a47ecfdd88550189c225db66a46 --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.5-thinking/48c1f2ff-9a78-42a0-828f-72f5d1b8dc61.json @@ -0,0 +1,101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/moonshotai_kimi-k2.5-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.5-thinking", + "id": "moonshotai/kimi-k2.5-thinking", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.5-thinking", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 84.335, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.335", + "raw_stderr": "0.874", + "latency": "110.011", + "cost_per_test": "0.019563", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.874, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.5-thinking/639138f6-f3e3-4818-9aa4-96efbd5cc4f7.json b/data/vals_ai/moonshotai/kimi-k2.5-thinking/639138f6-f3e3-4818-9aa4-96efbd5cc4f7.json new file mode 100644 index 0000000000000000000000000000000000000000..d37c609dba37f449766360a645202d25cdd01cb3 --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.5-thinking/639138f6-f3e3-4818-9aa4-96efbd5cc4f7.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/moonshotai_kimi-k2.5-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.5-thinking", + "id": "moonshotai/kimi-k2.5-thinking", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.5-thinking", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 74.96, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "74.96", + "raw_stderr": "1.221", + "latency": "103.935", + "cost_per_test": "0.018388", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.221, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 66.534, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.534", + "raw_stderr": "0.925", + "latency": "79.787", + "cost_per_test": "0.014918", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.925, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 58.108, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.108", + "raw_stderr": "1.391", + "latency": "55.639", + "cost_per_test": "0.011447", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.391, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.5-thinking/a789ef9a-1e3a-4a5f-b409-aa83417ab8c4.json b/data/vals_ai/moonshotai/kimi-k2.5-thinking/a789ef9a-1e3a-4a5f-b409-aa83417ab8c4.json new file mode 100644 index 0000000000000000000000000000000000000000..42a3f9ee47f200b9807b85bb21ceed98ee0cb40b --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.5-thinking/a789ef9a-1e3a-4a5f-b409-aa83417ab8c4.json @@ -0,0 +1,346 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/moonshotai_kimi-k2.5-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.5-thinking", + "id": "moonshotai/kimi-k2.5-thinking", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.5-thinking", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 44.444, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "44.444", + "raw_stderr": "6.236", + "latency": "175.937", + "cost_per_test": "0.026757", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 6.236, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 34.127, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "34.127", + "raw_stderr": "6.078", + "latency": "182.559", + "cost_per_test": "0.02803", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 6.078, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 17.284, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "17.284", + "raw_stderr": "5.24", + "latency": "210.765", + "cost_per_test": "0.055617", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 5.24, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 71.024, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "71.024", + "raw_stderr": "5.096", + "latency": "161.04", + "cost_per_test": "0.031422", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 5.096, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 49.865, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "49.865", + "raw_stderr": "3.363", + "latency": "173.179", + "cost_per_test": "0.028736", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.363, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.5-thinking/b6bf20e4-0b73-4dd2-b4d4-3af3f284b6ed.json b/data/vals_ai/moonshotai/kimi-k2.5-thinking/b6bf20e4-0b73-4dd2-b4d4-3af3f284b6ed.json new file mode 100644 index 0000000000000000000000000000000000000000..37757d146eaa21d5f599a5bae57964d9dcf7ed04 --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.5-thinking/b6bf20e4-0b73-4dd2-b4d4-3af3f284b6ed.json @@ -0,0 +1,361 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/moonshotai_kimi-k2.5-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.5-thinking", + "id": "moonshotai/kimi-k2.5-thinking", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.5-thinking", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 30.952, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "30.952", + "raw_stderr": "7.22", + "latency": "513.911", + "cost_per_test": "0.29421", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 7.22, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 66.284, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66.284", + "raw_stderr": "2.932", + "latency": "376.083", + "cost_per_test": "0.189295", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.932, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 84.536, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "84.536", + "raw_stderr": "2.603", + "latency": "250.94", + "cost_per_test": "0.110781", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.603, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "398.818", + "cost_per_test": "0.193336", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 70.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "70", + "raw_stderr": "2.051", + "latency": "339.242", + "cost_per_test": "0.167669", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.051, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.5-thinking/d00977a3-7a6c-4f36-b569-a602961acfb8.json b/data/vals_ai/moonshotai/kimi-k2.5-thinking/d00977a3-7a6c-4f36-b569-a602961acfb8.json new file mode 100644 index 0000000000000000000000000000000000000000..ee2de44eae1679d8b258797a27cb1ab356ac4731 --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.5-thinking/d00977a3-7a6c-4f36-b569-a602961acfb8.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/moonshotai_kimi-k2.5-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.5-thinking", + "id": "moonshotai/kimi-k2.5-thinking", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.5-thinking", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 12.167, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "12.167", + "raw_stderr": "5.041", + "latency": "3195.697", + "cost_per_test": "0.561257", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 5.041, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 23.167, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "23.167", + "raw_stderr": "11.074", + "latency": "4461.074", + "cost_per_test": "0.793265", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 11.074, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 17.667, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "17.667", + "raw_stderr": "6.084", + "latency": "3828.386", + "cost_per_test": "0.677261", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 6.084, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.5-thinking/d127549e-81f8-43be-b9b5-ff85e0b5215e.json b/data/vals_ai/moonshotai/kimi-k2.5-thinking/d127549e-81f8-43be-b9b5-ff85e0b5215e.json new file mode 100644 index 0000000000000000000000000000000000000000..19559949bb9114d5ed0d8b9c11ade41417f2b044 --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.5-thinking/d127549e-81f8-43be-b9b5-ff85e0b5215e.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/moonshotai_kimi-k2.5-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.5-thinking", + "id": "moonshotai/kimi-k2.5-thinking", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.5-thinking", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 17.536, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "17.536", + "raw_stderr": "3.258", + "latency": "2570.381", + "cost_per_test": "0.879292", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.258, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.5-thinking/dd344945-63f8-46ce-825d-1007a78437d3.json b/data/vals_ai/moonshotai/kimi-k2.5-thinking/dd344945-63f8-46ce-825d-1007a78437d3.json new file mode 100644 index 0000000000000000000000000000000000000000..8b241789ec896bda16c1c4d456196e49d2135ba7 --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.5-thinking/dd344945-63f8-46ce-825d-1007a78437d3.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/moonshotai_kimi-k2.5-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.5-thinking", + "id": "moonshotai/kimi-k2.5-thinking", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.5-thinking", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.026, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.026", + "raw_stderr": "0.96", + "latency": "28.765", + "cost_per_test": "0.005674", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.96, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.748, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.748", + "raw_stderr": "1.032", + "latency": "31.453", + "cost_per_test": "0.006461", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.032, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.254, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.254", + "raw_stderr": "0.847", + "latency": "56.981", + "cost_per_test": "0.011818", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.847, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.024, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.024", + "raw_stderr": "1.544", + "latency": "39.138", + "cost_per_test": "0.007604", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.544, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.863, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.863", + "raw_stderr": "1.083", + "latency": "24.636", + "cost_per_test": "0.004975", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.083, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.798, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.798", + "raw_stderr": "1.187", + "latency": "89.832", + "cost_per_test": "0.018821", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.187, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.807, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.807", + "raw_stderr": "1.377", + "latency": "32.16", + "cost_per_test": "0.006303", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.377, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.378, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.378", + "raw_stderr": "2.176", + "latency": "27.199", + "cost_per_test": "0.005502", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.176, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 70.3, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.3", + "raw_stderr": "1.377", + "latency": "50.75", + "cost_per_test": "0.009989", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.377, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 95.707, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "95.707", + "raw_stderr": "0.551", + "latency": "36.885", + "cost_per_test": "0.007629", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.551, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.333, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.333", + "raw_stderr": "1.226", + "latency": "24.362", + "cost_per_test": "0.004649", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.226, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.914, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.914", + "raw_stderr": "0.343", + "latency": "39.204", + "cost_per_test": "0.00793", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.343, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.164, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.164", + "raw_stderr": "1.714", + "latency": "26.244", + "cost_per_test": "0.005272", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.714, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.301, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.301", + "raw_stderr": "0.782", + "latency": "56.633", + "cost_per_test": "0.011735", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.782, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.09, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.09", + "raw_stderr": "1.225", + "latency": "23.815", + "cost_per_test": "0.00459", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.225, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.5-thinking/f45599a4-473b-4d8e-a0f9-f789e65c7103.json b/data/vals_ai/moonshotai/kimi-k2.5-thinking/f45599a4-473b-4d8e-a0f9-f789e65c7103.json new file mode 100644 index 0000000000000000000000000000000000000000..c71effbecfc635729f9ddbdff6efe78ed48f9fed --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.5-thinking/f45599a4-473b-4d8e-a0f9-f789e65c7103.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/moonshotai_kimi-k2.5-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.5-thinking", + "id": "moonshotai/kimi-k2.5-thinking", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.5-thinking", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 66.2, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.2", + "raw_stderr": "1.615", + "latency": "40.528", + "cost_per_test": "0.004139", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.615, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 68.415, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.415", + "raw_stderr": "1.587", + "latency": "132.479", + "cost_per_test": "0.109485", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.587, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 68.259, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.259", + "raw_stderr": "0.917", + "latency": "80.779", + "cost_per_test": "0.052412", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.917, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 70.163, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "70.163", + "raw_stderr": "1.562", + "latency": "69.331", + "cost_per_test": "0.043611", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.562, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.5-thinking/f69a11bb-50c7-4f99-a0e9-049350ba1a20.json b/data/vals_ai/moonshotai/kimi-k2.5-thinking/f69a11bb-50c7-4f99-a0e9-049350ba1a20.json new file mode 100644 index 0000000000000000000000000000000000000000..d0829e6f1babb93f99139a84d16654bc0fe060dd --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.5-thinking/f69a11bb-50c7-4f99-a0e9-049350ba1a20.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/moonshotai_kimi-k2.5-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.5-thinking", + "id": "moonshotai/kimi-k2.5-thinking", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.5-thinking", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 82.828, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.828", + "raw_stderr": "2.68", + "latency": "122.006", + "cost_per_test": "0.024807", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.68, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 85.354, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.354", + "raw_stderr": "3.324", + "latency": "181.68", + "cost_per_test": "0.035118", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.324, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:kimi/kimi-k2.5-thinking:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 84.091, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.091", + "raw_stderr": "2.135", + "latency": "151.843", + "cost_per_test": "0.029962", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.135, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.6/14f64459-b60c-40e9-81fe-e02fef2625cb.json b/data/vals_ai/moonshotai/kimi-k2.6/14f64459-b60c-40e9-81fe-e02fef2625cb.json new file mode 100644 index 0000000000000000000000000000000000000000..94df91ee117e5bc15a6667ef1472ec27f77a896f --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.6/14f64459-b60c-40e9-81fe-e02fef2625cb.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/moonshotai_kimi-k2.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.6", + "id": "moonshotai/kimi-k2.6", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.6", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 99.068, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "99.068", + "raw_stderr": "0.535", + "latency": "76.515", + "cost_per_test": "0.008417", + "top_p": "0.95", + "max_output_tokens": "256000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 71.429, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "71.429", + "raw_stderr": "2.415", + "latency": "1138.039", + "cost_per_test": "0.137579", + "top_p": "0.95", + "max_output_tokens": "256000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.415, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 89.817, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "89.817", + "raw_stderr": "1.545", + "latency": "372.634", + "cost_per_test": "0.050295", + "top_p": "0.95", + "max_output_tokens": "256000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.545, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 86.771, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "86.771", + "raw_stderr": "0.972", + "latency": "529.063", + "cost_per_test": "0.06543", + "top_p": "0.95", + "max_output_tokens": "256000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.972, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 256000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.6/1826368e-d61f-47c9-b7c4-0619ad0f215c.json b/data/vals_ai/moonshotai/kimi-k2.6/1826368e-d61f-47c9-b7c4-0619ad0f215c.json new file mode 100644 index 0000000000000000000000000000000000000000..602bb1b6ed7c12c3b9b8718e7f56517dcd6425f7 --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.6/1826368e-d61f-47c9-b7c4-0619ad0f215c.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/moonshotai_kimi-k2.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.6", + "id": "moonshotai/kimi-k2.6", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.6", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 37.891, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "37.891", + "raw_stderr": "4.912", + "latency": "2967.768", + "cost_per_test": "1.925351", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 4.912, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.6/19014c64-979a-4769-ad5e-b06ec7abd9d6.json b/data/vals_ai/moonshotai/kimi-k2.6/19014c64-979a-4769-ad5e-b06ec7abd9d6.json new file mode 100644 index 0000000000000000000000000000000000000000..d0cf6c5e715e7cefe7cb5377dab3a16c44299741 --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.6/19014c64-979a-4769-ad5e-b06ec7abd9d6.json @@ -0,0 +1,3286 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/hlab/moonshotai_kimi-k2.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Harvey's Legal Agent Benchmark", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.6", + "id": "moonshotai/kimi-k2.6", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.6", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "hlab:antitrust_competition:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1294.499", + "cost_per_test": "0.809466", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:arbitration_international_dispute_resolution:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1184.996", + "cost_per_test": "0.91661", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:banking_finance:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1314.303", + "cost_per_test": "0.794833", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:bankruptcy_restructuring:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1464.079", + "cost_per_test": "1.001911", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:capital_markets:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1097.435", + "cost_per_test": "0.64382", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:corporate_governance:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1311.711", + "cost_per_test": "0.826881", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:corporate_ma:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1282.05", + "cost_per_test": "0.73839", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Criteria Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Criteria Pass Rate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 74.063, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate", + "task_name": "Criteria Pass Rate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "74.063", + "raw_stderr": "0.527", + "latency": "1402.917", + "cost_per_test": "0.818741", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.527, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__antitrust_competition:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 63.768, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "63.768", + "raw_stderr": "2.899", + "latency": "1294.499", + "cost_per_test": "0.809466", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.899, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__arbitration_international_dispute_resolution:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 50.179, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "50.179", + "raw_stderr": "2.999", + "latency": "1184.996", + "cost_per_test": "0.91661", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.999, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__banking_finance:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 84.758, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.758", + "raw_stderr": "2.196", + "latency": "1314.303", + "cost_per_test": "0.794833", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.196, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__bankruptcy_restructuring:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 54.325, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "54.325", + "raw_stderr": "2.935", + "latency": "1464.079", + "cost_per_test": "1.001911", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.935, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__capital_markets:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 72.157, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "72.157", + "raw_stderr": "2.812", + "latency": "1097.435", + "cost_per_test": "0.64382", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.812, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_governance:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 86.986, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.986", + "raw_stderr": "1.972", + "latency": "1311.711", + "cost_per_test": "0.826881", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.972, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_ma:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 91.026, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "91.026", + "raw_stderr": "1.621", + "latency": "1282.05", + "cost_per_test": "0.73839", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.621, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__data_privacy_cybersecurity:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 74.558, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "74.558", + "raw_stderr": "2.594", + "latency": "985.892", + "cost_per_test": "0.697975", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.594, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__emerging_companies_venture_capital:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 81.57, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.57", + "raw_stderr": "2.269", + "latency": "1316.008", + "cost_per_test": "0.769426", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.269, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__employment_labor:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 73.034, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "73.034", + "raw_stderr": "2.721", + "latency": "1280.348", + "cost_per_test": "0.735629", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.721, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__energy_natural_resources:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 51.815, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "51.815", + "raw_stderr": "2.875", + "latency": "935.853", + "cost_per_test": "0.646701", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.875, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__environmental_esg:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 79.212, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.212", + "raw_stderr": "2.434", + "latency": "1709.282", + "cost_per_test": "0.994698", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.434, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__funds_asset_management:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 82.216, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.216", + "raw_stderr": "2.068", + "latency": "1785.657", + "cost_per_test": "1.192459", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.068, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__healthcare_life_sciences:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 63.366, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "63.366", + "raw_stderr": "2.772", + "latency": "1753.215", + "cost_per_test": "0.66545", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.772, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__immigration:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 78.182, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.182", + "raw_stderr": "2.495", + "latency": "1322.104", + "cost_per_test": "0.650266", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.495, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__insurance:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 80.22, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.22", + "raw_stderr": "2.415", + "latency": "1350.645", + "cost_per_test": "0.819488", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.415, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__intellectual_property:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 90.459, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.459", + "raw_stderr": "1.749", + "latency": "1400.063", + "cost_per_test": "0.808637", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.749, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__international_trade_sanctions:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 83.73, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.73", + "raw_stderr": "2.33", + "latency": "1627.862", + "cost_per_test": "0.83481", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.33, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__litigation_dispute_resolution:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 67.041, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "67.041", + "raw_stderr": "2.882", + "latency": "1618.179", + "cost_per_test": "0.908831", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.882, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__real_estate:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 81.924, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.924", + "raw_stderr": "2.081", + "latency": "1936.314", + "cost_per_test": "1.163256", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.081, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__structured_finance_securitization:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 71.429, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "71.429", + "raw_stderr": "2.549", + "latency": "1700.003", + "cost_per_test": "0.903082", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.549, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__tax:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 75.585, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "75.585", + "raw_stderr": "2.488", + "latency": "1716.131", + "cost_per_test": "0.890846", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.488, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__trusts_estates_private_client:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 73.282, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "73.282", + "raw_stderr": "2.739", + "latency": "1143.29", + "cost_per_test": "0.6436", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.739, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__white_collar_defense_investigations:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 64.983, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "64.983", + "raw_stderr": "2.773", + "latency": "1140.092", + "cost_per_test": "0.592721", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.773, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:data_privacy_cybersecurity:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "985.892", + "cost_per_test": "0.697975", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:emerging_companies_venture_capital:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1316.008", + "cost_per_test": "0.769426", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:employment_labor:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1280.348", + "cost_per_test": "0.735629", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:energy_natural_resources:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "935.853", + "cost_per_test": "0.646701", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:environmental_esg:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1709.282", + "cost_per_test": "0.994698", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:funds_asset_management:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1785.657", + "cost_per_test": "1.192459", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:healthcare_life_sciences:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1753.215", + "cost_per_test": "0.66545", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:immigration:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1322.104", + "cost_per_test": "0.650266", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:insurance:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1350.645", + "cost_per_test": "0.819488", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:intellectual_property:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "1400.063", + "cost_per_test": "0.808637", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:international_trade_sanctions:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1627.862", + "cost_per_test": "0.83481", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:litigation_dispute_resolution:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1618.179", + "cost_per_test": "0.908831", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:overall:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.overall", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Overall).", + "metric_id": "vals_ai.hlab.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 1.667, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "1.667", + "raw_stderr": "0.833", + "latency": "1402.917", + "cost_per_test": "0.818741", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.833, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:real_estate:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1936.314", + "cost_per_test": "1.163256", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:structured_finance_securitization:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1700.003", + "cost_per_test": "0.903082", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:tax:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1716.131", + "cost_per_test": "0.890846", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:trusts_estates_private_client:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1143.29", + "cost_per_test": "0.6436", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:white_collar_defense_investigations:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.hlab.white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1140.092", + "cost_per_test": "0.592721", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.6/239e6f5b-a408-4d0b-bdd1-26f2a148024c.json b/data/vals_ai/moonshotai/kimi-k2.6/239e6f5b-a408-4d0b-bdd1-26f2a148024c.json new file mode 100644 index 0000000000000000000000000000000000000000..a3561a60813aa7323ade6983ef46ad91d626d4fc --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.6/239e6f5b-a408-4d0b-bdd1-26f2a148024c.json @@ -0,0 +1,361 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/moonshotai_kimi-k2.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.6", + "id": "moonshotai/kimi-k2.6", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.6", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 40.476, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "40.476", + "raw_stderr": "7.666", + "latency": "850.848", + "cost_per_test": "0.491018", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 7.666, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 74.33, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "74.33", + "raw_stderr": "2.709", + "latency": "594.144", + "cost_per_test": "0.321761", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.709, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 87.629, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "87.629", + "raw_stderr": "2.37", + "latency": "340.865", + "cost_per_test": "0.163665", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.37, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "993.192", + "cost_per_test": "0.338884", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 76.2, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "76.2", + "raw_stderr": "1.906", + "latency": "519.829", + "cost_per_test": "0.27474", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.906, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.6/258cc054-2012-471f-9986-f48a802f72c7.json b/data/vals_ai/moonshotai/kimi-k2.6/258cc054-2012-471f-9986-f48a802f72c7.json new file mode 100644 index 0000000000000000000000000000000000000000..3302e04f2e2933b7379750a817f229b83d99a9a0 --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.6/258cc054-2012-471f-9986-f48a802f72c7.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/public-benefits-bench/moonshotai_kimi-k2.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Public Benefits Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "social mobility", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.6", + "id": "moonshotai/kimi-k2.6", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.6", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "public-benefits-bench:overall:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.public-benefits-bench.overall", + "source_data": { + "dataset_name": "Public Benefits Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Public Benefits Bench v1.1 (Overall).", + "metric_id": "vals_ai.public-benefits-bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "score_details": { + "score": 56.631, + "details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "social mobility", + "raw_score": "56.631", + "raw_stderr": "1.289", + "latency": "741.797", + "cost_per_test": "0.906347", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.289, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.6/27064304-652d-4b20-acdd-0ba56123df86.json b/data/vals_ai/moonshotai/kimi-k2.6/27064304-652d-4b20-acdd-0ba56123df86.json new file mode 100644 index 0000000000000000000000000000000000000000..11ed994c5abf519085963020602d6cae1260ec45 --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.6/27064304-652d-4b20-acdd-0ba56123df86.json @@ -0,0 +1,1152 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_research/moonshotai_kimi-k2.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Legal Research Bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.6", + "id": "moonshotai/kimi-k2.6", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.6", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_research:administrative_regulatory:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.legal_research.administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 18.182, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "18.182", + "raw_stderr": "4.784", + "latency": "1689.904", + "cost_per_test": "1.488125", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 4.784, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:business_commercial:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.legal_research.business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 20.93, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "20.93", + "raw_stderr": "6.277", + "latency": "1873.65", + "cost_per_test": "1.432344", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 6.277, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:civil_litigation:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.legal_research.civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 18.519, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "18.519", + "raw_stderr": "7.618", + "latency": "1733.848", + "cost_per_test": "1.440762", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 7.618, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:constitutional_civil_rights:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.legal_research.constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 21.053, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "21.053", + "raw_stderr": "6.702", + "latency": "1564.212", + "cost_per_test": "1.25835", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 6.702, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:criminal:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.legal_research.criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 14.035, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "14.035", + "raw_stderr": "4.642", + "latency": "1937.218", + "cost_per_test": "1.608029", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 4.642, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:family:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.legal_research.family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "2863.795", + "cost_per_test": "2.335677", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:health:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.legal_research.health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 30.0, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "30", + "raw_stderr": "8.51", + "latency": "1608.673", + "cost_per_test": "1.371176", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 8.51, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:immigration:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.legal_research.immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 18.182, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "18.182", + "raw_stderr": "12.197", + "latency": "1561.245", + "cost_per_test": "1.49714", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 12.197, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:overall:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.legal_research.overall", + "source_data": { + "dataset_name": "Legal Research Bench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Overall).", + "metric_id": "vals_ai.legal_research.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 15.865, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "legal", + "raw_score": "15.865", + "raw_stderr": "2.539", + "latency": "1853.133", + "cost_per_test": "1.520305", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.539, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass", + "source_data": { + "dataset_name": "Legal Research Bench - Weighted Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Weighted Pass).", + "metric_id": "vals_ai.legal_research.weighted_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 74.742, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass", + "task_name": "Weighted Pass", + "dataset_type": "private", + "industry": "legal", + "raw_score": "74.742", + "raw_stderr": "1.698", + "latency": "1853.133", + "cost_per_test": "1.520305", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.698, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__administrative_regulatory:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.weighted_pass__administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 75.531, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "75.531", + "raw_stderr": "3.082", + "latency": "1689.904", + "cost_per_test": "1.488125", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.082, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__business_commercial:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.weighted_pass__business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 78.436, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "78.436", + "raw_stderr": "3.571", + "latency": "1873.65", + "cost_per_test": "1.432344", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.571, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__civil_litigation:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.weighted_pass__civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 75.8, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "75.8", + "raw_stderr": "4.147", + "latency": "1733.848", + "cost_per_test": "1.440762", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 4.147, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__constitutional_civil_rights:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 79.224, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "79.224", + "raw_stderr": "3.12", + "latency": "1564.212", + "cost_per_test": "1.25835", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.12, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__criminal:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.weighted_pass__criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 75.057, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "75.057", + "raw_stderr": "2.901", + "latency": "1937.218", + "cost_per_test": "1.608029", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.901, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__family:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.weighted_pass__family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 60.959, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "60.959", + "raw_stderr": "6.121", + "latency": "2863.795", + "cost_per_test": "2.335677", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 6.121, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__health:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.weighted_pass__health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 81.669, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "81.669", + "raw_stderr": "3.886", + "latency": "1608.673", + "cost_per_test": "1.371176", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.886, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__immigration:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.weighted_pass__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 75.554, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "75.554", + "raw_stderr": "8.508", + "latency": "1561.245", + "cost_per_test": "1.49714", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 8.508, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.6/2852f5c3-bac9-4f37-87e0-13a1fe39a344.json b/data/vals_ai/moonshotai/kimi-k2.6/2852f5c3-bac9-4f37-87e0-13a1fe39a344.json new file mode 100644 index 0000000000000000000000000000000000000000..184f72780d5f23ac222b0d988bcfec2f9966ceaf --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.6/2852f5c3-bac9-4f37-87e0-13a1fe39a344.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/moonshotai_kimi-k2.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.6", + "id": "moonshotai/kimi-k2.6", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.6", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 83.333, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "83.333", + "raw_stderr": "8.333", + "latency": "20744.962", + "cost_per_test": "3.080492", + "top_p": "0.95", + "max_output_tokens": "256000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 8.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 25.556, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "25.556", + "raw_stderr": "2.222", + "latency": "2765.995", + "cost_per_test": "0.410732", + "top_p": "0.95", + "max_output_tokens": "256000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.222, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66.667", + "raw_stderr": "3.687", + "latency": "1508.725", + "cost_per_test": "0.224036", + "top_p": "0.95", + "max_output_tokens": "256000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.687, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 53.558, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "53.558", + "raw_stderr": "2.085", + "latency": "932.358", + "cost_per_test": "0.138449", + "top_p": "0.95", + "max_output_tokens": "256000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.085, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 256000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.6/314ed044-345d-47d7-b809-655c9bd451cd.json b/data/vals_ai/moonshotai/kimi-k2.6/314ed044-345d-47d7-b809-655c9bd451cd.json new file mode 100644 index 0000000000000000000000000000000000000000..25397a36e2f7c992bf6e20a3041e41b30db8250f --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.6/314ed044-345d-47d7-b809-655c9bd451cd.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/moonshotai_kimi-k2.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.6", + "id": "moonshotai/kimi-k2.6", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.6", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 62.715, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.715", + "raw_stderr": "1.383", + "latency": "267.484", + "cost_per_test": "0.029274", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.383, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 74.652, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "74.652", + "raw_stderr": "0.846", + "latency": "267.484", + "cost_per_test": "0.029274", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.846, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 86.59, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "86.59", + "raw_stderr": "0.974", + "latency": "267.484", + "cost_per_test": "0.029274", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.974, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.6/31d5ac41-6294-4b19-9274-e29a8aaec8f6.json b/data/vals_ai/moonshotai/kimi-k2.6/31d5ac41-6294-4b19-9274-e29a8aaec8f6.json new file mode 100644 index 0000000000000000000000000000000000000000..24f4a900db6d6e2eb136f6bd773cfff9e334fdb6 --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.6/31d5ac41-6294-4b19-9274-e29a8aaec8f6.json @@ -0,0 +1,346 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/moonshotai_kimi-k2.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.6", + "id": "moonshotai/kimi-k2.6", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.6", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 39.683, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "39.683", + "raw_stderr": "6.148", + "latency": "306.764", + "cost_per_test": "0.063769", + "top_p": "0.95", + "max_output_tokens": "96000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 6.148, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 96000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 42.857, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "42.857", + "raw_stderr": "6.36", + "latency": "284.178", + "cost_per_test": "0.060206", + "top_p": "0.95", + "max_output_tokens": "96000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 6.36, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 96000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 22.839, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "22.839", + "raw_stderr": "5.48", + "latency": "143.574", + "cost_per_test": "0.129492", + "top_p": "0.95", + "max_output_tokens": "96000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 5.48, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 96000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 68.132, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "68.132", + "raw_stderr": "5.254", + "latency": "221.926", + "cost_per_test": "0.080619", + "top_p": "0.95", + "max_output_tokens": "96000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 5.254, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 96000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 50.224, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "50.224", + "raw_stderr": "3.429", + "latency": "270.956", + "cost_per_test": "0.068198", + "top_p": "0.95", + "max_output_tokens": "96000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.429, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 96000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.6/4e2d8e28-233c-4640-aba1-129ff9f58929.json b/data/vals_ai/moonshotai/kimi-k2.6/4e2d8e28-233c-4640-aba1-129ff9f58929.json new file mode 100644 index 0000000000000000000000000000000000000000..c2cdce988aef1f7cf71967bc0fee233bf22ffe20 --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.6/4e2d8e28-233c-4640-aba1-129ff9f58929.json @@ -0,0 +1,398 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/moonshotai_kimi-k2.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.6", + "id": "moonshotai/kimi-k2.6", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.6", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 68.182, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "68.182", + "raw_stderr": "1.59", + "latency": "96.974", + "cost_per_test": "0.042064", + "top_p": "0.95", + "max_output_tokens": "48000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.59, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 48000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 44.932, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "44.932", + "raw_stderr": "1.709", + "latency": "1528.223", + "cost_per_test": "1.082722", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.709, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:overall:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 55.166, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "55.166", + "raw_stderr": "1.924", + "latency": "1184.397", + "cost_per_test": "0.702299", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.924, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 74.51, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "74.51", + "raw_stderr": "4.315", + "latency": "396.661", + "cost_per_test": "0.322908" + }, + "uncertainty": { + "standard_error": { + "value": 4.315, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 53.558, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "53.558", + "raw_stderr": "2.085", + "latency": "932.358", + "cost_per_test": "0.138449", + "top_p": "0.95", + "max_output_tokens": "256000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.085, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 42.323, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "42.323", + "raw_stderr": "8.392", + "latency": "2967.768", + "cost_per_test": "1.925351", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 8.392, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.6/6f98f222-5bcf-4c80-9185-0b6b9e49b73c.json b/data/vals_ai/moonshotai/kimi-k2.6/6f98f222-5bcf-4c80-9185-0b6b9e49b73c.json new file mode 100644 index 0000000000000000000000000000000000000000..28f68f506b8a43bb0c41476e049b7cb435522d29 --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.6/6f98f222-5bcf-4c80-9185-0b6b9e49b73c.json @@ -0,0 +1,426 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/moonshotai_kimi-k2.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.6", + "id": "moonshotai/kimi-k2.6", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.6", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 89.413, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.413", + "raw_stderr": "0.796", + "latency": "52.705", + "cost_per_test": "0.005986", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.796, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.197, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.197", + "raw_stderr": "0.198", + "latency": "39.793", + "cost_per_test": "0.005464", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.198, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 90.089, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.089", + "raw_stderr": "0.559", + "latency": "24.471", + "cost_per_test": "0.003533", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.559, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 84.738, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.738", + "raw_stderr": "0.449", + "latency": "50.067", + "cost_per_test": "0.0065", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.449, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.4, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.4", + "raw_stderr": "1.179", + "latency": "23.913", + "cost_per_test": "0.003209", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.179, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.592, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.592", + "raw_stderr": "1.635", + "latency": "109.453", + "cost_per_test": "0.01431", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.635, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.6/814c5f2b-3bc9-430d-ba5e-9362207b624b.json b/data/vals_ai/moonshotai/kimi-k2.6/814c5f2b-3bc9-430d-ba5e-9362207b624b.json new file mode 100644 index 0000000000000000000000000000000000000000..3e02649859c54bf5e1cf7d203eb3f3830007fc8e --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.6/814c5f2b-3bc9-430d-ba5e-9362207b624b.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/moonshotai_kimi-k2.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.6", + "id": "moonshotai/kimi-k2.6", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.6", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 89.899, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.899", + "raw_stderr": "2.277", + "latency": "500.4", + "cost_per_test": "0.075267", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.277, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 88.384, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.384", + "raw_stderr": "3.281", + "latency": "589.087", + "cost_per_test": "0.084991", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.281, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 89.142, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.142", + "raw_stderr": "1.997", + "latency": "544.744", + "cost_per_test": "0.080129", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.997, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.6/81fce48d-1f00-4a88-9e16-d9d19338e261.json b/data/vals_ai/moonshotai/kimi-k2.6/81fce48d-1f00-4a88-9e16-d9d19338e261.json new file mode 100644 index 0000000000000000000000000000000000000000..0b49c6ec329db4a4b223b6ed31b4f0ef976796f4 --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.6/81fce48d-1f00-4a88-9e16-d9d19338e261.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/moonshotai_kimi-k2.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.6", + "id": "moonshotai/kimi-k2.6", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.6", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.724, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.724", + "raw_stderr": "0.906", + "latency": "96.93", + "cost_per_test": "0.010782", + "top_p": "0.95", + "max_output_tokens": "96000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.906, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 96000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.635, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.635", + "raw_stderr": "0.986", + "latency": "102.866", + "cost_per_test": "0.01534", + "top_p": "0.95", + "max_output_tokens": "96000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.986, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 96000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.491, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.491", + "raw_stderr": "0.808", + "latency": "211.305", + "cost_per_test": "0.02316", + "top_p": "0.95", + "max_output_tokens": "96000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.808, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 96000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.22, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.22", + "raw_stderr": "1.398", + "latency": "154.119", + "cost_per_test": "0.017371", + "top_p": "0.95", + "max_output_tokens": "96000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.398, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 96000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.929, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.929", + "raw_stderr": "1.036", + "latency": "82.769", + "cost_per_test": "0.01056", + "top_p": "0.95", + "max_output_tokens": "96000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.036, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 96000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.1, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.1", + "raw_stderr": "1.175", + "latency": "463.727", + "cost_per_test": "0.053424", + "top_p": "0.95", + "max_output_tokens": "96000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.175, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 96000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.685, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.685", + "raw_stderr": "1.38", + "latency": "123.418", + "cost_per_test": "0.013682", + "top_p": "0.95", + "max_output_tokens": "96000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.38, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 96000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.478, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.478", + "raw_stderr": "2.105", + "latency": "96.434", + "cost_per_test": "0.011629", + "top_p": "0.95", + "max_output_tokens": "96000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.105, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 96000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.296, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.296", + "raw_stderr": "1.323", + "latency": "272.45", + "cost_per_test": "0.028894", + "top_p": "0.95", + "max_output_tokens": "96000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.323, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 96000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 95.633, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "95.633", + "raw_stderr": "0.556", + "latency": "140.155", + "cost_per_test": "0.014751", + "top_p": "0.95", + "max_output_tokens": "96000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.556, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 96000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.957, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.957", + "raw_stderr": "1.176", + "latency": "98.047", + "cost_per_test": "0.011346", + "top_p": "0.95", + "max_output_tokens": "96000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.176, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 96000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.572, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.572", + "raw_stderr": "0.33", + "latency": "156.347", + "cost_per_test": "0.018214", + "top_p": "0.95", + "max_output_tokens": "96000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.33, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 96000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.569, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.569", + "raw_stderr": "1.617", + "latency": "70.384", + "cost_per_test": "0.011206", + "top_p": "0.95", + "max_output_tokens": "96000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.617, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 96000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.071, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.071", + "raw_stderr": "0.753", + "latency": "186.541", + "cost_per_test": "0.022914", + "top_p": "0.95", + "max_output_tokens": "96000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.753, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 96000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.223, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.223", + "raw_stderr": "1.211", + "latency": "89.709", + "cost_per_test": "0.009941", + "top_p": "0.95", + "max_output_tokens": "96000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.211, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 96000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.6/862497ed-c16e-4c8b-9da1-875196a054a3.json b/data/vals_ai/moonshotai/kimi-k2.6/862497ed-c16e-4c8b-9da1-875196a054a3.json new file mode 100644 index 0000000000000000000000000000000000000000..604eac9bc99be5063b7073f600ed92c91cf4a96e --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.6/862497ed-c16e-4c8b-9da1-875196a054a3.json @@ -0,0 +1,522 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_multimodal_index/moonshotai_kimi-k2.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Multimodal Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.6", + "id": "moonshotai/kimi-k2.6", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.6", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_multimodal_index:corp_fin_v2:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.vals_multimodal_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Multimodal Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (CorpFin v2).", + "metric_id": "vals_ai.vals_multimodal_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 68.182, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "68.182", + "raw_stderr": "1.59", + "latency": "96.974", + "cost_per_test": "0.042064", + "top_p": "0.95", + "max_output_tokens": "48000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.59, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 48000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:finance_agent:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.vals_multimodal_index.finance_agent", + "source_data": { + "dataset_name": "Vals Multimodal Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_multimodal_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 44.932, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "44.932", + "raw_stderr": "1.709", + "latency": "1528.223", + "cost_per_test": "1.082722", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.709, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:mortgage_tax:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.vals_multimodal_index.mortgage_tax", + "source_data": { + "dataset_name": "Vals Multimodal Index - Mortgage Tax", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "mortgage_tax", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Mortgage Tax).", + "metric_id": "vals_ai.vals_multimodal_index.mortgage_tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 65.818, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "mortgage_tax", + "task_name": "Mortgage Tax", + "dataset_type": "private", + "industry": "index", + "raw_score": "65.818", + "raw_stderr": "0.93", + "latency": "109.76", + "cost_per_test": "0.02869", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.93, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:overall:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.vals_multimodal_index.overall", + "source_data": { + "dataset_name": "Vals Multimodal Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Overall).", + "metric_id": "vals_ai.vals_multimodal_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 56.434, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "56.434", + "raw_stderr": "1.735", + "latency": "900.386", + "cost_per_test": "0.515483", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.735, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:sage:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.vals_multimodal_index.sage", + "source_data": { + "dataset_name": "Vals Multimodal Index - SAGE", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "sage", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SAGE).", + "metric_id": "vals_ai.vals_multimodal_index.sage.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 50.224, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "sage", + "task_name": "SAGE", + "dataset_type": "private", + "industry": "index", + "raw_score": "50.224", + "raw_stderr": "3.429", + "latency": "270.956", + "cost_per_test": "0.068198", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.429, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:swebench:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.vals_multimodal_index.swebench", + "source_data": { + "dataset_name": "Vals Multimodal Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SWE-bench).", + "metric_id": "vals_ai.vals_multimodal_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 74.51, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "74.51", + "raw_stderr": "4.315", + "latency": "396.661", + "cost_per_test": "0.322908" + }, + "uncertainty": { + "standard_error": { + "value": 4.315, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:terminal_bench_2_1:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.vals_multimodal_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Multimodal Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_multimodal_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 53.558, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "53.558", + "raw_stderr": "2.085", + "latency": "932.358", + "cost_per_test": "0.138449", + "top_p": "0.95", + "max_output_tokens": "256000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.085, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:vibe_code_bench:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.vals_multimodal_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Multimodal Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_multimodal_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 42.323, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "42.323", + "raw_stderr": "8.392", + "latency": "2967.768", + "cost_per_test": "1.925351", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 8.392, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.6/93343464-a501-4379-9510-433220fb0408.json b/data/vals_ai/moonshotai/kimi-k2.6/93343464-a501-4379-9510-433220fb0408.json new file mode 100644 index 0000000000000000000000000000000000000000..f33c11a4a47b314165678e38b443bfc32562180a --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.6/93343464-a501-4379-9510-433220fb0408.json @@ -0,0 +1,656 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/emb/moonshotai_kimi-k2.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - EMB", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.6", + "id": "moonshotai/kimi-k2.6", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.6", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "emb:comps:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.emb.comps", + "source_data": { + "dataset_name": "EMB - Comps", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "comps", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Comps).", + "metric_id": "vals_ai.emb.comps.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 61.395, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "comps", + "task_name": "Comps", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.395", + "raw_stderr": "4.513", + "latency": "2652.675", + "cost_per_test": "2.521027", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 4.513, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:dataroom:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.emb.dataroom", + "source_data": { + "dataset_name": "EMB - Dataroom", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dataroom", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Dataroom).", + "metric_id": "vals_ai.emb.dataroom.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 75.069, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dataroom", + "task_name": "Dataroom", + "dataset_type": "private", + "industry": "finance", + "raw_score": "75.069", + "raw_stderr": "6.099", + "latency": "1788.453", + "cost_per_test": "1.852169", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 6.099, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:dcf:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.emb.dcf", + "source_data": { + "dataset_name": "EMB - DCF", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dcf", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (DCF).", + "metric_id": "vals_ai.emb.dcf.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 49.675, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dcf", + "task_name": "DCF", + "dataset_type": "private", + "industry": "finance", + "raw_score": "49.675", + "raw_stderr": "6.323", + "latency": "2553.91", + "cost_per_test": "2.595617", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 6.323, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:lbo:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.emb.lbo", + "source_data": { + "dataset_name": "EMB - LBO", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "lbo", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (LBO).", + "metric_id": "vals_ai.emb.lbo.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 43.117, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "lbo", + "task_name": "LBO", + "dataset_type": "private", + "industry": "finance", + "raw_score": "43.117", + "raw_stderr": "5.123", + "latency": "3314.566", + "cost_per_test": "3.032183", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 5.123, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:ma:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.emb.ma", + "source_data": { + "dataset_name": "EMB - M&A", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "ma", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (M&A).", + "metric_id": "vals_ai.emb.ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 64.205, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "ma", + "task_name": "M&A", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.205", + "raw_stderr": "5.335", + "latency": "2075.852", + "cost_per_test": "1.570029", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 5.335, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:misc:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.emb.misc", + "source_data": { + "dataset_name": "EMB - Misc", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "misc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Misc).", + "metric_id": "vals_ai.emb.misc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 54.599, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "misc", + "task_name": "Misc", + "dataset_type": "private", + "industry": "finance", + "raw_score": "54.599", + "raw_stderr": "6.131", + "latency": "2093.729", + "cost_per_test": "1.835205", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 6.131, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:operating:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.emb.operating", + "source_data": { + "dataset_name": "EMB - Operating", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "operating", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Operating).", + "metric_id": "vals_ai.emb.operating.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 56.923, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "operating", + "task_name": "Operating", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.923", + "raw_stderr": "7.443", + "latency": "2067.12", + "cost_per_test": "1.804151", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 7.443, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:overall:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.emb.overall", + "source_data": { + "dataset_name": "EMB - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Overall).", + "metric_id": "vals_ai.emb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 57.855, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.855", + "raw_stderr": "2.739", + "latency": "2387.544", + "cost_per_test": "2.195977", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.739, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:scratch:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.emb.scratch", + "source_data": { + "dataset_name": "EMB - Scratch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "scratch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Scratch).", + "metric_id": "vals_ai.emb.scratch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 59.678, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "scratch", + "task_name": "Scratch", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.678", + "raw_stderr": "2.915", + "latency": "2428.134", + "cost_per_test": "2.154897", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.915, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:template:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.emb.template", + "source_data": { + "dataset_name": "EMB - Template", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "template", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Template).", + "metric_id": "vals_ai.emb.template.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 55.683, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "template", + "task_name": "Template", + "dataset_type": "private", + "industry": "finance", + "raw_score": "55.683", + "raw_stderr": "3.761", + "latency": "2346.955", + "cost_per_test": "2.237057", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.761, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.6/94df650a-876b-455d-be6e-be40c1b095d5.json b/data/vals_ai/moonshotai/kimi-k2.6/94df650a-876b-455d-be6e-be40c1b095d5.json new file mode 100644 index 0000000000000000000000000000000000000000..7a536f8de8d5074356101078ee8cd1ed6b1c8958 --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.6/94df650a-876b-455d-be6e-be40c1b095d5.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/moonshotai_kimi-k2.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.6", + "id": "moonshotai/kimi-k2.6", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.6", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 64.802, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.802", + "raw_stderr": "1.63", + "latency": "38.306", + "cost_per_test": "0.006637", + "top_p": "0.95", + "max_output_tokens": "48000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.63, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 48000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 67.249, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.249", + "raw_stderr": "1.602", + "latency": "103.491", + "cost_per_test": "0.171139", + "top_p": "0.95", + "max_output_tokens": "48000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.602, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 48000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 66.744, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.744", + "raw_stderr": "0.928", + "latency": "79.59", + "cost_per_test": "0.07328", + "top_p": "0.95", + "max_output_tokens": "48000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.928, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 48000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 68.182, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.182", + "raw_stderr": "1.59", + "latency": "96.974", + "cost_per_test": "0.042064", + "top_p": "0.95", + "max_output_tokens": "48000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.59, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 48000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.6/b877d509-00f4-4cc6-a0ac-a3efac5ad667.json b/data/vals_ai/moonshotai/kimi-k2.6/b877d509-00f4-4cc6-a0ac-a3efac5ad667.json new file mode 100644 index 0000000000000000000000000000000000000000..6caac0935b73f9c672278648ea11413f182e527c --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.6/b877d509-00f4-4cc6-a0ac-a3efac5ad667.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/programbench/moonshotai_kimi-k2.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProgramBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.6", + "id": "moonshotai/kimi-k2.6", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.6", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "programbench:almost:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.programbench.almost", + "source_data": { + "dataset_name": "ProgramBench - Almost Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "almost", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Almost Resolved).", + "metric_id": "vals_ai.programbench.almost.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.5, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "almost", + "task_name": "Almost Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0.5", + "raw_stderr": "0.5", + "latency": "4794.524", + "cost_per_test": "6.993877", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.5, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "programbench:overall:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.programbench.overall", + "source_data": { + "dataset_name": "ProgramBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Overall).", + "metric_id": "vals_ai.programbench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "4794.524", + "cost_per_test": "6.993877", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "programbench:partial:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.programbench.partial", + "source_data": { + "dataset_name": "ProgramBench - Raw Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "partial", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Raw Pass Rate).", + "metric_id": "vals_ai.programbench.partial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 47.952, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "partial", + "task_name": "Raw Pass Rate", + "dataset_type": "public", + "industry": "coding", + "raw_score": "47.952", + "raw_stderr": "1.955", + "latency": "4794.524", + "cost_per_test": "6.993877", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.955, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "programbench:strict:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.programbench.strict", + "source_data": { + "dataset_name": "ProgramBench - Fully Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "strict", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Fully Resolved).", + "metric_id": "vals_ai.programbench.strict.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "strict", + "task_name": "Fully Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "4794.524", + "cost_per_test": "6.993877", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.6/c4a039ab-f587-40f3-af9a-d72b17c2595f.json b/data/vals_ai/moonshotai/kimi-k2.6/c4a039ab-f587-40f3-af9a-d72b17c2595f.json new file mode 100644 index 0000000000000000000000000000000000000000..f8627f929a1530bf7e65d7192fd3c05cfeaef2d8 --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.6/c4a039ab-f587-40f3-af9a-d72b17c2595f.json @@ -0,0 +1,195 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/cyber/moonshotai_kimi-k2.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CyberBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "beta", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.6", + "id": "moonshotai/kimi-k2.6", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.6", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "cyber:overall:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.cyber.overall", + "source_data": { + "dataset_name": "CyberBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (Overall).", + "metric_id": "vals_ai.cyber.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 74.351, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "beta", + "raw_score": "74.351", + "raw_stderr": "6.178", + "latency": "1186.456", + "cost_per_test": "1.045297", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 6.178, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "cyber:patch:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.cyber.patch", + "source_data": { + "dataset_name": "CyberBench - Patch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "patch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (Patch).", + "metric_id": "vals_ai.cyber.patch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 71.429, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "patch", + "task_name": "Patch", + "dataset_type": "private", + "industry": "beta", + "raw_score": "71.429", + "raw_stderr": "6.037", + "latency": "1321.537", + "cost_per_test": "0.817664", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 6.037, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "cyber:poc:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.cyber.poc", + "source_data": { + "dataset_name": "CyberBench - PoC", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "poc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (PoC).", + "metric_id": "vals_ai.cyber.poc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 77.273, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "poc", + "task_name": "PoC", + "dataset_type": "private", + "industry": "beta", + "raw_score": "77.273", + "raw_stderr": "6.318", + "latency": "1051.374", + "cost_per_test": "1.272931", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 6.318, + "method": "vals_reported" + } + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.6/c6c33f9b-f376-444c-a523-15aaa9844add.json b/data/vals_ai/moonshotai/kimi-k2.6/c6c33f9b-f376-444c-a523-15aaa9844add.json new file mode 100644 index 0000000000000000000000000000000000000000..6f9585505ec7f0dab873a37857ff190f4b2f4bb6 --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.6/c6c33f9b-f376-444c-a523-15aaa9844add.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/code-migration/moonshotai_kimi-k2.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Code Migration", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.6", + "id": "moonshotai/kimi-k2.6", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.6", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "code-migration:cli:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.code-migration.cli", + "source_data": { + "dataset_name": "Code Migration - CLI", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cli", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (CLI).", + "metric_id": "vals_ai.code-migration.cli.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 17.977, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cli", + "task_name": "CLI", + "dataset_type": "private", + "industry": "coding", + "raw_score": "17.977", + "raw_stderr": "1.675", + "latency": "6170.995", + "cost_per_test": "6.079502", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.675, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "code-migration:cobol:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.code-migration.cobol", + "source_data": { + "dataset_name": "Code Migration - COBOL", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cobol", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (COBOL).", + "metric_id": "vals_ai.code-migration.cobol.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 57.143, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cobol", + "task_name": "COBOL", + "dataset_type": "private", + "industry": "coding", + "raw_score": "57.143", + "raw_stderr": "15.793", + "latency": "3081.279", + "cost_per_test": "2.209705", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 15.793, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "code-migration:code_quality:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.code-migration.code_quality", + "source_data": { + "dataset_name": "Code Migration - Code Quality", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "code_quality", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Code Quality).", + "metric_id": "vals_ai.code-migration.code_quality.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 3.149, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "code_quality", + "task_name": "Code Quality", + "dataset_type": "private", + "industry": "coding", + "raw_score": "3.149", + "raw_stderr": "0.071", + "latency": "5398.566", + "cost_per_test": "5.112053", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.071, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "code-migration:overall:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.code-migration.overall", + "source_data": { + "dataset_name": "Code Migration - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Overall).", + "metric_id": "vals_ai.code-migration.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 27.768, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "27.768", + "raw_stderr": "4.144", + "latency": "5398.566", + "cost_per_test": "5.112053", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 4.144, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.6/cf8b1ebc-c98d-4979-9be2-415ae9d2b94b.json b/data/vals_ai/moonshotai/kimi-k2.6/cf8b1ebc-c98d-4979-9be2-415ae9d2b94b.json new file mode 100644 index 0000000000000000000000000000000000000000..ebca04768ddfaf1e6a8369668870e3afdc02d2a9 --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.6/cf8b1ebc-c98d-4979-9be2-415ae9d2b94b.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/moonshotai_kimi-k2.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.6", + "id": "moonshotai/kimi-k2.6", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.6", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 74.483, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "74.483", + "raw_stderr": "1.229", + "latency": "130.666", + "cost_per_test": "0.037373", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.229, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 65.818, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.818", + "raw_stderr": "0.93", + "latency": "109.76", + "cost_per_test": "0.02869", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.93, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 57.154, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.154", + "raw_stderr": "1.395", + "latency": "88.854", + "cost_per_test": "0.020008", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.395, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.6/e3f35e4d-d616-417f-b923-79acb3a10834.json b/data/vals_ai/moonshotai/kimi-k2.6/e3f35e4d-d616-417f-b923-79acb3a10834.json new file mode 100644 index 0000000000000000000000000000000000000000..b41762d69e2326a8f6f1b8bee78214f0a4bc461c --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.6/e3f35e4d-d616-417f-b923-79acb3a10834.json @@ -0,0 +1,101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/moonshotai_kimi-k2.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.6", + "id": "moonshotai/kimi-k2.6", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.6", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 86.301, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.301", + "raw_stderr": "0.827", + "latency": "241.908", + "cost_per_test": "0.034867", + "top_p": "0.95", + "max_output_tokens": "96000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.827, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 96000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.6/e4583558-5378-4ebd-9ff0-caac1e962ac7.json b/data/vals_ai/moonshotai/kimi-k2.6/e4583558-5378-4ebd-9ff0-caac1e962ac7.json new file mode 100644 index 0000000000000000000000000000000000000000..0425c890e44316c27baa3d1a2c113d0ddf1f6c27 --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.6/e4583558-5378-4ebd-9ff0-caac1e962ac7.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/moonshotai_kimi-k2.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.6", + "id": "moonshotai/kimi-k2.6", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.6", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 78.149, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "78.149", + "raw_stderr": "1.792", + "latency": "494.142", + "cost_per_test": "0.055962", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.792, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.6/ec8ac69b-788e-4542-93ec-277ef5b89998.json b/data/vals_ai/moonshotai/kimi-k2.6/ec8ac69b-788e-4542-93ec-277ef5b89998.json new file mode 100644 index 0000000000000000000000000000000000000000..04789d390acd98b69cad902077f68ca94bb7df25 --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.6/ec8ac69b-788e-4542-93ec-277ef5b89998.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/moonshotai_kimi-k2.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.6", + "id": "moonshotai/kimi-k2.6", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.6", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 16.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "16", + "raw_stderr": "3.666", + "latency": "1386.754", + "cost_per_test": "0.295811", + "top_p": "0.95", + "max_output_tokens": "48000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.666, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 48000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.6/f9feab0e-96ac-4180-b74c-9eabe34484ba.json b/data/vals_ai/moonshotai/kimi-k2.6/f9feab0e-96ac-4180-b74c-9eabe34484ba.json new file mode 100644 index 0000000000000000000000000000000000000000..f8d210c445cb89b25552d8652c05acaea0012de0 --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.6/f9feab0e-96ac-4180-b74c-9eabe34484ba.json @@ -0,0 +1,718 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/moonshotai_kimi-k2.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.6", + "id": "moonshotai/kimi-k2.6", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.6", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 37.934, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "37.934", + "raw_stderr": "2.645", + "latency": "1483.007", + "cost_per_test": "1.064336", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.645, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 32.584, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "32.584", + "raw_stderr": "0.465", + "latency": "1629.639", + "cost_per_test": "1.141587", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.465, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 41.069, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "41.069", + "raw_stderr": "0.105", + "latency": "2307.494", + "cost_per_test": "1.642419", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.105, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 52.124, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.124", + "raw_stderr": "2.522", + "latency": "1512.854", + "cost_per_test": "1.275044", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.522, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 63.824, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.824", + "raw_stderr": "0.746", + "latency": "847.882", + "cost_per_test": "0.491388", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.746, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 14.201, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "14.201", + "raw_stderr": "1.558", + "latency": "1736.611", + "cost_per_test": "1.336633", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.558, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 57.295, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.295", + "raw_stderr": "2.228", + "latency": "1520.92", + "cost_per_test": "1.003368", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.228, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 65.13, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.13", + "raw_stderr": "2.429", + "latency": "1549.406", + "cost_per_test": "1.111564", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.429, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 53.053, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "53.053", + "raw_stderr": "3.915", + "latency": "1002.477", + "cost_per_test": "0.577509", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.915, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:overall:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 44.899, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "44.899", + "raw_stderr": "0.733", + "latency": "1629.639", + "cost_per_test": "1.141587", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.733, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:kimi/kimi-k2.6:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 19.459, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "19.459", + "raw_stderr": "1.755", + "latency": "2890.369", + "cost_per_test": "1.843727", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.755, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.7-code/4f518176-482d-4573-a86a-65cbfe97f9fc.json b/data/vals_ai/moonshotai/kimi-k2.7-code/4f518176-482d-4573-a86a-65cbfe97f9fc.json new file mode 100644 index 0000000000000000000000000000000000000000..e6b51ee29b63978398544243a8cfb140c3f9908c --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.7-code/4f518176-482d-4573-a86a-65cbfe97f9fc.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/code-migration/moonshotai_kimi-k2.7-code/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Code Migration", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.7-code", + "id": "moonshotai/kimi-k2.7-code", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.7-code", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "code-migration:cli:kimi/kimi-k2.7-code:score", + "evaluation_name": "vals_ai.code-migration.cli", + "source_data": { + "dataset_name": "Code Migration - CLI", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cli", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (CLI).", + "metric_id": "vals_ai.code-migration.cli.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 13.44, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cli", + "task_name": "CLI", + "dataset_type": "private", + "industry": "coding", + "raw_score": "13.44", + "raw_stderr": "1.683", + "latency": "6689.822", + "cost_per_test": "5.885386", + "top_p": "0.95", + "max_output_tokens": "32768", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.683, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "code-migration:cobol:kimi/kimi-k2.7-code:score", + "evaluation_name": "vals_ai.code-migration.cobol", + "source_data": { + "dataset_name": "Code Migration - COBOL", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cobol", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (COBOL).", + "metric_id": "vals_ai.code-migration.cobol.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 61.25, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cobol", + "task_name": "COBOL", + "dataset_type": "private", + "industry": "coding", + "raw_score": "61.25", + "raw_stderr": "15.861", + "latency": "3713.208", + "cost_per_test": "2.507493", + "top_p": "0.95", + "max_output_tokens": "32768", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 15.861, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "code-migration:code_quality:kimi/kimi-k2.7-code:score", + "evaluation_name": "vals_ai.code-migration.code_quality", + "source_data": { + "dataset_name": "Code Migration - Code Quality", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "code_quality", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Code Quality).", + "metric_id": "vals_ai.code-migration.code_quality.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 3.547, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "code_quality", + "task_name": "Code Quality", + "dataset_type": "private", + "industry": "coding", + "raw_score": "3.547", + "raw_stderr": "0.065", + "latency": "5945.669", + "cost_per_test": "5.040913", + "top_p": "0.95", + "max_output_tokens": "32768", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.065, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "code-migration:overall:kimi/kimi-k2.7-code:score", + "evaluation_name": "vals_ai.code-migration.overall", + "source_data": { + "dataset_name": "Code Migration - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Overall).", + "metric_id": "vals_ai.code-migration.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 25.392, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "25.392", + "raw_stderr": "4.161", + "latency": "5945.669", + "cost_per_test": "5.040913", + "top_p": "0.95", + "max_output_tokens": "32768", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 4.161, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.7-code/5c558b2d-27f5-4879-b180-c9df8c23bbbe.json b/data/vals_ai/moonshotai/kimi-k2.7-code/5c558b2d-27f5-4879-b180-c9df8c23bbbe.json new file mode 100644 index 0000000000000000000000000000000000000000..5b7f1d912fde13c24f9f8282b9343a8d8c95dafe --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.7-code/5c558b2d-27f5-4879-b180-c9df8c23bbbe.json @@ -0,0 +1,260 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/moonshotai_kimi-k2.7-code/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.7-code", + "id": "moonshotai/kimi-k2.7-code", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.7-code", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:kimi/kimi-k2.7-code:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 91.667, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "91.667", + "raw_stderr": "8.333", + "latency": "17447.537", + "cost_per_test": "5.707198", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 8.333, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:kimi/kimi-k2.7-code:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 45.556, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "45.556", + "raw_stderr": "4.006", + "latency": "2326.338", + "cost_per_test": "0.76096", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 4.006, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:kimi/kimi-k2.7-code:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 76.97, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "76.97", + "raw_stderr": "3.207", + "latency": "1268.912", + "cost_per_test": "0.415069", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.207, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:kimi/kimi-k2.7-code:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 67.041, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "67.041", + "raw_stderr": "0.375", + "latency": "784.159", + "cost_per_test": "0.256503", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.375, + "method": "vals_reported" + } + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.7-code/5fd6a6ca-118e-4b7a-8aa8-fa9fb5ac291f.json b/data/vals_ai/moonshotai/kimi-k2.7-code/5fd6a6ca-118e-4b7a-8aa8-fa9fb5ac291f.json new file mode 100644 index 0000000000000000000000000000000000000000..0b649e6ec2c38150ffbddcf7d7b5f65f1b0a0ace --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.7-code/5fd6a6ca-118e-4b7a-8aa8-fa9fb5ac291f.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/moonshotai_kimi-k2.7-code/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.7-code", + "id": "moonshotai/kimi-k2.7-code", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.7-code", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:kimi/kimi-k2.7-code:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 47.211, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "47.211", + "raw_stderr": "5.191", + "latency": "10318.246", + "cost_per_test": "3.835806", + "top_p": "0.95", + "max_output_tokens": "32768", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 5.191, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.7-code/9da7f07d-4191-46e4-bd3f-098fae2a03ce.json b/data/vals_ai/moonshotai/kimi-k2.7-code/9da7f07d-4191-46e4-bd3f-098fae2a03ce.json new file mode 100644 index 0000000000000000000000000000000000000000..acf53c5234527e9e9f6edcbe8779deba51c74b60 --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.7-code/9da7f07d-4191-46e4-bd3f-098fae2a03ce.json @@ -0,0 +1,260 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/moonshotai_kimi-k2.7-code/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.7-code", + "id": "moonshotai/kimi-k2.7-code", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.7-code", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:kimi/kimi-k2.7-code:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 97.826, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "97.826", + "raw_stderr": "0.813", + "latency": "70.383", + "cost_per_test": "0.002724", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.813, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "lcb:hard:kimi/kimi-k2.7-code:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 60.857, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "60.857", + "raw_stderr": "2.609", + "latency": "388.241", + "cost_per_test": "0.060639", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.609, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "lcb:medium:kimi/kimi-k2.7-code:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 87.467, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "87.467", + "raw_stderr": "1.692", + "latency": "183.911", + "cost_per_test": "0.018683", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.692, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "lcb:overall:kimi/kimi-k2.7-code:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 82.05, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "82.05", + "raw_stderr": "1.071", + "latency": "214.178", + "cost_per_test": "0.027349", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.071, + "method": "vals_reported" + } + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.7-code/ced114ca-ce73-4410-8f5b-7fba16839192.json b/data/vals_ai/moonshotai/kimi-k2.7-code/ced114ca-ce73-4410-8f5b-7fba16839192.json new file mode 100644 index 0000000000000000000000000000000000000000..64c4f9054bc8c05f18dbc551be82722d520aeb6b --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.7-code/ced114ca-ce73-4410-8f5b-7fba16839192.json @@ -0,0 +1,316 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/moonshotai_kimi-k2.7-code/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.7-code", + "id": "moonshotai/kimi-k2.7-code", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.7-code", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:kimi/kimi-k2.7-code:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 50.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "50", + "raw_stderr": "7.809", + "latency": "1477.273", + "cost_per_test": "0.464016", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 7.809, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:kimi/kimi-k2.7-code:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 75.479, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "75.479", + "raw_stderr": "2.668", + "latency": "1041.18", + "cost_per_test": "0.312089", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.668, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:kimi/kimi-k2.7-code:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 88.66, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "88.66", + "raw_stderr": "2.282", + "latency": "561.729", + "cost_per_test": "0.154689", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.282, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:kimi/kimi-k2.7-code:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "1862.15", + "cost_per_test": "0.561509", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "swebench:overall:kimi/kimi-k2.7-code:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 78.2, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "78.2", + "raw_stderr": "1.848", + "latency": "896.71", + "cost_per_test": "0.265276", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.848, + "method": "vals_reported" + } + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/moonshotai/kimi-k2.7-code/deab0949-a6b5-4272-abae-4d0364bb47ee.json b/data/vals_ai/moonshotai/kimi-k2.7-code/deab0949-a6b5-4272-abae-4d0364bb47ee.json new file mode 100644 index 0000000000000000000000000000000000000000..2605e058e8cf4443b4463f6da2e34993f39c58f5 --- /dev/null +++ b/data/vals_ai/moonshotai/kimi-k2.7-code/deab0949-a6b5-4272-abae-4d0364bb47ee.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/programbench/moonshotai_kimi-k2.7-code/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProgramBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "kimi-k2.7-code", + "id": "moonshotai/kimi-k2.7-code", + "developer": "moonshotai", + "additional_details": { + "vals_model_id": "kimi/kimi-k2.7-code", + "vals_provider": "Moonshot AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "programbench:almost:kimi/kimi-k2.7-code:score", + "evaluation_name": "vals_ai.programbench.almost", + "source_data": { + "dataset_name": "ProgramBench - Almost Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "almost", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Almost Resolved).", + "metric_id": "vals_ai.programbench.almost.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.5, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "almost", + "task_name": "Almost Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0.5", + "raw_stderr": "0.5", + "latency": "3860.332", + "cost_per_test": "3.185154", + "top_p": "0.95", + "max_output_tokens": "32768", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.5, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "programbench:overall:kimi/kimi-k2.7-code:score", + "evaluation_name": "vals_ai.programbench.overall", + "source_data": { + "dataset_name": "ProgramBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Overall).", + "metric_id": "vals_ai.programbench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "3860.332", + "cost_per_test": "3.185154", + "top_p": "0.95", + "max_output_tokens": "32768", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "programbench:partial:kimi/kimi-k2.7-code:score", + "evaluation_name": "vals_ai.programbench.partial", + "source_data": { + "dataset_name": "ProgramBench - Raw Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "partial", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Raw Pass Rate).", + "metric_id": "vals_ai.programbench.partial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 49.435, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "partial", + "task_name": "Raw Pass Rate", + "dataset_type": "public", + "industry": "coding", + "raw_score": "49.435", + "raw_stderr": "1.799", + "latency": "3860.332", + "cost_per_test": "3.185154", + "top_p": "0.95", + "max_output_tokens": "32768", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.799, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "programbench:strict:kimi/kimi-k2.7-code:score", + "evaluation_name": "vals_ai.programbench.strict", + "source_data": { + "dataset_name": "ProgramBench - Fully Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "strict", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Fully Resolved).", + "metric_id": "vals_ai.programbench.strict.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "strict", + "task_name": "Fully Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "3860.332", + "cost_per_test": "3.185154", + "top_p": "0.95", + "max_output_tokens": "32768", + "provider": "Moonshot AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/0e6f80a1-fd0c-4e2f-968f-c930db786679.json b/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/0e6f80a1-fd0c-4e2f-968f-c930db786679.json new file mode 100644 index 0000000000000000000000000000000000000000..521a7cef0d16f58e29cb3505684a09223fa28c17 --- /dev/null +++ b/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/0e6f80a1-fd0c-4e2f-968f-c930db786679.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/nvidia_nemotron-3-ultra-550b-a55b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "nemotron-3-ultra-550b-a55b", + "id": "nvidia/nemotron-3-ultra-550b-a55b", + "developer": "nvidia", + "additional_details": { + "vals_model_id": "nvidia/nemotron-3-ultra-550b-a55b", + "vals_provider": "Nvidia" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 86.364, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.364", + "raw_stderr": "2.439", + "latency": "142.062", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 2.439, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 85.859, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.859", + "raw_stderr": "2.476", + "latency": "174.88", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 2.476, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 86.112, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.112", + "raw_stderr": "1.738", + "latency": "158.471", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 1.738, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/1722c51e-604f-40cf-83bc-52134f20ee9e.json b/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/1722c51e-604f-40cf-83bc-52134f20ee9e.json new file mode 100644 index 0000000000000000000000000000000000000000..9520fa9d2879700883fd4f6246c444ba06142a94 --- /dev/null +++ b/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/1722c51e-604f-40cf-83bc-52134f20ee9e.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/nvidia_nemotron-3-ultra-550b-a55b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "nemotron-3-ultra-550b-a55b", + "id": "nvidia/nemotron-3-ultra-550b-a55b", + "developer": "nvidia", + "additional_details": { + "vals_model_id": "nvidia/nemotron-3-ultra-550b-a55b", + "vals_provider": "Nvidia" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 91.667, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "91.667", + "raw_stderr": "8.333", + "latency": "12996.574", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 8.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 24.444, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "24.444", + "raw_stderr": "4.006", + "latency": "1732.877", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 4.006, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 62.424, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "62.424", + "raw_stderr": "2.424", + "latency": "945.205", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 2.424, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 50.936, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "50.936", + "raw_stderr": "2.925", + "latency": "584.116", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 2.925, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/3441c4db-ad00-418a-9e54-966618f8bd4c.json b/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/3441c4db-ad00-418a-9e54-966618f8bd4c.json new file mode 100644 index 0000000000000000000000000000000000000000..5dbede202abd1b8ecfb68d0ee8f9b0169de569bd --- /dev/null +++ b/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/3441c4db-ad00-418a-9e54-966618f8bd4c.json @@ -0,0 +1,97 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/nvidia_nemotron-3-ultra-550b-a55b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "nemotron-3-ultra-550b-a55b", + "id": "nvidia/nemotron-3-ultra-550b-a55b", + "developer": "nvidia", + "additional_details": { + "vals_model_id": "nvidia/nemotron-3-ultra-550b-a55b", + "vals_provider": "Nvidia" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 2.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "2", + "raw_stderr": "1.407", + "latency": "2552.288", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 1.407, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/3b5f2772-8951-4902-93d9-b66fc657938c.json b/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/3b5f2772-8951-4902-93d9-b66fc657938c.json new file mode 100644 index 0000000000000000000000000000000000000000..35f681cc4fd8f51b6a47639804a728d014467569 --- /dev/null +++ b/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/3b5f2772-8951-4902-93d9-b66fc657938c.json @@ -0,0 +1,97 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/nvidia_nemotron-3-ultra-550b-a55b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "nemotron-3-ultra-550b-a55b", + "id": "nvidia/nemotron-3-ultra-550b-a55b", + "developer": "nvidia", + "additional_details": { + "vals_model_id": "nvidia/nemotron-3-ultra-550b-a55b", + "vals_provider": "Nvidia" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 7.636, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "7.636", + "raw_stderr": "2.783", + "latency": "2217.462", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 2.783, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/6650b968-33d5-482c-bad1-cf6f0b8d7818.json b/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/6650b968-33d5-482c-bad1-cf6f0b8d7818.json new file mode 100644 index 0000000000000000000000000000000000000000..3242a9dd405dc66481e0ff099938efa322f4c3bd --- /dev/null +++ b/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/6650b968-33d5-482c-bad1-cf6f0b8d7818.json @@ -0,0 +1,392 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/nvidia_nemotron-3-ultra-550b-a55b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "nemotron-3-ultra-550b-a55b", + "id": "nvidia/nemotron-3-ultra-550b-a55b", + "developer": "nvidia", + "additional_details": { + "vals_model_id": "nvidia/nemotron-3-ultra-550b-a55b", + "vals_provider": "Nvidia" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 64.918, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "64.918", + "raw_stderr": "1.629", + "latency": "15.701", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 1.629, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 38.16, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "38.16", + "raw_stderr": "2.661", + "latency": "239.629", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 2.661, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:overall:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 43.987, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "43.987", + "raw_stderr": "1.263", + "latency": "638.492", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 1.263, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 67.647, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "67.647", + "raw_stderr": "4.632", + "latency": "328.551" + }, + "uncertainty": { + "standard_error": { + "value": 4.632, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 50.936, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "50.936", + "raw_stderr": "2.925", + "latency": "584.116", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 2.925, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 7.104, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "7.104", + "raw_stderr": "3.207", + "latency": "2024.461", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 3.207, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/66617183-2181-47a2-ad43-4a2e071813f5.json b/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/66617183-2181-47a2-ad43-4a2e071813f5.json new file mode 100644 index 0000000000000000000000000000000000000000..92ab815ffce9a70e798d5da640760060909f7793 --- /dev/null +++ b/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/66617183-2181-47a2-ad43-4a2e071813f5.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/programbench/nvidia_nemotron-3-ultra-550b-a55b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProgramBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "nemotron-3-ultra-550b-a55b", + "id": "nvidia/nemotron-3-ultra-550b-a55b", + "developer": "nvidia", + "additional_details": { + "vals_model_id": "nvidia/nemotron-3-ultra-550b-a55b", + "vals_provider": "Nvidia" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "programbench:almost:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.programbench.almost", + "source_data": { + "dataset_name": "ProgramBench - Almost Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "almost", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Almost Resolved).", + "metric_id": "vals_ai.programbench.almost.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.5, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "almost", + "task_name": "Almost Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0.5", + "raw_stderr": "0.5", + "latency": "3718.555", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 0.5, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "programbench:overall:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.programbench.overall", + "source_data": { + "dataset_name": "ProgramBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Overall).", + "metric_id": "vals_ai.programbench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "3718.555", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "programbench:partial:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.programbench.partial", + "source_data": { + "dataset_name": "ProgramBench - Raw Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "partial", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Raw Pass Rate).", + "metric_id": "vals_ai.programbench.partial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 9.967, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "partial", + "task_name": "Raw Pass Rate", + "dataset_type": "public", + "industry": "coding", + "raw_score": "9.967", + "raw_stderr": "1.376", + "latency": "3718.555", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 1.376, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "programbench:strict:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.programbench.strict", + "source_data": { + "dataset_name": "ProgramBench - Fully Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "strict", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Fully Resolved).", + "metric_id": "vals_ai.programbench.strict.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "strict", + "task_name": "Fully Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "3718.555", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/69b31a1c-b677-4588-a619-8abc30e0ce07.json b/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/69b31a1c-b677-4588-a619-8abc30e0ce07.json new file mode 100644 index 0000000000000000000000000000000000000000..d50b465209f9759d0007c32b3266106d7028a681 --- /dev/null +++ b/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/69b31a1c-b677-4588-a619-8abc30e0ce07.json @@ -0,0 +1,356 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/nvidia_nemotron-3-ultra-550b-a55b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "nemotron-3-ultra-550b-a55b", + "id": "nvidia/nemotron-3-ultra-550b-a55b", + "developer": "nvidia", + "additional_details": { + "vals_model_id": "nvidia/nemotron-3-ultra-550b-a55b", + "vals_provider": "Nvidia" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 38.095, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "38.095", + "raw_stderr": "7.584", + "latency": "400.802", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 7.584, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 65.517, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "65.517", + "raw_stderr": "2.948", + "latency": "313.181", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 2.948, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 80.928, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "80.928", + "raw_stderr": "2.828", + "latency": "214.147", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 2.828, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "475.334", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 69.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "69", + "raw_stderr": "2.07", + "latency": "283.089", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 2.07, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/776ca076-866d-45f4-81aa-e287b67da429.json b/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/776ca076-866d-45f4-81aa-e287b67da429.json new file mode 100644 index 0000000000000000000000000000000000000000..84bd0330c1542bbe06b15b5f40cdd9e960ebc6f9 --- /dev/null +++ b/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/776ca076-866d-45f4-81aa-e287b67da429.json @@ -0,0 +1,996 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/nvidia_nemotron-3-ultra-550b-a55b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "nemotron-3-ultra-550b-a55b", + "id": "nvidia/nemotron-3-ultra-550b-a55b", + "developer": "nvidia", + "additional_details": { + "vals_model_id": "nvidia/nemotron-3-ultra-550b-a55b", + "vals_provider": "Nvidia" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.19, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.19", + "raw_stderr": "1.002", + "latency": "14.605", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 1.002, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.001, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.001", + "raw_stderr": "1.019", + "latency": "29.998", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 1.019, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.696, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.696", + "raw_stderr": "0.82", + "latency": "48.434", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 0.82, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.756, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.756", + "raw_stderr": "1.498", + "latency": "31.773", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 1.498, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.507, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.507", + "raw_stderr": "1.098", + "latency": "16.623", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 1.098, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.766, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.766", + "raw_stderr": "1.213", + "latency": "132.631", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 1.213, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.462, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.462", + "raw_stderr": "1.412", + "latency": "17.419", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 1.412, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.115, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.115", + "raw_stderr": "2.184", + "latency": "13.577", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 2.184, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 72.934, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.934", + "raw_stderr": "1.339", + "latency": "33.366", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 1.339, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 95.559, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "95.559", + "raw_stderr": "0.56", + "latency": "30.914", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 0.56, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.9, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.9", + "raw_stderr": "1.239", + "latency": "15.648", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 1.239, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.761, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.761", + "raw_stderr": "0.345", + "latency": "32.92", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 0.345, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.962, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.962", + "raw_stderr": "1.758", + "latency": "16.114", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 1.758, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.225, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.225", + "raw_stderr": "0.743", + "latency": "46.02", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 0.743, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.586, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.586", + "raw_stderr": "1.278", + "latency": "13.753", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 1.278, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/7a9a5e8c-287c-4c1f-b394-c3289e3bf26f.json b/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/7a9a5e8c-287c-4c1f-b394-c3289e3bf26f.json new file mode 100644 index 0000000000000000000000000000000000000000..40138620c37caa6c7bd820a1bab926d6e1d3f119 --- /dev/null +++ b/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/7a9a5e8c-287c-4c1f-b394-c3289e3bf26f.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/nvidia_nemotron-3-ultra-550b-a55b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "nemotron-3-ultra-550b-a55b", + "id": "nvidia/nemotron-3-ultra-550b-a55b", + "developer": "nvidia", + "additional_details": { + "vals_model_id": "nvidia/nemotron-3-ultra-550b-a55b", + "vals_provider": "Nvidia" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.47, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.47", + "raw_stderr": "0.772", + "latency": "6.253", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 0.772, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.266, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.266", + "raw_stderr": "0.209", + "latency": "6.225", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 0.209, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.318, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.318", + "raw_stderr": "0.59", + "latency": "2.071", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 0.59, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.066, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.066", + "raw_stderr": "0.46", + "latency": "5.179", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 0.46, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.885, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.885", + "raw_stderr": "1.088", + "latency": "3.366", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 1.088, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.393, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.393", + "raw_stderr": "1.763", + "latency": "7.98", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 1.763, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/9287d013-fcae-4408-b2ef-2c177f85e1dc.json b/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/9287d013-fcae-4408-b2ef-2c177f85e1dc.json new file mode 100644 index 0000000000000000000000000000000000000000..74ae257b85f4f978ff95df410007b22f7809f414 --- /dev/null +++ b/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/9287d013-fcae-4408-b2ef-2c177f85e1dc.json @@ -0,0 +1,219 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/nvidia_nemotron-3-ultra-550b-a55b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "nemotron-3-ultra-550b-a55b", + "id": "nvidia/nemotron-3-ultra-550b-a55b", + "developer": "nvidia", + "additional_details": { + "vals_model_id": "nvidia/nemotron-3-ultra-550b-a55b", + "vals_provider": "Nvidia" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 61.652, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.652", + "raw_stderr": "1.39", + "latency": "36.485", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 1.39, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 73.099, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "73.099", + "raw_stderr": "0.866", + "latency": "36.485", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 0.866, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 84.546, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "84.546", + "raw_stderr": "1.034", + "latency": "36.485", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 1.034, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/b2b69d45-4941-48db-8e01-2cfe135e9900.json b/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/b2b69d45-4941-48db-8e01-2cfe135e9900.json new file mode 100644 index 0000000000000000000000000000000000000000..3db23cedf6aa70ff6b353a3cd27aaeaea9f36289 --- /dev/null +++ b/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/b2b69d45-4941-48db-8e01-2cfe135e9900.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/nvidia_nemotron-3-ultra-550b-a55b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "nemotron-3-ultra-550b-a55b", + "id": "nvidia/nemotron-3-ultra-550b-a55b", + "developer": "nvidia", + "additional_details": { + "vals_model_id": "nvidia/nemotron-3-ultra-550b-a55b", + "vals_provider": "Nvidia" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.447, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.447", + "raw_stderr": "0.689", + "latency": "16.96", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 0.689, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 69.143, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "69.143", + "raw_stderr": "2.469", + "latency": "366.032", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 2.469, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 90.339, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "90.339", + "raw_stderr": "1.51", + "latency": "100.608", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 1.51, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 85.976, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "85.976", + "raw_stderr": "0.992", + "latency": "161.2", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 0.992, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/c098504e-f1e3-4ede-b4b2-3943a7f8c9c8.json b/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/c098504e-f1e3-4ede-b4b2-3943a7f8c9c8.json new file mode 100644 index 0000000000000000000000000000000000000000..08017a735901ad8e2573314b1644a4bbddfdd36f --- /dev/null +++ b/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/c098504e-f1e3-4ede-b4b2-3943a7f8c9c8.json @@ -0,0 +1,280 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/nvidia_nemotron-3-ultra-550b-a55b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "nemotron-3-ultra-550b-a55b", + "id": "nvidia/nemotron-3-ultra-550b-a55b", + "developer": "nvidia", + "additional_details": { + "vals_model_id": "nvidia/nemotron-3-ultra-550b-a55b", + "vals_provider": "Nvidia" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 64.336, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.336", + "raw_stderr": "1.635", + "latency": "6.738", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 1.635, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 67.133, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.133", + "raw_stderr": "1.604", + "latency": "22.607", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 1.604, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 65.462, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.462", + "raw_stderr": "0.937", + "latency": "15.015", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 0.937, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 64.918, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.918", + "raw_stderr": "1.629", + "latency": "15.701", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 1.629, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/d8104b4b-30de-4072-a6e8-199c7155155e.json b/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/d8104b4b-30de-4072-a6e8-199c7155155e.json new file mode 100644 index 0000000000000000000000000000000000000000..5dd6de4b5ed87eef8641bd02ededaa027749e1f2 --- /dev/null +++ b/data/vals_ai/nvidia/nemotron-3-ultra-550b-a55b/d8104b4b-30de-4072-a6e8-199c7155155e.json @@ -0,0 +1,707 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/nvidia_nemotron-3-ultra-550b-a55b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "nemotron-3-ultra-550b-a55b", + "id": "nvidia/nemotron-3-ultra-550b-a55b", + "developer": "nvidia", + "additional_details": { + "vals_model_id": "nvidia/nemotron-3-ultra-550b-a55b", + "vals_provider": "Nvidia" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 34.449, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "34.449", + "raw_stderr": "3.045", + "latency": "255.224", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 3.045, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 24.802, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "24.802", + "raw_stderr": "0.728", + "latency": "253.882", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 0.728, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 37.181, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "37.181", + "raw_stderr": "1.6", + "latency": "288.432", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 1.6, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 51.307, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "51.307", + "raw_stderr": "0.192", + "latency": "250.558", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 0.192, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 40.623, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "40.623", + "raw_stderr": "1.887", + "latency": "203.776", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 1.887, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 7.071, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "7.071", + "raw_stderr": "1.766", + "latency": "312.219", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 1.766, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 58.487, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.487", + "raw_stderr": "3.971", + "latency": "231.138", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 3.971, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 55.802, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "55.802", + "raw_stderr": "3.002", + "latency": "213.531", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 3.002, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 41.27, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "41.27", + "raw_stderr": "1.624", + "latency": "208.881", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 1.624, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:overall:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 37.667, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "37.667", + "raw_stderr": "0.149", + "latency": "253.882", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 0.149, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:nvidia/nemotron-3-ultra-550b-a55b:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 12.816, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "12.816", + "raw_stderr": "3.421", + "latency": "333.089", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Nvidia" + }, + "uncertainty": { + "standard_error": { + "value": 3.421, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-3.5-turbo/0171ecaa-781c-4aae-929a-03e2dc94620e.json b/data/vals_ai/openai/gpt-3.5-turbo/0171ecaa-781c-4aae-929a-03e2dc94620e.json new file mode 100644 index 0000000000000000000000000000000000000000..aa2489da3f36c2310b52aeeea82affbf944ec16b --- /dev/null +++ b/data/vals_ai/openai/gpt-3.5-turbo/0171ecaa-781c-4aae-929a-03e2dc94620e.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/openai_gpt-3.5-turbo/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-3.5-turbo", + "id": "openai/gpt-3.5-turbo", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-3.5-turbo", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:openai/gpt-3.5-turbo:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 28.788, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "28.788", + "raw_stderr": "3.201", + "latency": "2.834", + "cost_per_test": "0.001296", + "temperature": "1", + "max_output_tokens": "4096", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.201, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 4096, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:openai/gpt-3.5-turbo:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 32.323, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "32.323", + "raw_stderr": "3.266", + "latency": "2.423", + "cost_per_test": "0.000427", + "temperature": "1", + "max_output_tokens": "4096", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.266, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 4096, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:openai/gpt-3.5-turbo:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 30.556, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "30.556", + "raw_stderr": "2.287", + "latency": "2.628", + "cost_per_test": "0.000862", + "temperature": "1", + "max_output_tokens": "4096", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.287, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 4096, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-3.5-turbo/2759d766-0160-4c1b-a48a-3585695135f8.json b/data/vals_ai/openai/gpt-3.5-turbo/2759d766-0160-4c1b-a48a-3585695135f8.json new file mode 100644 index 0000000000000000000000000000000000000000..ecccfc9c21d19d711782ab5754af7ce136b39644 --- /dev/null +++ b/data/vals_ai/openai/gpt-3.5-turbo/2759d766-0160-4c1b-a48a-3585695135f8.json @@ -0,0 +1,444 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/openai_gpt-3.5-turbo/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-3.5-turbo", + "id": "openai/gpt-3.5-turbo", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-3.5-turbo", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:openai/gpt-3.5-turbo:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 67.401, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "67.401", + "raw_stderr": "1.13", + "latency": "0.345", + "temperature": "1", + "max_output_tokens": "4096", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.13, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 4096, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:openai/gpt-3.5-turbo:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 65.515, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "65.515", + "raw_stderr": "0.21", + "latency": "0.46", + "temperature": "1", + "max_output_tokens": "4096", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.21, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 4096, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:openai/gpt-3.5-turbo:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 57.054, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "57.054", + "raw_stderr": "0.515", + "latency": "0.665", + "temperature": "1", + "max_output_tokens": "4096", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.515, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 4096, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:openai/gpt-3.5-turbo:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 64.372, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "64.372", + "raw_stderr": "0.628", + "latency": "0.542", + "temperature": "1", + "max_output_tokens": "4096", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.628, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 4096, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:openai/gpt-3.5-turbo:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 65.714, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "65.714", + "raw_stderr": "1.095", + "latency": "0.465", + "temperature": "1", + "max_output_tokens": "4096", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.095, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 4096, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:openai/gpt-3.5-turbo:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 66.174, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "66.174", + "raw_stderr": "2.659", + "latency": "0.777", + "temperature": "1", + "max_output_tokens": "4096", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.659, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 4096, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4-turbo/82eb40cf-9008-4cc8-8605-4e0d3ca67757.json b/data/vals_ai/openai/gpt-4-turbo/82eb40cf-9008-4cc8-8605-4e0d3ca67757.json new file mode 100644 index 0000000000000000000000000000000000000000..060ea047ee7e9edd2f83d0cb81bc27dc44860565 --- /dev/null +++ b/data/vals_ai/openai/gpt-4-turbo/82eb40cf-9008-4cc8-8605-4e0d3ca67757.json @@ -0,0 +1,444 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/openai_gpt-4-turbo/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4-turbo", + "id": "openai/gpt-4-turbo", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4-turbo", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:openai/gpt-4-turbo:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.547, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.547", + "raw_stderr": "0.954", + "latency": "0.69", + "temperature": "1", + "max_output_tokens": "4096", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.954, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 4096, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:openai/gpt-4-turbo:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 76.501, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "76.501", + "raw_stderr": "0.233", + "latency": "0.939", + "temperature": "1", + "max_output_tokens": "4096", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.233, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 4096, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:openai/gpt-4-turbo:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.88, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.88", + "raw_stderr": "0.499", + "latency": "1.166", + "temperature": "1", + "max_output_tokens": "4096", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.499, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 4096, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:openai/gpt-4-turbo:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.462, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.462", + "raw_stderr": "0.482", + "latency": "1.405", + "temperature": "1", + "max_output_tokens": "4096", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.482, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 4096, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:openai/gpt-4-turbo:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.12, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.12", + "raw_stderr": "0.997", + "latency": "0.95", + "temperature": "1", + "max_output_tokens": "4096", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.997, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 4096, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:openai/gpt-4-turbo:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.26, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.26", + "raw_stderr": "1.897", + "latency": "3.282", + "temperature": "1", + "max_output_tokens": "4096", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.897, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 4096, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4.1-2025-04-14/18847d55-0ec4-4bc6-b4de-4715b1365c49.json b/data/vals_ai/openai/gpt-4.1-2025-04-14/18847d55-0ec4-4bc6-b4de-4715b1365c49.json new file mode 100644 index 0000000000000000000000000000000000000000..3aac634a9b3b7c9dd63e4c5e32f813318f881dc9 --- /dev/null +++ b/data/vals_ai/openai/gpt-4.1-2025-04-14/18847d55-0ec4-4bc6-b4de-4715b1365c49.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/openai_gpt-4.1-2025-04-14/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4.1-2025-04-14", + "id": "openai/gpt-4.1-2025-04-14", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4.1-2025-04-14", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 63.859, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.859", + "raw_stderr": "1.374", + "latency": "7.89", + "cost_per_test": "0.006214", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.374, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 75.061, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "75.061", + "raw_stderr": "0.845", + "latency": "7.89", + "cost_per_test": "0.006214", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.845, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 86.263, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "86.263", + "raw_stderr": "0.984", + "latency": "7.89", + "cost_per_test": "0.006214", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.984, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4.1-2025-04-14/1dd1c65d-b697-4b13-a9af-495d620cf7cc.json b/data/vals_ai/openai/gpt-4.1-2025-04-14/1dd1c65d-b697-4b13-a9af-495d620cf7cc.json new file mode 100644 index 0000000000000000000000000000000000000000..8d2cd673934edba4cfe49efacd453e6c20d3cfe5 --- /dev/null +++ b/data/vals_ai/openai/gpt-4.1-2025-04-14/1dd1c65d-b697-4b13-a9af-495d620cf7cc.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/openai_gpt-4.1-2025-04-14/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4.1-2025-04-14", + "id": "openai/gpt-4.1-2025-04-14", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4.1-2025-04-14", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 64.141, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "64.141", + "raw_stderr": "3.419", + "latency": "20.444", + "cost_per_test": "0.017619", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.419, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.667", + "raw_stderr": "3.375", + "latency": "25.835", + "cost_per_test": "0.010086", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.375, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 65.404, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "65.404", + "raw_stderr": "2.402", + "latency": "23.14", + "cost_per_test": "0.013852", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.402, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4.1-2025-04-14/30b22f44-e127-49da-93da-edcfbffb361f.json b/data/vals_ai/openai/gpt-4.1-2025-04-14/30b22f44-e127-49da-93da-edcfbffb361f.json new file mode 100644 index 0000000000000000000000000000000000000000..1b6e75f5987e9f145537aadbe534073e043481da --- /dev/null +++ b/data/vals_ai/openai/gpt-4.1-2025-04-14/30b22f44-e127-49da-93da-edcfbffb361f.json @@ -0,0 +1,444 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/openai_gpt-4.1-2025-04-14/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4.1-2025-04-14", + "id": "openai/gpt-4.1-2025-04-14", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4.1-2025-04-14", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.139, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.139", + "raw_stderr": "0.875", + "latency": "0.386", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.875, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.768, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.768", + "raw_stderr": "0.192", + "latency": "0.411", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.192, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.158, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.158", + "raw_stderr": "0.592", + "latency": "0.418", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.592, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.1, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.1", + "raw_stderr": "0.427", + "latency": "0.509", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.427, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.218, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.218", + "raw_stderr": "0.834", + "latency": "0.4", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.834, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.218, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.218", + "raw_stderr": "1.645", + "latency": "0.931", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.645, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4.1-2025-04-14/3108ba96-a13f-4dd1-b4a5-5c118c256e7f.json b/data/vals_ai/openai/gpt-4.1-2025-04-14/3108ba96-a13f-4dd1-b4a5-5c118c256e7f.json new file mode 100644 index 0000000000000000000000000000000000000000..470c0511472d9af6ec086c04396762aec6ed56f3 --- /dev/null +++ b/data/vals_ai/openai/gpt-4.1-2025-04-14/3108ba96-a13f-4dd1-b4a5-5c118c256e7f.json @@ -0,0 +1,105 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/openai_gpt-4.1-2025-04-14/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4.1-2025-04-14", + "id": "openai/gpt-4.1-2025-04-14", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4.1-2025-04-14", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 72.386, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.386", + "raw_stderr": "1.075", + "latency": "12", + "cost_per_test": "0.007959", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.075, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4.1-2025-04-14/533d3641-2c8a-4c49-9eee-da14927bf2db.json b/data/vals_ai/openai/gpt-4.1-2025-04-14/533d3641-2c8a-4c49-9eee-da14927bf2db.json new file mode 100644 index 0000000000000000000000000000000000000000..a713024b7b17f951b00763d3f00002852103ee40 --- /dev/null +++ b/data/vals_ai/openai/gpt-4.1-2025-04-14/533d3641-2c8a-4c49-9eee-da14927bf2db.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/openai_gpt-4.1-2025-04-14/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4.1-2025-04-14", + "id": "openai/gpt-4.1-2025-04-14", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4.1-2025-04-14", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 94.72, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "94.72", + "raw_stderr": "1.246", + "latency": "2.191", + "cost_per_test": "0.001899", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.246, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:hard:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 13.143, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "13.143", + "raw_stderr": "1.806", + "latency": "79.95", + "cost_per_test": "0.018997", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.806, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:medium:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 56.136, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "56.136", + "raw_stderr": "2.536", + "latency": "9.553", + "cost_per_test": "0.007034", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.536, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:overall:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 54.666, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "54.666", + "raw_stderr": "1.118", + "latency": "30.565", + "cost_per_test": "0.00931", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.118, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4.1-2025-04-14/5d559bae-b900-4912-af01-7b8bebd2ae68.json b/data/vals_ai/openai/gpt-4.1-2025-04-14/5d559bae-b900-4912-af01-7b8bebd2ae68.json new file mode 100644 index 0000000000000000000000000000000000000000..aa87fee68346f2ee60489ba16ec2bf0b7564b7b8 --- /dev/null +++ b/data/vals_ai/openai/gpt-4.1-2025-04-14/5d559bae-b900-4912-af01-7b8bebd2ae68.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/openai_gpt-4.1-2025-04-14/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4.1-2025-04-14", + "id": "openai/gpt-4.1-2025-04-14", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4.1-2025-04-14", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 62.821, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.821", + "raw_stderr": "1.65", + "latency": "3.975", + "cost_per_test": "0.005077", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.65, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.839, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.839", + "raw_stderr": "1.666", + "latency": "15.579", + "cost_per_test": "0.261029", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.666, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 63.054, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.054", + "raw_stderr": "0.951", + "latency": "10.14", + "cost_per_test": "0.13202", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.951, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 65.501, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.501", + "raw_stderr": "1.623", + "latency": "10.865", + "cost_per_test": "0.129954", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.623, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4.1-2025-04-14/a40820c5-06b0-4afd-a21e-833556151548.json b/data/vals_ai/openai/gpt-4.1-2025-04-14/a40820c5-06b0-4afd-a21e-833556151548.json new file mode 100644 index 0000000000000000000000000000000000000000..e5748686fbc1a731b4552d1a8574f5e4af64a8a1 --- /dev/null +++ b/data/vals_ai/openai/gpt-4.1-2025-04-14/a40820c5-06b0-4afd-a21e-833556151548.json @@ -0,0 +1,1071 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/openai_gpt-4.1-2025-04-14/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4.1-2025-04-14", + "id": "openai/gpt-4.1-2025-04-14", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4.1-2025-04-14", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.958, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.958", + "raw_stderr": "1.122", + "latency": "9.333", + "cost_per_test": "0.005755", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.122, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.298, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.298", + "raw_stderr": "1.261", + "latency": "5.622", + "cost_per_test": "0.00618", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.261, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.951, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.951", + "raw_stderr": "1.118", + "latency": "20.885", + "cost_per_test": "0.010238", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.118, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.122, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.122", + "raw_stderr": "1.758", + "latency": "5.688", + "cost_per_test": "0.00729", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.758, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.019, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.019", + "raw_stderr": "1.194", + "latency": "5.216", + "cost_per_test": "0.004358", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.194, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 67.802, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "67.802", + "raw_stderr": "1.501", + "latency": "16.506", + "cost_per_test": "0.011514", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.501, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.995, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.995", + "raw_stderr": "1.448", + "latency": "2.974", + "cost_per_test": "0.004239", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.448, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 72.178, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.178", + "raw_stderr": "2.296", + "latency": "3.257", + "cost_per_test": "0.006572", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.296, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 65.577, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "65.577", + "raw_stderr": "1.432", + "latency": "4.849", + "cost_per_test": "0.006256", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.432, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.637, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.637", + "raw_stderr": "0.829", + "latency": "8.747", + "cost_per_test": "0.010057", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.829, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.87, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.87", + "raw_stderr": "1.319", + "latency": "2.19", + "cost_per_test": "0.00323", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.319, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.495, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.495", + "raw_stderr": "0.385", + "latency": "8.177", + "cost_per_test": "0.007141", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.385, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.353, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.353", + "raw_stderr": "1.902", + "latency": "18.234", + "cost_per_test": "0.013185", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.902, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.834, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.834", + "raw_stderr": "0.995", + "latency": "8.881", + "cost_per_test": "0.007688", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.995, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.333, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.333", + "raw_stderr": "1.319", + "latency": "2.093", + "cost_per_test": "0.003408", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.319, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4.1-2025-04-14/d051d852-bbc0-4b51-939c-84a6a05de70d.json b/data/vals_ai/openai/gpt-4.1-2025-04-14/d051d852-bbc0-4b51-939c-84a6a05de70d.json new file mode 100644 index 0000000000000000000000000000000000000000..64fb4a738d71a49dbd5f933d1b53c9831b4e5065 --- /dev/null +++ b/data/vals_ai/openai/gpt-4.1-2025-04-14/d051d852-bbc0-4b51-939c-84a6a05de70d.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/openai_gpt-4.1-2025-04-14/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4.1-2025-04-14", + "id": "openai/gpt-4.1-2025-04-14", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4.1-2025-04-14", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 76.232, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "76.232", + "raw_stderr": "1.2", + "latency": "4.837", + "cost_per_test": "0.006059", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.2, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 65.938, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.938", + "raw_stderr": "0.922", + "latency": "5.034", + "cost_per_test": "0.006212", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.922, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:openai/gpt-4.1-2025-04-14:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 55.644, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "55.644", + "raw_stderr": "1.401", + "latency": "5.23", + "cost_per_test": "0.006366", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.401, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4.1-mini-2025-04-14/01c8029b-13b2-410b-b58d-f3757e0843f3.json b/data/vals_ai/openai/gpt-4.1-mini-2025-04-14/01c8029b-13b2-410b-b58d-f3757e0843f3.json new file mode 100644 index 0000000000000000000000000000000000000000..65b6404ef6f8f3b1964d83553d515c4b01159b2f --- /dev/null +++ b/data/vals_ai/openai/gpt-4.1-mini-2025-04-14/01c8029b-13b2-410b-b58d-f3757e0843f3.json @@ -0,0 +1,105 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/openai_gpt-4.1-mini-2025-04-14/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4.1-mini-2025-04-14", + "id": "openai/gpt-4.1-mini-2025-04-14", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4.1-mini-2025-04-14", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 70.537, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.537", + "raw_stderr": "1.096", + "latency": "8.225", + "cost_per_test": "0.001788", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.096, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4.1-mini-2025-04-14/194916f0-d46a-4aec-b83d-aa8d5acfdfcb.json b/data/vals_ai/openai/gpt-4.1-mini-2025-04-14/194916f0-d46a-4aec-b83d-aa8d5acfdfcb.json new file mode 100644 index 0000000000000000000000000000000000000000..17ea2e84a293bafd8c422a4ac13d4b32f6655f65 --- /dev/null +++ b/data/vals_ai/openai/gpt-4.1-mini-2025-04-14/194916f0-d46a-4aec-b83d-aa8d5acfdfcb.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/openai_gpt-4.1-mini-2025-04-14/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4.1-mini-2025-04-14", + "id": "openai/gpt-4.1-mini-2025-04-14", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4.1-mini-2025-04-14", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 72.337, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "72.337", + "raw_stderr": "1.261", + "latency": "4.59", + "cost_per_test": "0.002781", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.261, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 65.501, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.501", + "raw_stderr": "0.938", + "latency": "4.598", + "cost_per_test": "0.002797", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.938, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 58.665, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.665", + "raw_stderr": "1.388", + "latency": "4.606", + "cost_per_test": "0.002813", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.388, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4.1-mini-2025-04-14/665b383d-3fd6-4cc9-98eb-ce24b885d668.json b/data/vals_ai/openai/gpt-4.1-mini-2025-04-14/665b383d-3fd6-4cc9-98eb-ce24b885d668.json new file mode 100644 index 0000000000000000000000000000000000000000..e94e5123e0acf357a3ac9e18b575405863cce001 --- /dev/null +++ b/data/vals_ai/openai/gpt-4.1-mini-2025-04-14/665b383d-3fd6-4cc9-98eb-ce24b885d668.json @@ -0,0 +1,1071 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/openai_gpt-4.1-mini-2025-04-14/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4.1-mini-2025-04-14", + "id": "openai/gpt-4.1-mini-2025-04-14", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4.1-mini-2025-04-14", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.285, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.285", + "raw_stderr": "1.201", + "latency": "3.176", + "cost_per_test": "0.001175", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.201, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.129, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.129", + "raw_stderr": "1.364", + "latency": "3.578", + "cost_per_test": "0.001132", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.364, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.657, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.657", + "raw_stderr": "1.099", + "latency": "5.939", + "cost_per_test": "0.001817", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.099, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.927, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.927", + "raw_stderr": "1.858", + "latency": "4.076", + "cost_per_test": "0.001393", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.858, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.597, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.597", + "raw_stderr": "1.243", + "latency": "2.854", + "cost_per_test": "0.001048", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.243, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 65.325, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "65.325", + "raw_stderr": "1.529", + "latency": "7.809", + "cost_per_test": "0.002335", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.529, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.672, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.672", + "raw_stderr": "1.5", + "latency": "2.805", + "cost_per_test": "0.000979", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.5, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 67.192, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "67.192", + "raw_stderr": "2.405", + "latency": "3.02", + "cost_per_test": "0.001395", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.405, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 53.769, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "53.769", + "raw_stderr": "1.503", + "latency": "3.66", + "cost_per_test": "0.001316", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.503, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.674, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.674", + "raw_stderr": "0.791", + "latency": "4.803", + "cost_per_test": "0.001774", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.791, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.758, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.758", + "raw_stderr": "1.41", + "latency": "2.07", + "cost_per_test": "0.000755", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.41, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.225, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.225", + "raw_stderr": "0.405", + "latency": "3.832", + "cost_per_test": "0.001316", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.405, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 68.737, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "68.737", + "raw_stderr": "2.075", + "latency": "2.836", + "cost_per_test": "0.00094", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.075, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.603, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.603", + "raw_stderr": "1.027", + "latency": "4.687", + "cost_per_test": "0.001481", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.027, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.822, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.822", + "raw_stderr": "1.446", + "latency": "2.333", + "cost_per_test": "0.000888", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.446, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4.1-mini-2025-04-14/7ff71d04-b0e0-4456-9e0e-2448d16b906d.json b/data/vals_ai/openai/gpt-4.1-mini-2025-04-14/7ff71d04-b0e0-4456-9e0e-2448d16b906d.json new file mode 100644 index 0000000000000000000000000000000000000000..f554d79a751e29c8b7de6a7e616b7f039137ee04 --- /dev/null +++ b/data/vals_ai/openai/gpt-4.1-mini-2025-04-14/7ff71d04-b0e0-4456-9e0e-2448d16b906d.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/openai_gpt-4.1-mini-2025-04-14/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4.1-mini-2025-04-14", + "id": "openai/gpt-4.1-mini-2025-04-14", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4.1-mini-2025-04-14", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 59.441, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.441", + "raw_stderr": "1.676", + "latency": "3.307", + "cost_per_test": "0.000891", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.676, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 55.012, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "55.012", + "raw_stderr": "1.698", + "latency": "8.221", + "cost_per_test": "0.051823", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.698, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 57.926, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.926", + "raw_stderr": "0.972", + "latency": "6.443", + "cost_per_test": "0.026113", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.972, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 59.324, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.324", + "raw_stderr": "1.677", + "latency": "7.8", + "cost_per_test": "0.025626", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.677, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4.1-mini-2025-04-14/ab610b21-90fc-4ffd-9b5d-7f6bbad5c59a.json b/data/vals_ai/openai/gpt-4.1-mini-2025-04-14/ab610b21-90fc-4ffd-9b5d-7f6bbad5c59a.json new file mode 100644 index 0000000000000000000000000000000000000000..61830f425667f758aa8433a96037fcfa01224102 --- /dev/null +++ b/data/vals_ai/openai/gpt-4.1-mini-2025-04-14/ab610b21-90fc-4ffd-9b5d-7f6bbad5c59a.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/openai_gpt-4.1-mini-2025-04-14/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4.1-mini-2025-04-14", + "id": "openai/gpt-4.1-mini-2025-04-14", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4.1-mini-2025-04-14", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 95.342, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "95.342", + "raw_stderr": "1.174", + "latency": "2.25", + "cost_per_test": "0.000377", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.174, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:hard:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 21.429, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "21.429", + "raw_stderr": "2.193", + "latency": "62.607", + "cost_per_test": "0.002221", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.193, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:medium:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 57.702, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "57.702", + "raw_stderr": "2.524", + "latency": "17.92", + "cost_per_test": "0.000936", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.524, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:overall:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 58.158, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "58.158", + "raw_stderr": "1.181", + "latency": "27.592", + "cost_per_test": "0.001178", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.181, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4.1-mini-2025-04-14/cd112a35-f0c1-4f85-a275-78379c0d7acb.json b/data/vals_ai/openai/gpt-4.1-mini-2025-04-14/cd112a35-f0c1-4f85-a275-78379c0d7acb.json new file mode 100644 index 0000000000000000000000000000000000000000..efa0a87c16d1e1c8ae194cf8cc7c800aa0296c1a --- /dev/null +++ b/data/vals_ai/openai/gpt-4.1-mini-2025-04-14/cd112a35-f0c1-4f85-a275-78379c0d7acb.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/openai_gpt-4.1-mini-2025-04-14/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4.1-mini-2025-04-14", + "id": "openai/gpt-4.1-mini-2025-04-14", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4.1-mini-2025-04-14", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 60.834, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.834", + "raw_stderr": "1.396", + "latency": "5.921", + "cost_per_test": "0.001233", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.396, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 71.914, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.914", + "raw_stderr": "0.881", + "latency": "5.921", + "cost_per_test": "0.001233", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.881, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 82.993, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "82.993", + "raw_stderr": "1.074", + "latency": "5.921", + "cost_per_test": "0.001233", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.074, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4.1-mini-2025-04-14/d274a146-d1de-47a4-9a54-00299f93f567.json b/data/vals_ai/openai/gpt-4.1-mini-2025-04-14/d274a146-d1de-47a4-9a54-00299f93f567.json new file mode 100644 index 0000000000000000000000000000000000000000..578e08030f404a3fba5f3bf748278c2f5a017564 --- /dev/null +++ b/data/vals_ai/openai/gpt-4.1-mini-2025-04-14/d274a146-d1de-47a4-9a54-00299f93f567.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/openai_gpt-4.1-mini-2025-04-14/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4.1-mini-2025-04-14", + "id": "openai/gpt-4.1-mini-2025-04-14", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4.1-mini-2025-04-14", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 64.646, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "64.646", + "raw_stderr": "3.397", + "latency": "8.974", + "cost_per_test": "0.002537", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.397, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 71.212, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.212", + "raw_stderr": "3.25", + "latency": "9.589", + "cost_per_test": "0.002112", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.25, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 67.929, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "67.929", + "raw_stderr": "2.351", + "latency": "9.282", + "cost_per_test": "0.002324", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.351, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4.1-mini-2025-04-14/df8385af-ba27-4e34-93e5-f51de705e508.json b/data/vals_ai/openai/gpt-4.1-mini-2025-04-14/df8385af-ba27-4e34-93e5-f51de705e508.json new file mode 100644 index 0000000000000000000000000000000000000000..842ca72a7d4c7b207bbcbba29fa049b6284b6d3c --- /dev/null +++ b/data/vals_ai/openai/gpt-4.1-mini-2025-04-14/df8385af-ba27-4e34-93e5-f51de705e508.json @@ -0,0 +1,444 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/openai_gpt-4.1-mini-2025-04-14/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4.1-mini-2025-04-14", + "id": "openai/gpt-4.1-mini-2025-04-14", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4.1-mini-2025-04-14", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.674, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.674", + "raw_stderr": "1.001", + "latency": "0.395", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.001, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 75.92, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "75.92", + "raw_stderr": "0.215", + "latency": "0.402", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.215, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.242, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.242", + "raw_stderr": "0.663", + "latency": "0.41", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.663, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.044, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.044", + "raw_stderr": "0.419", + "latency": "0.467", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.419, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.775, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.775", + "raw_stderr": "0.836", + "latency": "0.437", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.836, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:openai/gpt-4.1-mini-2025-04-14:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 75.61, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "75.61", + "raw_stderr": "1.484", + "latency": "0.69", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.484, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4.1-nano-2025-04-14/518101dd-252b-4abc-8610-957ddf5d5e6c.json b/data/vals_ai/openai/gpt-4.1-nano-2025-04-14/518101dd-252b-4abc-8610-957ddf5d5e6c.json new file mode 100644 index 0000000000000000000000000000000000000000..b0136adabe3ec6d433ce62730dfa7bc8d441f5d2 --- /dev/null +++ b/data/vals_ai/openai/gpt-4.1-nano-2025-04-14/518101dd-252b-4abc-8610-957ddf5d5e6c.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/openai_gpt-4.1-nano-2025-04-14/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4.1-nano-2025-04-14", + "id": "openai/gpt-4.1-nano-2025-04-14", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4.1-nano-2025-04-14", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 85.404, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "85.404", + "raw_stderr": "1.968", + "latency": "2.331", + "cost_per_test": "0.000107", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.968, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:hard:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 8.286, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "8.286", + "raw_stderr": "1.473", + "latency": "22.324", + "cost_per_test": "0.000588", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.473, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:medium:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 34.465, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "34.465", + "raw_stderr": "2.428", + "latency": "10.747", + "cost_per_test": "0.000199", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.428, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:overall:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 42.718, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "42.718", + "raw_stderr": "1.152", + "latency": "11.801", + "cost_per_test": "0.000298", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.152, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4.1-nano-2025-04-14/5cb336fd-6c23-4423-ada4-37e8eee98d12.json b/data/vals_ai/openai/gpt-4.1-nano-2025-04-14/5cb336fd-6c23-4423-ada4-37e8eee98d12.json new file mode 100644 index 0000000000000000000000000000000000000000..1118993c8d9ad51493952396b2c715cc24c61a46 --- /dev/null +++ b/data/vals_ai/openai/gpt-4.1-nano-2025-04-14/5cb336fd-6c23-4423-ada4-37e8eee98d12.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/openai_gpt-4.1-nano-2025-04-14/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4.1-nano-2025-04-14", + "id": "openai/gpt-4.1-nano-2025-04-14", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4.1-nano-2025-04-14", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 48.951, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "48.951", + "raw_stderr": "1.707", + "latency": "3.869", + "cost_per_test": "0.000183", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.707, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 38.462, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "38.462", + "raw_stderr": "1.661", + "latency": "5.948", + "cost_per_test": "0.012861", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.661, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 42.075, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "42.075", + "raw_stderr": "0.968", + "latency": "4.954", + "cost_per_test": "0.006454", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.968, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 38.811, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "38.811", + "raw_stderr": "1.664", + "latency": "5.045", + "cost_per_test": "0.006318", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.664, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4.1-nano-2025-04-14/68ed0542-fecf-4632-88d7-452191f9d61e.json b/data/vals_ai/openai/gpt-4.1-nano-2025-04-14/68ed0542-fecf-4632-88d7-452191f9d61e.json new file mode 100644 index 0000000000000000000000000000000000000000..408cb2d0b346d2c86fdba2db45f95d7ef4a47d11 --- /dev/null +++ b/data/vals_ai/openai/gpt-4.1-nano-2025-04-14/68ed0542-fecf-4632-88d7-452191f9d61e.json @@ -0,0 +1,1071 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/openai_gpt-4.1-nano-2025-04-14/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4.1-nano-2025-04-14", + "id": "openai/gpt-4.1-nano-2025-04-14", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4.1-nano-2025-04-14", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.474, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.474", + "raw_stderr": "1.5", + "latency": "1.914", + "cost_per_test": "0.000292", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.5, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 72.37, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.37", + "raw_stderr": "1.634", + "latency": "1.992", + "cost_per_test": "0.000272", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.634, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 64.223, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "64.223", + "raw_stderr": "1.44", + "latency": "3.327", + "cost_per_test": "0.000433", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.44, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 63.659, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "63.659", + "raw_stderr": "2.385", + "latency": "2.425", + "cost_per_test": "0.000351", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.385, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.223, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.223", + "raw_stderr": "1.562", + "latency": "2.07", + "cost_per_test": "0.00028", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.562, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 35.913, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "35.913", + "raw_stderr": "1.527", + "latency": "5.105", + "cost_per_test": "0.000638", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.527, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 69.682, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.682", + "raw_stderr": "1.626", + "latency": "1.938", + "cost_per_test": "0.000255", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.626, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 55.381, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "55.381", + "raw_stderr": "2.547", + "latency": "1.88", + "cost_per_test": "0.000367", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.547, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 37.511, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "37.511", + "raw_stderr": "1.456", + "latency": "2.468", + "cost_per_test": "0.000364", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.456, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.238, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.238", + "raw_stderr": "1.162", + "latency": "2.936", + "cost_per_test": "0.000442", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.162, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 64.394, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "64.394", + "raw_stderr": "1.577", + "latency": "1.481", + "cost_per_test": "0.000196", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.577, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 63.479, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "63.479", + "raw_stderr": "0.462", + "latency": "2.403", + "cost_per_test": "0.000337", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.462, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 52.906, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "52.906", + "raw_stderr": "2.237", + "latency": "1.806", + "cost_per_test": "0.000241", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.237, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 69.053, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.053", + "raw_stderr": "1.299", + "latency": "2.658", + "cost_per_test": "0.000363", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.299, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.679, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.679", + "raw_stderr": "1.603", + "latency": "1.642", + "cost_per_test": "0.000228", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.603, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4.1-nano-2025-04-14/734aa25c-c1e7-48cb-85e4-fa521ead3871.json b/data/vals_ai/openai/gpt-4.1-nano-2025-04-14/734aa25c-c1e7-48cb-85e4-fa521ead3871.json new file mode 100644 index 0000000000000000000000000000000000000000..6393de801298a72fe8cc701b37664722a648fae3 --- /dev/null +++ b/data/vals_ai/openai/gpt-4.1-nano-2025-04-14/734aa25c-c1e7-48cb-85e4-fa521ead3871.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/openai_gpt-4.1-nano-2025-04-14/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4.1-nano-2025-04-14", + "id": "openai/gpt-4.1-nano-2025-04-14", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4.1-nano-2025-04-14", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 48.569, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "48.569", + "raw_stderr": "1.429", + "latency": "2.998", + "cost_per_test": "0.000291", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.429, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 60.752, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.752", + "raw_stderr": "0.956", + "latency": "2.998", + "cost_per_test": "0.000291", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.956, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 72.935, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "72.935", + "raw_stderr": "1.27", + "latency": "2.998", + "cost_per_test": "0.000291", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.27, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4.1-nano-2025-04-14/bf1b3941-88ff-40aa-9be4-55858aecc0f2.json b/data/vals_ai/openai/gpt-4.1-nano-2025-04-14/bf1b3941-88ff-40aa-9be4-55858aecc0f2.json new file mode 100644 index 0000000000000000000000000000000000000000..4cea8ae967c19ff75dc25762b8f975da94b5ec8b --- /dev/null +++ b/data/vals_ai/openai/gpt-4.1-nano-2025-04-14/bf1b3941-88ff-40aa-9be4-55858aecc0f2.json @@ -0,0 +1,444 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/openai_gpt-4.1-nano-2025-04-14/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4.1-nano-2025-04-14", + "id": "openai/gpt-4.1-nano-2025-04-14", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4.1-nano-2025-04-14", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 65.648, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "65.648", + "raw_stderr": "1.18", + "latency": "0.31", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.18, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 57.822, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "57.822", + "raw_stderr": "0.254", + "latency": "0.334", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.254, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 52.269, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "52.269", + "raw_stderr": "0.83", + "latency": "0.366", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.83, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 61.056, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "61.056", + "raw_stderr": "0.595", + "latency": "0.357", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.595, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 62.804, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "62.804", + "raw_stderr": "1.596", + "latency": "0.353", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.596, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 66.738, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "66.738", + "raw_stderr": "2.04", + "latency": "0.422", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.04, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4.1-nano-2025-04-14/eea587da-8929-451a-b3ed-10d0698ab2e0.json b/data/vals_ai/openai/gpt-4.1-nano-2025-04-14/eea587da-8929-451a-b3ed-10d0698ab2e0.json new file mode 100644 index 0000000000000000000000000000000000000000..81b43ae93e540523b85284a93cf3fedcef72a36c --- /dev/null +++ b/data/vals_ai/openai/gpt-4.1-nano-2025-04-14/eea587da-8929-451a-b3ed-10d0698ab2e0.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/openai_gpt-4.1-nano-2025-04-14/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4.1-nano-2025-04-14", + "id": "openai/gpt-4.1-nano-2025-04-14", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4.1-nano-2025-04-14", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 51.01, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "51.01", + "raw_stderr": "3.438", + "latency": "4.255", + "cost_per_test": "0.000576", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.438, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 50.505, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "50.505", + "raw_stderr": "3.512", + "latency": "4.128", + "cost_per_test": "0.0004", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.512, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 50.758, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "50.758", + "raw_stderr": "2.457", + "latency": "4.192", + "cost_per_test": "0.000488", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.457, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4.1-nano-2025-04-14/f714f1dd-618e-4b19-92c0-3fcf727a6cd2.json b/data/vals_ai/openai/gpt-4.1-nano-2025-04-14/f714f1dd-618e-4b19-92c0-3fcf727a6cd2.json new file mode 100644 index 0000000000000000000000000000000000000000..400297e459aa50be881b87f879949c93f6064769 --- /dev/null +++ b/data/vals_ai/openai/gpt-4.1-nano-2025-04-14/f714f1dd-618e-4b19-92c0-3fcf727a6cd2.json @@ -0,0 +1,105 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/openai_gpt-4.1-nano-2025-04-14/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4.1-nano-2025-04-14", + "id": "openai/gpt-4.1-nano-2025-04-14", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4.1-nano-2025-04-14", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 55.055, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "55.055", + "raw_stderr": "1.196", + "latency": "7.251", + "cost_per_test": "0.000566", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.196, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4.1-nano-2025-04-14/f8a4a1ce-95a4-449a-a025-5a6045e8e028.json b/data/vals_ai/openai/gpt-4.1-nano-2025-04-14/f8a4a1ce-95a4-449a-a025-5a6045e8e028.json new file mode 100644 index 0000000000000000000000000000000000000000..3d1d94ced06d21814ffd59ddce9be1b033d66b90 --- /dev/null +++ b/data/vals_ai/openai/gpt-4.1-nano-2025-04-14/f8a4a1ce-95a4-449a-a025-5a6045e8e028.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/openai_gpt-4.1-nano-2025-04-14/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4.1-nano-2025-04-14", + "id": "openai/gpt-4.1-nano-2025-04-14", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4.1-nano-2025-04-14", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 54.849, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "54.849", + "raw_stderr": "1.403", + "latency": "3.828", + "cost_per_test": "0.000997", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.403, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 52.822, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.822", + "raw_stderr": "0.995", + "latency": "3.858", + "cost_per_test": "0.001003", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.995, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:openai/gpt-4.1-nano-2025-04-14:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 50.795, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "50.795", + "raw_stderr": "1.41", + "latency": "3.889", + "cost_per_test": "0.001009", + "temperature": "1", + "max_output_tokens": "32768", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.41, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4o-2024-05-13/f03f2246-f90b-4438-8ed4-980b5658a646.json b/data/vals_ai/openai/gpt-4o-2024-05-13/f03f2246-f90b-4438-8ed4-980b5658a646.json new file mode 100644 index 0000000000000000000000000000000000000000..387d438a460992176124a81808dec3d9cd50a031 --- /dev/null +++ b/data/vals_ai/openai/gpt-4o-2024-05-13/f03f2246-f90b-4438-8ed4-980b5658a646.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/openai_gpt-4o-2024-05-13/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4o-2024-05-13", + "id": "openai/gpt-4o-2024-05-13", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4o-2024-05-13", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:openai/gpt-4o-2024-05-13:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 50.505, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "50.505", + "raw_stderr": "3.553", + "latency": "10.468", + "cost_per_test": "0.019577", + "temperature": "1", + "max_output_tokens": "4096", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.553, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 4096, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:openai/gpt-4o-2024-05-13:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 50.0, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "50", + "raw_stderr": "3.553", + "latency": "9.688", + "cost_per_test": "0.011302", + "temperature": "1", + "max_output_tokens": "4096", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.553, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 4096, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:openai/gpt-4o-2024-05-13:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 50.252, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "50.252", + "raw_stderr": "2.512", + "latency": "10.078", + "cost_per_test": "0.01544", + "temperature": "1", + "max_output_tokens": "4096", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.512, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 4096, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4o-2024-08-06/0fa6e42d-195b-464b-aeed-03ea04a6256e.json b/data/vals_ai/openai/gpt-4o-2024-08-06/0fa6e42d-195b-464b-aeed-03ea04a6256e.json new file mode 100644 index 0000000000000000000000000000000000000000..811fbe9198ab2b65cac46e7b5e3257bd325365e8 --- /dev/null +++ b/data/vals_ai/openai/gpt-4o-2024-08-06/0fa6e42d-195b-464b-aeed-03ea04a6256e.json @@ -0,0 +1,444 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/openai_gpt-4o-2024-08-06/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4o-2024-08-06", + "id": "openai/gpt-4o-2024-08-06", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4o-2024-08-06", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 86.442, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.442", + "raw_stderr": "0.773", + "latency": "0.343", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.773, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.968, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.968", + "raw_stderr": "0.199", + "latency": "0.367", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.199, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.596, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.596", + "raw_stderr": "0.653", + "latency": "0.42", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.653, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.12, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.12", + "raw_stderr": "0.538", + "latency": "0.532", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.538, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.878, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.878", + "raw_stderr": "0.777", + "latency": "0.353", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.777, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 74.718, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "74.718", + "raw_stderr": "2.358", + "latency": "1.178", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.358, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4o-2024-08-06/6ecc197a-cbf1-46d4-aafc-4010892cb72a.json b/data/vals_ai/openai/gpt-4o-2024-08-06/6ecc197a-cbf1-46d4-aafc-4010892cb72a.json new file mode 100644 index 0000000000000000000000000000000000000000..15628bc453a0726b1678cb88ef37c93429b06fc8 --- /dev/null +++ b/data/vals_ai/openai/gpt-4o-2024-08-06/6ecc197a-cbf1-46d4-aafc-4010892cb72a.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/openai_gpt-4o-2024-08-06/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4o-2024-08-06", + "id": "openai/gpt-4o-2024-08-06", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4o-2024-08-06", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 50.233, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "50.233", + "raw_stderr": "1.707", + "latency": "0.073", + "cost_per_test": "0.004761", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.707, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 36.364, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "36.364", + "raw_stderr": "1.642", + "latency": "0.057", + "cost_per_test": "0.283471", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.642, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 39.433, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "39.433", + "raw_stderr": "0.951", + "latency": "0.046", + "cost_per_test": "0.147391", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.951, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 31.702, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "31.702", + "raw_stderr": "1.589", + "latency": "0.009", + "cost_per_test": "0.15394", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.589, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4o-2024-08-06/9579b2ee-e34b-47d8-9903-92575e8aa70f.json b/data/vals_ai/openai/gpt-4o-2024-08-06/9579b2ee-e34b-47d8-9903-92575e8aa70f.json new file mode 100644 index 0000000000000000000000000000000000000000..f65d4c4dded2dae25d82395ad76d057bb4356f86 --- /dev/null +++ b/data/vals_ai/openai/gpt-4o-2024-08-06/9579b2ee-e34b-47d8-9903-92575e8aa70f.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/openai_gpt-4o-2024-08-06/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4o-2024-08-06", + "id": "openai/gpt-4o-2024-08-06", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4o-2024-08-06", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 60.834, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.834", + "raw_stderr": "1.396", + "latency": "9.543", + "cost_per_test": "0.004417", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.396, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 71.136, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.136", + "raw_stderr": "0.892", + "latency": "9.543", + "cost_per_test": "0.004417", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.892, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 81.439, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "81.439", + "raw_stderr": "1.112", + "latency": "9.543", + "cost_per_test": "0.004417", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.112, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4o-2024-08-06/9e53a04a-e782-4648-9d9a-2e4342bb957f.json b/data/vals_ai/openai/gpt-4o-2024-08-06/9e53a04a-e782-4648-9d9a-2e4342bb957f.json new file mode 100644 index 0000000000000000000000000000000000000000..ec412e3f02f82e9e44cf1f8bdbf60d03e0765a2d --- /dev/null +++ b/data/vals_ai/openai/gpt-4o-2024-08-06/9e53a04a-e782-4648-9d9a-2e4342bb957f.json @@ -0,0 +1,105 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/openai_gpt-4o-2024-08-06/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4o-2024-08-06", + "id": "openai/gpt-4o-2024-08-06", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4o-2024-08-06", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 64.009, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "64.009", + "raw_stderr": "1.154", + "latency": "47.52", + "cost_per_test": "0.006251", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.154, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4o-2024-08-06/c3707a61-d91c-4a9e-8d45-02fa6d4fb2da.json b/data/vals_ai/openai/gpt-4o-2024-08-06/c3707a61-d91c-4a9e-8d45-02fa6d4fb2da.json new file mode 100644 index 0000000000000000000000000000000000000000..7928a0e1da0d4c1695c0de21c213ce9e30f8ff12 --- /dev/null +++ b/data/vals_ai/openai/gpt-4o-2024-08-06/c3707a61-d91c-4a9e-8d45-02fa6d4fb2da.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/openai_gpt-4o-2024-08-06/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4o-2024-08-06", + "id": "openai/gpt-4o-2024-08-06", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4o-2024-08-06", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 71.145, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.145", + "raw_stderr": "2.501", + "latency": "7.274", + "cost_per_test": "0.006779", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.501, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 60.97, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.97", + "raw_stderr": "1.862", + "latency": "7.356", + "cost_per_test": "0.006956", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.862, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 50.795, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "50.795", + "raw_stderr": "2.758", + "latency": "7.437", + "cost_per_test": "0.007133", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.758, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4o-2024-08-06/ec47b0dd-c231-42e9-a573-b1e346f7de01.json b/data/vals_ai/openai/gpt-4o-2024-08-06/ec47b0dd-c231-42e9-a573-b1e346f7de01.json new file mode 100644 index 0000000000000000000000000000000000000000..208de1220057fa558c62d17ff753212a3f6f666a --- /dev/null +++ b/data/vals_ai/openai/gpt-4o-2024-08-06/ec47b0dd-c231-42e9-a573-b1e346f7de01.json @@ -0,0 +1,1071 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/openai_gpt-4o-2024-08-06/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4o-2024-08-06", + "id": "openai/gpt-4o-2024-08-06", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4o-2024-08-06", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.145, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.145", + "raw_stderr": "1.207", + "latency": "7.182", + "cost_per_test": "0.006568", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.207, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.087, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.087", + "raw_stderr": "1.448", + "latency": "7.893", + "cost_per_test": "0.006438", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.448, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.322, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.322", + "raw_stderr": "1.315", + "latency": "14.513", + "cost_per_test": "0.009883", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.315, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.049, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.049", + "raw_stderr": "2.044", + "latency": "10.061", + "cost_per_test": "0.007916", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.044, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.858, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.858", + "raw_stderr": "1.381", + "latency": "6.954", + "cost_per_test": "0.006085", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.381, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 55.108, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "55.108", + "raw_stderr": "1.598", + "latency": "16.182", + "cost_per_test": "0.010385", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.598, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.528, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.528", + "raw_stderr": "1.482", + "latency": "5.636", + "cost_per_test": "0.005575", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.482, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.916, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.916", + "raw_stderr": "2.302", + "latency": "6.494", + "cost_per_test": "0.007865", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.302, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 56.131, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "56.131", + "raw_stderr": "1.496", + "latency": "8.064", + "cost_per_test": "0.00696", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.496, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.905, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.905", + "raw_stderr": "1.11", + "latency": "11.455", + "cost_per_test": "0.009151", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.11, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.165, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.165", + "raw_stderr": "1.381", + "latency": "5.932", + "cost_per_test": "0.004937", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.381, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.13, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.13", + "raw_stderr": "0.42", + "latency": "8.998", + "cost_per_test": "0.007217", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.42, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 67.535, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "67.535", + "raw_stderr": "2.096", + "latency": "5.941", + "cost_per_test": "0.005383", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.096, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.366, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.366", + "raw_stderr": "1.196", + "latency": "12.704", + "cost_per_test": "0.008408", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.196, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:openai/gpt-4o-2024-08-06:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.702, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.702", + "raw_stderr": "1.4", + "latency": "6.961", + "cost_per_test": "0.005488", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.4, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4o-2024-11-20/007e9378-fe78-41ff-b90d-5bc6202aa88f.json b/data/vals_ai/openai/gpt-4o-2024-11-20/007e9378-fe78-41ff-b90d-5bc6202aa88f.json new file mode 100644 index 0000000000000000000000000000000000000000..c224c936a9406e1c1ebc6fa46318906ab3c67df6 --- /dev/null +++ b/data/vals_ai/openai/gpt-4o-2024-11-20/007e9378-fe78-41ff-b90d-5bc6202aa88f.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/openai_gpt-4o-2024-11-20/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4o-2024-11-20", + "id": "openai/gpt-4o-2024-11-20", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4o-2024-11-20", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 55.245, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "55.245", + "raw_stderr": "1.698", + "latency": "2.215", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.698, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 42.657, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "42.657", + "raw_stderr": "1.688", + "latency": "10.92", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.688, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 45.921, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "45.921", + "raw_stderr": "0.973", + "latency": "6.461", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.973, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 39.86, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "39.86", + "raw_stderr": "1.672", + "latency": "6.248", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.672, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4o-2024-11-20/0e3d198e-a758-4354-b4a4-aaa8d5fe0acd.json b/data/vals_ai/openai/gpt-4o-2024-11-20/0e3d198e-a758-4354-b4a4-aaa8d5fe0acd.json new file mode 100644 index 0000000000000000000000000000000000000000..5d48c881b194f2f10e345163af60e00638ac2736 --- /dev/null +++ b/data/vals_ai/openai/gpt-4o-2024-11-20/0e3d198e-a758-4354-b4a4-aaa8d5fe0acd.json @@ -0,0 +1,105 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/openai_gpt-4o-2024-11-20/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4o-2024-11-20", + "id": "openai/gpt-4o-2024-11-20", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4o-2024-11-20", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 62.161, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "62.161", + "raw_stderr": "1.166", + "latency": "45.873", + "cost_per_test": "0.006822", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.166, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4o-2024-11-20/2efcac0e-e268-4d84-97fc-278de6de25be.json b/data/vals_ai/openai/gpt-4o-2024-11-20/2efcac0e-e268-4d84-97fc-278de6de25be.json new file mode 100644 index 0000000000000000000000000000000000000000..d1e8714351335635518473b24fcb3173e4c0c11d --- /dev/null +++ b/data/vals_ai/openai/gpt-4o-2024-11-20/2efcac0e-e268-4d84-97fc-278de6de25be.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/openai_gpt-4o-2024-11-20/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4o-2024-11-20", + "id": "openai/gpt-4o-2024-11-20", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4o-2024-11-20", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 54.04, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "54.04", + "raw_stderr": "3.542", + "latency": "15.717", + "cost_per_test": "0.011827", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.542, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 53.535, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "53.535", + "raw_stderr": "3.55", + "latency": "14.131", + "cost_per_test": "0.007103", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.55, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 53.788, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "53.788", + "raw_stderr": "2.507", + "latency": "14.924", + "cost_per_test": "0.009465", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.507, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4o-2024-11-20/30519b14-4116-41a4-801b-b33828a2835d.json b/data/vals_ai/openai/gpt-4o-2024-11-20/30519b14-4116-41a4-801b-b33828a2835d.json new file mode 100644 index 0000000000000000000000000000000000000000..493e76f0d1d760956f90090aa83480c714b2609a --- /dev/null +++ b/data/vals_ai/openai/gpt-4o-2024-11-20/30519b14-4116-41a4-801b-b33828a2835d.json @@ -0,0 +1,1071 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/openai_gpt-4o-2024-11-20/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4o-2024-11-20", + "id": "openai/gpt-4o-2024-11-20", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4o-2024-11-20", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.1, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.1", + "raw_stderr": "1.468", + "latency": "8.08", + "cost_per_test": "0.007273", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.468, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.693, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.693", + "raw_stderr": "1.533", + "latency": "7.518", + "cost_per_test": "0.006093", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.533, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.265, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.265", + "raw_stderr": "1.315", + "latency": "13.454", + "cost_per_test": "0.010192", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.315, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.854, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.854", + "raw_stderr": "2.258", + "latency": "9.572", + "cost_per_test": "0.008429", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.258, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.28, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.28", + "raw_stderr": "1.407", + "latency": "7.483", + "cost_per_test": "0.00664", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.407, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 52.632, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "52.632", + "raw_stderr": "1.606", + "latency": "15.557", + "cost_per_test": "0.010722", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.606, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.306, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.306", + "raw_stderr": "1.525", + "latency": "6.075", + "cost_per_test": "0.005606", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.525, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 65.617, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "65.617", + "raw_stderr": "2.54", + "latency": "5.575", + "cost_per_test": "0.008195", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.54, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 48.683, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "48.683", + "raw_stderr": "1.481", + "latency": "13.023", + "cost_per_test": "0.008599", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.481, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.871, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.871", + "raw_stderr": "1.263", + "latency": "10.95", + "cost_per_test": "0.009548", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.263, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.273, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.273", + "raw_stderr": "1.397", + "latency": "5.526", + "cost_per_test": "0.004344", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.397, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 72.563, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.563", + "raw_stderr": "0.441", + "latency": "9.143", + "cost_per_test": "0.007465", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.441, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.142, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.142", + "raw_stderr": "2.056", + "latency": "6.384", + "cost_per_test": "0.005199", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.056, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.211, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.211", + "raw_stderr": "1.271", + "latency": "12.097", + "cost_per_test": "0.008214", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.271, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.957, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.957", + "raw_stderr": "1.35", + "latency": "6.706", + "cost_per_test": "0.005454", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.35, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4o-2024-11-20/44f74082-b3b7-49cf-ba32-f5727c181bc1.json b/data/vals_ai/openai/gpt-4o-2024-11-20/44f74082-b3b7-49cf-ba32-f5727c181bc1.json new file mode 100644 index 0000000000000000000000000000000000000000..c04b3dc318fc51bc800a0d968eea996bd23139a0 --- /dev/null +++ b/data/vals_ai/openai/gpt-4o-2024-11-20/44f74082-b3b7-49cf-ba32-f5727c181bc1.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/openai_gpt-4o-2024-11-20/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4o-2024-11-20", + "id": "openai/gpt-4o-2024-11-20", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4o-2024-11-20", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 86.025, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "86.025", + "raw_stderr": "1.932", + "latency": "2.741", + "cost_per_test": "0.00269", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.932, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:hard:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 5.143, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "5.143", + "raw_stderr": "1.181", + "latency": "5.656", + "cost_per_test": "0.005244", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.181, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:medium:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 39.164, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "39.164", + "raw_stderr": "2.494", + "latency": "4.643", + "cost_per_test": "0.004063", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.494, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:overall:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 43.444, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "43.444", + "raw_stderr": "1.123", + "latency": "4.347", + "cost_per_test": "0.003999", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.123, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4o-2024-11-20/47da4134-44f1-4c29-a50c-ddaceb6863fc.json b/data/vals_ai/openai/gpt-4o-2024-11-20/47da4134-44f1-4c29-a50c-ddaceb6863fc.json new file mode 100644 index 0000000000000000000000000000000000000000..e0bf9b5f988e1d81e55f5c189216234fe1681822 --- /dev/null +++ b/data/vals_ai/openai/gpt-4o-2024-11-20/47da4134-44f1-4c29-a50c-ddaceb6863fc.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/openai_gpt-4o-2024-11-20/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4o-2024-11-20", + "id": "openai/gpt-4o-2024-11-20", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4o-2024-11-20", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 64.865, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.865", + "raw_stderr": "2.634", + "latency": "5.85", + "cost_per_test": "0.006935", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.634, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 57.432, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.432", + "raw_stderr": "1.907", + "latency": "5.984", + "cost_per_test": "0.007141", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.907, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 50.0, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "50", + "raw_stderr": "2.759", + "latency": "6.119", + "cost_per_test": "0.007347", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.759, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4o-2024-11-20/9b4fb1ac-1ccb-4ec2-a2d7-6246dcdf6877.json b/data/vals_ai/openai/gpt-4o-2024-11-20/9b4fb1ac-1ccb-4ec2-a2d7-6246dcdf6877.json new file mode 100644 index 0000000000000000000000000000000000000000..fc3a678fef2176cd2e844c3fe2e0b561e0a8cc9f --- /dev/null +++ b/data/vals_ai/openai/gpt-4o-2024-11-20/9b4fb1ac-1ccb-4ec2-a2d7-6246dcdf6877.json @@ -0,0 +1,444 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/openai_gpt-4o-2024-11-20/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4o-2024-11-20", + "id": "openai/gpt-4o-2024-11-20", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4o-2024-11-20", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.208, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.208", + "raw_stderr": "0.818", + "latency": "0.328", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.818, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.91, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.91", + "raw_stderr": "0.209", + "latency": "0.325", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.209, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.568, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.568", + "raw_stderr": "0.615", + "latency": "0.381", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.615, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.214, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.214", + "raw_stderr": "0.463", + "latency": "0.452", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.463, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.954, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.954", + "raw_stderr": "0.817", + "latency": "0.343", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.817, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.429, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.429", + "raw_stderr": "1.9", + "latency": "0.882", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.9, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4o-2024-11-20/ca858926-5079-4289-944b-24252d091eaf.json b/data/vals_ai/openai/gpt-4o-2024-11-20/ca858926-5079-4289-944b-24252d091eaf.json new file mode 100644 index 0000000000000000000000000000000000000000..53ec86b1b450159db18177fcc7dae5bdbc0eb6e5 --- /dev/null +++ b/data/vals_ai/openai/gpt-4o-2024-11-20/ca858926-5079-4289-944b-24252d091eaf.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/openai_gpt-4o-2024-11-20/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4o-2024-11-20", + "id": "openai/gpt-4o-2024-11-20", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4o-2024-11-20", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 64.841, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.841", + "raw_stderr": "1.365", + "latency": "5.855", + "cost_per_test": "0.005806", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.365, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 74.53, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "74.53", + "raw_stderr": "0.859", + "latency": "5.855", + "cost_per_test": "0.005806", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.859, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:openai/gpt-4o-2024-11-20:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 84.219, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "84.219", + "raw_stderr": "1.042", + "latency": "5.855", + "cost_per_test": "0.005806", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.042, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4o-mini-2024-07-18/00d2ce85-85b4-4e13-bcc2-d9d6cf17428b.json b/data/vals_ai/openai/gpt-4o-mini-2024-07-18/00d2ce85-85b4-4e13-bcc2-d9d6cf17428b.json new file mode 100644 index 0000000000000000000000000000000000000000..60424f875a57e8747aa02ccda6048369935d644b --- /dev/null +++ b/data/vals_ai/openai/gpt-4o-mini-2024-07-18/00d2ce85-85b4-4e13-bcc2-d9d6cf17428b.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/openai_gpt-4o-mini-2024-07-18/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4o-mini-2024-07-18", + "id": "openai/gpt-4o-mini-2024-07-18", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4o-mini-2024-07-18", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 60.175, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.175", + "raw_stderr": "2.701", + "latency": "5.931", + "cost_per_test": "0.000727", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.701, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 54.492, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "54.492", + "raw_stderr": "1.93", + "latency": "6.059", + "cost_per_test": "0.000822", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.93, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 48.808, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "48.808", + "raw_stderr": "2.758", + "latency": "6.187", + "cost_per_test": "0.000916", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.758, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4o-mini-2024-07-18/1940a7a9-913d-456e-b0a6-aab0ccdb4c32.json b/data/vals_ai/openai/gpt-4o-mini-2024-07-18/1940a7a9-913d-456e-b0a6-aab0ccdb4c32.json new file mode 100644 index 0000000000000000000000000000000000000000..78546f0f9f37d33bc900075b9316ccfe43e4d02a --- /dev/null +++ b/data/vals_ai/openai/gpt-4o-mini-2024-07-18/1940a7a9-913d-456e-b0a6-aab0ccdb4c32.json @@ -0,0 +1,105 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/openai_gpt-4o-mini-2024-07-18/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4o-mini-2024-07-18", + "id": "openai/gpt-4o-mini-2024-07-18", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4o-mini-2024-07-18", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 56.557, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "56.557", + "raw_stderr": "1.191", + "latency": "45.016", + "cost_per_test": "0.00037", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.191, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4o-mini-2024-07-18/558782f5-4265-4af2-987f-bf6271e21a83.json b/data/vals_ai/openai/gpt-4o-mini-2024-07-18/558782f5-4265-4af2-987f-bf6271e21a83.json new file mode 100644 index 0000000000000000000000000000000000000000..c34ba45f9cfbd3d85b7020597ff422cb983f3e17 --- /dev/null +++ b/data/vals_ai/openai/gpt-4o-mini-2024-07-18/558782f5-4265-4af2-987f-bf6271e21a83.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/openai_gpt-4o-mini-2024-07-18/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4o-mini-2024-07-18", + "id": "openai/gpt-4o-mini-2024-07-18", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4o-mini-2024-07-18", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 52.214, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.214", + "raw_stderr": "1.705", + "latency": "0.012", + "cost_per_test": "0.000268", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.705, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 41.026, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "41.026", + "raw_stderr": "1.679", + "latency": "0.055", + "cost_per_test": "0.017023", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.679, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 45.455, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "45.455", + "raw_stderr": "0.977", + "latency": "0.041", + "cost_per_test": "0.008848", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.977, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 43.124, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "43.124", + "raw_stderr": "1.691", + "latency": "0.056", + "cost_per_test": "0.009254", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.691, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4o-mini-2024-07-18/a83ec126-e5ff-4f6d-a805-181c8dfac5e2.json b/data/vals_ai/openai/gpt-4o-mini-2024-07-18/a83ec126-e5ff-4f6d-a805-181c8dfac5e2.json new file mode 100644 index 0000000000000000000000000000000000000000..504d339c6f7bbfb4c23d68de074e5eadb3955ae5 --- /dev/null +++ b/data/vals_ai/openai/gpt-4o-mini-2024-07-18/a83ec126-e5ff-4f6d-a805-181c8dfac5e2.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/openai_gpt-4o-mini-2024-07-18/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4o-mini-2024-07-18", + "id": "openai/gpt-4o-mini-2024-07-18", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4o-mini-2024-07-18", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 44.949, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "44.949", + "raw_stderr": "3.535", + "latency": "15.039", + "cost_per_test": "0.00067", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 43.434, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "43.434", + "raw_stderr": "3.523", + "latency": "15.6", + "cost_per_test": "0.000403", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.523, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 44.192, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "44.192", + "raw_stderr": "2.495", + "latency": "15.32", + "cost_per_test": "0.000536", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.495, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4o-mini-2024-07-18/e1dfe529-7af3-4768-a730-00aba936d323.json b/data/vals_ai/openai/gpt-4o-mini-2024-07-18/e1dfe529-7af3-4768-a730-00aba936d323.json new file mode 100644 index 0000000000000000000000000000000000000000..1e8a29d4e76345d04d06f493bf64a85d9c1419dc --- /dev/null +++ b/data/vals_ai/openai/gpt-4o-mini-2024-07-18/e1dfe529-7af3-4768-a730-00aba936d323.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/openai_gpt-4o-mini-2024-07-18/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4o-mini-2024-07-18", + "id": "openai/gpt-4o-mini-2024-07-18", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4o-mini-2024-07-18", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 60.87, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "60.87", + "raw_stderr": "2.72", + "latency": "6.066", + "cost_per_test": "0.000237", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.72, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:hard:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 1.429, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "1.429", + "raw_stderr": "0.634", + "latency": "8.949", + "cost_per_test": "0.000362", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.634, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:medium:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 16.971, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "16.971", + "raw_stderr": "1.918", + "latency": "7.076", + "cost_per_test": "0.000295", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.918, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:overall:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 26.423, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "26.423", + "raw_stderr": "1.129", + "latency": "7.364", + "cost_per_test": "0.000298", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.129, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4o-mini-2024-07-18/f7dad6b9-12a8-473f-80d1-3384c29ea860.json b/data/vals_ai/openai/gpt-4o-mini-2024-07-18/f7dad6b9-12a8-473f-80d1-3384c29ea860.json new file mode 100644 index 0000000000000000000000000000000000000000..d3dce38ea10f6903351a7297739b84236ef935c9 --- /dev/null +++ b/data/vals_ai/openai/gpt-4o-mini-2024-07-18/f7dad6b9-12a8-473f-80d1-3384c29ea860.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/openai_gpt-4o-mini-2024-07-18/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4o-mini-2024-07-18", + "id": "openai/gpt-4o-mini-2024-07-18", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4o-mini-2024-07-18", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 49.55, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "49.55", + "raw_stderr": "1.43", + "latency": "8.899", + "cost_per_test": "0.000323", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.43, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 60.548, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.548", + "raw_stderr": "0.963", + "latency": "8.899", + "cost_per_test": "0.000323", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.963, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 71.545, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.545", + "raw_stderr": "1.29", + "latency": "8.899", + "cost_per_test": "0.000323", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.29, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-4o-mini-2024-07-18/f97b3782-e167-4ceb-9104-686fdd9f3dba.json b/data/vals_ai/openai/gpt-4o-mini-2024-07-18/f97b3782-e167-4ceb-9104-686fdd9f3dba.json new file mode 100644 index 0000000000000000000000000000000000000000..bf37de6d7938cc7ceae233eb872e9407680d1be3 --- /dev/null +++ b/data/vals_ai/openai/gpt-4o-mini-2024-07-18/f97b3782-e167-4ceb-9104-686fdd9f3dba.json @@ -0,0 +1,1071 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/openai_gpt-4o-mini-2024-07-18/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-4o-mini-2024-07-18", + "id": "openai/gpt-4o-mini-2024-07-18", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-4o-mini-2024-07-18", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.195, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.195", + "raw_stderr": "1.488", + "latency": "4.753", + "cost_per_test": "0.000416", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.488, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 70.342, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.342", + "raw_stderr": "1.626", + "latency": "5.379", + "cost_per_test": "0.000405", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.626, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 62.279, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "62.279", + "raw_stderr": "1.441", + "latency": "8.068", + "cost_per_test": "0.000573", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.441, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 64.39, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "64.39", + "raw_stderr": "2.365", + "latency": "5.662", + "cost_per_test": "0.000462", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.365, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.341, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.341", + "raw_stderr": "1.522", + "latency": "4.476", + "cost_per_test": "0.000352", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.522, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 39.216, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "39.216", + "raw_stderr": "1.568", + "latency": "8.948", + "cost_per_test": "0.000561", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.568, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 67.482, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "67.482", + "raw_stderr": "1.638", + "latency": "3.66", + "cost_per_test": "0.000298", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.638, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 54.593, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "54.593", + "raw_stderr": "2.551", + "latency": "2.482", + "cost_per_test": "0.00042", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.551, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 37.693, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "37.693", + "raw_stderr": "1.461", + "latency": "3.593", + "cost_per_test": "0.000363", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.461, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.241, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.241", + "raw_stderr": "1.19", + "latency": "6.517", + "cost_per_test": "0.000567", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.19, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 66.45, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.45", + "raw_stderr": "1.553", + "latency": "2.835", + "cost_per_test": "0.000253", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.553, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 62.735, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "62.735", + "raw_stderr": "0.461", + "latency": "5.086", + "cost_per_test": "0.000412", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.461, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 50.701, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "50.701", + "raw_stderr": "2.238", + "latency": "4.171", + "cost_per_test": "0.000319", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.238, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 64.434, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "64.434", + "raw_stderr": "1.328", + "latency": "7.722", + "cost_per_test": "0.000508", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.328, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:openai/gpt-4o-mini-2024-07-18:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 72.932, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.932", + "raw_stderr": "1.573", + "latency": "2.934", + "cost_per_test": "0.000277", + "temperature": "1", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.573, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-2025-08-07/01bc71cf-3e96-4bcb-a313-8c4d8f704f62.json b/data/vals_ai/openai/gpt-5-2025-08-07/01bc71cf-3e96-4bcb-a313-8c4d8f704f62.json new file mode 100644 index 0000000000000000000000000000000000000000..f1299d1ca6c2f0f27ce346c0b465752054abe2c9 --- /dev/null +++ b/data/vals_ai/openai/gpt-5-2025-08-07/01bc71cf-3e96-4bcb-a313-8c4d8f704f62.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/openai_gpt-5-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-2025-08-07", + "id": "openai/gpt-5-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 84.848, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.848", + "raw_stderr": "2.548", + "latency": "191.702", + "cost_per_test": "0.074134", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.548, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 86.364, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.364", + "raw_stderr": "2.439", + "latency": "147.729", + "cost_per_test": "0.071283", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.439, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 85.606, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.606", + "raw_stderr": "1.764", + "latency": "169.716", + "cost_per_test": "0.072708", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.764, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-2025-08-07/1563affc-84b6-4a7d-a570-3dd95dbb964f.json b/data/vals_ai/openai/gpt-5-2025-08-07/1563affc-84b6-4a7d-a570-3dd95dbb964f.json new file mode 100644 index 0000000000000000000000000000000000000000..b59592ce1e0f9426d723d081d07ac16d01025290 --- /dev/null +++ b/data/vals_ai/openai/gpt-5-2025-08-07/1563affc-84b6-4a7d-a570-3dd95dbb964f.json @@ -0,0 +1,1071 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/openai_gpt-5-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-2025-08-07", + "id": "openai/gpt-5-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.026, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.026", + "raw_stderr": "0.951", + "latency": "24.712", + "cost_per_test": "0.01654", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.951, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.381, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.381", + "raw_stderr": "0.999", + "latency": "49.206", + "cost_per_test": "0.023832", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.999, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.841, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.841", + "raw_stderr": "0.898", + "latency": "43.47", + "cost_per_test": "0.032013", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.898, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.805, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.805", + "raw_stderr": "1.616", + "latency": "30.035", + "cost_per_test": "0.021717", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.616, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.573, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.573", + "raw_stderr": "1.052", + "latency": "18.946", + "cost_per_test": "0.014362", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.052, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.599, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.599", + "raw_stderr": "1.27", + "latency": "98.624", + "cost_per_test": "0.062582", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.27, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.807, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.807", + "raw_stderr": "1.377", + "latency": "28.123", + "cost_per_test": "0.017724", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.377, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.265, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.265", + "raw_stderr": "2.077", + "latency": "23.049", + "cost_per_test": "0.01705", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.077, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.295, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.295", + "raw_stderr": "1.3", + "latency": "54.392", + "cost_per_test": "0.034743", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.3, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.745, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.745", + "raw_stderr": "0.607", + "latency": "36.702", + "cost_per_test": "0.023739", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.607, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.957, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.957", + "raw_stderr": "1.176", + "latency": "22.307", + "cost_per_test": "0.015416", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.176, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.544, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.544", + "raw_stderr": "0.336", + "latency": "37.239", + "cost_per_test": "0.024135", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.336, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.172, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.172", + "raw_stderr": "1.545", + "latency": "25.258", + "cost_per_test": "0.016946", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.545, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.685, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.685", + "raw_stderr": "0.806", + "latency": "44.486", + "cost_per_test": "0.027143", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.806, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.469, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.469", + "raw_stderr": "1.172", + "latency": "22.032", + "cost_per_test": "0.014089", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.172, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-2025-08-07/17cf9c67-340d-466e-b5eb-224c1fa3e6ad.json b/data/vals_ai/openai/gpt-5-2025-08-07/17cf9c67-340d-466e-b5eb-224c1fa3e6ad.json new file mode 100644 index 0000000000000000000000000000000000000000..fddd454864b2a24c2db8d0cbfbd67109ba81e165 --- /dev/null +++ b/data/vals_ai/openai/gpt-5-2025-08-07/17cf9c67-340d-466e-b5eb-224c1fa3e6ad.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/openai_gpt-5-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-2025-08-07", + "id": "openai/gpt-5-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 11.0, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "11", + "raw_stderr": "4.967", + "latency": "2627.572", + "cost_per_test": "1.870345", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.967, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 29.0, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "29", + "raw_stderr": "7.699", + "latency": "2588.931", + "cost_per_test": "1.388182", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.699, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "ioi:overall:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "20", + "raw_stderr": "4.581", + "latency": "2608.252", + "cost_per_test": "1.629264", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.581, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-2025-08-07/1872e24c-9f3f-4364-95f9-7b0b87bf61a4.json b/data/vals_ai/openai/gpt-5-2025-08-07/1872e24c-9f3f-4364-95f9-7b0b87bf61a4.json new file mode 100644 index 0000000000000000000000000000000000000000..de01711d66a4d4e01b0cdc9d1650a49462b6d077 --- /dev/null +++ b/data/vals_ai/openai/gpt-5-2025-08-07/1872e24c-9f3f-4364-95f9-7b0b87bf61a4.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/openai_gpt-5-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-2025-08-07", + "id": "openai/gpt-5-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 18.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "18", + "raw_stderr": "3.842", + "latency": "4579.325", + "cost_per_test": "0.852361", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.842, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-2025-08-07/3320f7b8-1f22-4062-a69c-98e214cbbc3a.json b/data/vals_ai/openai/gpt-5-2025-08-07/3320f7b8-1f22-4062-a69c-98e214cbbc3a.json new file mode 100644 index 0000000000000000000000000000000000000000..3f654e6f77cab37c1f4a6c7287444700d7b02efb --- /dev/null +++ b/data/vals_ai/openai/gpt-5-2025-08-07/3320f7b8-1f22-4062-a69c-98e214cbbc3a.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/openai_gpt-5-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-2025-08-07", + "id": "openai/gpt-5-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 20.088, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "20.088", + "raw_stderr": "3.409", + "latency": "1852.149", + "cost_per_test": "1.525167", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.409, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-2025-08-07/3537a4d3-c765-4464-8489-816efc49c6c2.json b/data/vals_ai/openai/gpt-5-2025-08-07/3537a4d3-c765-4464-8489-816efc49c6c2.json new file mode 100644 index 0000000000000000000000000000000000000000..b65f460a14c977dc06e4fe31bb636fb22e54a304 --- /dev/null +++ b/data/vals_ai/openai/gpt-5-2025-08-07/3537a4d3-c765-4464-8489-816efc49c6c2.json @@ -0,0 +1,356 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/openai_gpt-5-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-2025-08-07", + "id": "openai/gpt-5-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 39.683, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "39.683", + "raw_stderr": "6.139", + "latency": "71.949", + "cost_per_test": "0.03839", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.139, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:automata:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 29.762, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "29.762", + "raw_stderr": "5.782", + "latency": "51.314", + "cost_per_test": "0.027903", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.782, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:calculus:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 62.963, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "62.963", + "raw_stderr": "6.225", + "latency": "94.04", + "cost_per_test": "0.060657", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.225, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 61.596, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "61.596", + "raw_stderr": "5.463", + "latency": "52.517", + "cost_per_test": "0.030379", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.463, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:overall:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 43.68, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "43.68", + "raw_stderr": "3.349", + "latency": "58.593", + "cost_per_test": "0.032224", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.349, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-2025-08-07/68c5d5be-b5f0-4c3e-a115-a93b9f34ef60.json b/data/vals_ai/openai/gpt-5-2025-08-07/68c5d5be-b5f0-4c3e-a115-a93b9f34ef60.json new file mode 100644 index 0000000000000000000000000000000000000000..5b3a8c20a9d8142a453289bf2e374f837bc43f43 --- /dev/null +++ b/data/vals_ai/openai/gpt-5-2025-08-07/68c5d5be-b5f0-4c3e-a115-a93b9f34ef60.json @@ -0,0 +1,444 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/openai_gpt-5-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-2025-08-07", + "id": "openai/gpt-5-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 90.884, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.884", + "raw_stderr": "0.601", + "latency": "25.395", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.601, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.031, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.031", + "raw_stderr": "0.197", + "latency": "14.309", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.197, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 89.55, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.55", + "raw_stderr": "0.581", + "latency": "10.666", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.581, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 86.023, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.023", + "raw_stderr": "0.379", + "latency": "18.59", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.379, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.968, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.968", + "raw_stderr": "0.854", + "latency": "9.146", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.854, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 85.682, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.682", + "raw_stderr": "1.456", + "latency": "33.436", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.456, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-2025-08-07/784c3ad3-b645-457c-a7f8-02a49363ca12.json b/data/vals_ai/openai/gpt-5-2025-08-07/784c3ad3-b645-457c-a7f8-02a49363ca12.json new file mode 100644 index 0000000000000000000000000000000000000000..7979601df831118b22fb16487840ff85a6887f7e --- /dev/null +++ b/data/vals_ai/openai/gpt-5-2025-08-07/784c3ad3-b645-457c-a7f8-02a49363ca12.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/openai_gpt-5-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-2025-08-07", + "id": "openai/gpt-5-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 62.796, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.796", + "raw_stderr": "1.382", + "latency": "78.668", + "cost_per_test": "0.053873", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.382, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 73.385, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "73.385", + "raw_stderr": "0.868", + "latency": "78.668", + "cost_per_test": "0.053873", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.868, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 83.974, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "83.974", + "raw_stderr": "1.049", + "latency": "78.668", + "cost_per_test": "0.053873", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.049, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-2025-08-07/82392bec-832b-4e0d-ac60-1498c97eebb4.json b/data/vals_ai/openai/gpt-5-2025-08-07/82392bec-832b-4e0d-ac60-1498c97eebb4.json new file mode 100644 index 0000000000000000000000000000000000000000..8f73d46d5d9c4dcb2b633c6378a9977c55ec0623 --- /dev/null +++ b/data/vals_ai/openai/gpt-5-2025-08-07/82392bec-832b-4e0d-ac60-1498c97eebb4.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/openai_gpt-5-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-2025-08-07", + "id": "openai/gpt-5-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.758, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.758", + "raw_stderr": "0.617", + "latency": "27.107", + "cost_per_test": "0.018785", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.617, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "lcb:hard:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 66.286, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66.286", + "raw_stderr": "2.527", + "latency": "225.456", + "cost_per_test": "0.107337", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.527, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "lcb:medium:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 92.689, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "92.689", + "raw_stderr": "1.33", + "latency": "198.761", + "cost_per_test": "0.06513", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.33, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "lcb:overall:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 85.911, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "85.911", + "raw_stderr": "0.974", + "latency": "150.441", + "cost_per_test": "0.063751", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.974, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-2025-08-07/9b72ec42-6839-4c81-8a68-074bcbccb60a.json b/data/vals_ai/openai/gpt-5-2025-08-07/9b72ec42-6839-4c81-8a68-074bcbccb60a.json new file mode 100644 index 0000000000000000000000000000000000000000..c9efecbcb0b0794346e463cd1e8eac6d42b915bb --- /dev/null +++ b/data/vals_ai/openai/gpt-5-2025-08-07/9b72ec42-6839-4c81-8a68-074bcbccb60a.json @@ -0,0 +1,371 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/openai_gpt-5-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-2025-08-07", + "id": "openai/gpt-5-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 26.19, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "26.19", + "raw_stderr": "6.867", + "latency": "332.943", + "cost_per_test": "0.407381", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.867, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 65.134, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "65.134", + "raw_stderr": "2.955", + "latency": "248.171", + "cost_per_test": "0.294607", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.955, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 84.021, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "84.021", + "raw_stderr": "2.638", + "latency": "189.415", + "cost_per_test": "0.205578", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.638, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "306.121", + "cost_per_test": "0.420087", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:overall:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 69.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "69", + "raw_stderr": "2.07", + "latency": "232.842", + "cost_per_test": "0.27029", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.07, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-2025-08-07/a88ffdbf-fe58-469e-bffc-b66e06faf00b.json b/data/vals_ai/openai/gpt-5-2025-08-07/a88ffdbf-fe58-469e-bffc-b66e06faf00b.json new file mode 100644 index 0000000000000000000000000000000000000000..c959b16d2e3e15df72548482ff36d6fa5fd0f1cb --- /dev/null +++ b/data/vals_ai/openai/gpt-5-2025-08-07/a88ffdbf-fe58-469e-bffc-b66e06faf00b.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/openai_gpt-5-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-2025-08-07", + "id": "openai/gpt-5-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 76.773, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "76.773", + "raw_stderr": "1.058", + "latency": "79.525", + "cost_per_test": "0.037789", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.058, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 65.454, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.454", + "raw_stderr": "0.879", + "latency": "62.828", + "cost_per_test": "0.028906", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.879, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 54.134, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "54.134", + "raw_stderr": "1.405", + "latency": "46.132", + "cost_per_test": "0.020023", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.405, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-2025-08-07/ac21ed86-479a-4430-9ca4-454535b8f59e.json b/data/vals_ai/openai/gpt-5-2025-08-07/ac21ed86-479a-4430-9ca4-454535b8f59e.json new file mode 100644 index 0000000000000000000000000000000000000000..8ffb1cc9c719b4f13a29027d34286f8b4418e4fb --- /dev/null +++ b/data/vals_ai/openai/gpt-5-2025-08-07/ac21ed86-479a-4430-9ca4-454535b8f59e.json @@ -0,0 +1,105 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/openai_gpt-5-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-2025-08-07", + "id": "openai/gpt-5-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 81.503, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.503", + "raw_stderr": "0.934", + "latency": "72.114", + "cost_per_test": "0.032532", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.934, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-2025-08-07/b3e902ff-68d1-4564-9078-2c53ac46e96f.json b/data/vals_ai/openai/gpt-5-2025-08-07/b3e902ff-68d1-4564-9078-2c53ac46e96f.json new file mode 100644 index 0000000000000000000000000000000000000000..55475544a0eaa91b05017889f4dc36ffc144022d --- /dev/null +++ b/data/vals_ai/openai/gpt-5-2025-08-07/b3e902ff-68d1-4564-9078-2c53ac46e96f.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/openai_gpt-5-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-2025-08-07", + "id": "openai/gpt-5-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 59.79, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.79", + "raw_stderr": "1.674", + "latency": "148.152", + "cost_per_test": "0.003682", + "max_output_tokens": "32000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.674, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 32000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.072, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.072", + "raw_stderr": "1.665", + "latency": "175.462", + "cost_per_test": "0.175312", + "max_output_tokens": "32000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.665, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 32000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.072, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.072", + "raw_stderr": "0.961", + "latency": "160.705", + "cost_per_test": "0.086116", + "max_output_tokens": "32000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.961, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 32000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 62.354, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.354", + "raw_stderr": "1.654", + "latency": "158.5", + "cost_per_test": "0.079354", + "max_output_tokens": "32000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.654, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 32000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-2025-08-07/c420f553-2683-45fb-9ae6-e600b1097ca4.json b/data/vals_ai/openai/gpt-5-2025-08-07/c420f553-2683-45fb-9ae6-e600b1097ca4.json new file mode 100644 index 0000000000000000000000000000000000000000..07cfad8fe6a0af729dae2d6e6b667c5fb979905a --- /dev/null +++ b/data/vals_ai/openai/gpt-5-2025-08-07/c420f553-2683-45fb-9ae6-e600b1097ca4.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/openai_gpt-5-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-2025-08-07", + "id": "openai/gpt-5-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:openai/gpt-5-2025-08-07:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 83.65, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "83.65", + "raw_stderr": "1.936", + "latency": "184.536", + "cost_per_test": "0.1015", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.936, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-codex/4b8d8ecd-18d2-410a-b037-f65dd0a5d8de.json b/data/vals_ai/openai/gpt-5-codex/4b8d8ecd-18d2-410a-b037-f65dd0a5d8de.json new file mode 100644 index 0000000000000000000000000000000000000000..930b9dc433bdc0e24a25c38d565e73607724a5f3 --- /dev/null +++ b/data/vals_ai/openai/gpt-5-codex/4b8d8ecd-18d2-410a-b037-f65dd0a5d8de.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/openai_gpt-5-codex/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-codex", + "id": "openai/gpt-5-codex", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-codex", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:openai/gpt-5-codex:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 2.833, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "2.833", + "raw_stderr": "2.586", + "latency": "1650.37", + "cost_per_test": "2.583712", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.586, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:openai/gpt-5-codex:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 16.667, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "16.667", + "raw_stderr": "15.215", + "latency": "1435.417", + "cost_per_test": "2.756198", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 15.215, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "ioi:overall:openai/gpt-5-codex:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 9.75, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "9.75", + "raw_stderr": "7.717", + "latency": "1542.894", + "cost_per_test": "2.669955", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.717, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-codex/b2453a75-912e-4544-a194-7d04cdeef964.json b/data/vals_ai/openai/gpt-5-codex/b2453a75-912e-4544-a194-7d04cdeef964.json new file mode 100644 index 0000000000000000000000000000000000000000..d3128b9724b50f71f177e342fbc702d8b65495c0 --- /dev/null +++ b/data/vals_ai/openai/gpt-5-codex/b2453a75-912e-4544-a194-7d04cdeef964.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/openai_gpt-5-codex/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-codex", + "id": "openai/gpt-5-codex", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-codex", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:openai/gpt-5-codex:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.137, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.137", + "raw_stderr": "0.754", + "latency": "8.824", + "cost_per_test": "0.006587", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.754, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:hard:openai/gpt-5-codex:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 68.571, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "68.571", + "raw_stderr": "2.481", + "latency": "323.06", + "cost_per_test": "0.228984", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.481, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:medium:openai/gpt-5-codex:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 87.467, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "87.467", + "raw_stderr": "1.692", + "latency": "73.006", + "cost_per_test": "0.056783", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.692, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:overall:openai/gpt-5-codex:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 84.725, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "84.725", + "raw_stderr": "1.032", + "latency": "134.963", + "cost_per_test": "0.097451", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.032, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-mini-2025-08-07/141b0cea-e484-4455-ab58-eea67b25dbba.json b/data/vals_ai/openai/gpt-5-mini-2025-08-07/141b0cea-e484-4455-ab58-eea67b25dbba.json new file mode 100644 index 0000000000000000000000000000000000000000..05e4f3f5103b81da278901a18f6196359734b360 --- /dev/null +++ b/data/vals_ai/openai/gpt-5-mini-2025-08-07/141b0cea-e484-4455-ab58-eea67b25dbba.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/openai_gpt-5-mini-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-mini-2025-08-07", + "id": "openai/gpt-5-mini-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-mini-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 75.437, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "75.437", + "raw_stderr": "1.214", + "latency": "15.693", + "cost_per_test": "0.005493", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.214, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 66.892, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.892", + "raw_stderr": "0.923", + "latency": "24.864", + "cost_per_test": "0.00436", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.923, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 58.347, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.347", + "raw_stderr": "1.39", + "latency": "34.036", + "cost_per_test": "0.003226", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.39, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-mini-2025-08-07/68b08a45-265b-4ec2-909e-886fd47c2fc1.json b/data/vals_ai/openai/gpt-5-mini-2025-08-07/68b08a45-265b-4ec2-909e-886fd47c2fc1.json new file mode 100644 index 0000000000000000000000000000000000000000..b32840dab69a4502e761da5948352989b26cd9ab --- /dev/null +++ b/data/vals_ai/openai/gpt-5-mini-2025-08-07/68b08a45-265b-4ec2-909e-886fd47c2fc1.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/openai_gpt-5-mini-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-mini-2025-08-07", + "id": "openai/gpt-5-mini-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-mini-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 14.171, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "14.171", + "raw_stderr": "3.539", + "latency": "698.246", + "cost_per_test": "0.248587", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.539, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-mini-2025-08-07/6c6f81d5-0a99-418c-a473-8b712ab8de65.json b/data/vals_ai/openai/gpt-5-mini-2025-08-07/6c6f81d5-0a99-418c-a473-8b712ab8de65.json new file mode 100644 index 0000000000000000000000000000000000000000..9c88256b9f04a1e57a0d6272ab9b72d6d2a803b9 --- /dev/null +++ b/data/vals_ai/openai/gpt-5-mini-2025-08-07/6c6f81d5-0a99-418c-a473-8b712ab8de65.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/openai_gpt-5-mini-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-mini-2025-08-07", + "id": "openai/gpt-5-mini-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-mini-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 9.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "9", + "raw_stderr": "2.862", + "latency": "3349.817", + "cost_per_test": "0.22021", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.862, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-mini-2025-08-07/87729c12-2b19-42bb-8650-bac1df107522.json b/data/vals_ai/openai/gpt-5-mini-2025-08-07/87729c12-2b19-42bb-8650-bac1df107522.json new file mode 100644 index 0000000000000000000000000000000000000000..d0842f22fd9175269c65c4ae1d53c00dadc57abd --- /dev/null +++ b/data/vals_ai/openai/gpt-5-mini-2025-08-07/87729c12-2b19-42bb-8650-bac1df107522.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/openai_gpt-5-mini-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-mini-2025-08-07", + "id": "openai/gpt-5-mini-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-mini-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 80.577, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "80.577", + "raw_stderr": "1.924", + "latency": "275.717", + "cost_per_test": "0.033478", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.924, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-mini-2025-08-07/8da89793-e131-4c20-8ff3-619cabb7f601.json b/data/vals_ai/openai/gpt-5-mini-2025-08-07/8da89793-e131-4c20-8ff3-619cabb7f601.json new file mode 100644 index 0000000000000000000000000000000000000000..0bac272af123777956422f63a47a3c8b5ed034d7 --- /dev/null +++ b/data/vals_ai/openai/gpt-5-mini-2025-08-07/8da89793-e131-4c20-8ff3-619cabb7f601.json @@ -0,0 +1,1071 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/openai_gpt-5-mini-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-mini-2025-08-07", + "id": "openai/gpt-5-mini-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-mini-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.353, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.353", + "raw_stderr": "1.05", + "latency": "10.41", + "cost_per_test": "0.003454", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.05, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.593, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.593", + "raw_stderr": "1.132", + "latency": "23.791", + "cost_per_test": "0.006643", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.132, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.046, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.046", + "raw_stderr": "0.928", + "latency": "21.705", + "cost_per_test": "0.0064", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.928, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.829, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.829", + "raw_stderr": "1.67", + "latency": "15.009", + "cost_per_test": "0.004107", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.67, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.73, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.73", + "raw_stderr": "1.168", + "latency": "27.505", + "cost_per_test": "0.00379", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.168, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.613, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.613", + "raw_stderr": "1.398", + "latency": "89.219", + "cost_per_test": "0.013803", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.398, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.484, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.484", + "raw_stderr": "1.437", + "latency": "10.916", + "cost_per_test": "0.003669", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.437, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.654, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.654", + "raw_stderr": "2.309", + "latency": "14.023", + "cost_per_test": "0.003525", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.309, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 62.761, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "62.761", + "raw_stderr": "1.457", + "latency": "22.692", + "cost_per_test": "0.006819", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.457, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.301, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.301", + "raw_stderr": "0.631", + "latency": "12.822", + "cost_per_test": "0.004542", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.631, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.545, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.545", + "raw_stderr": "1.327", + "latency": "14.719", + "cost_per_test": "0.003481", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.327, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.226, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.226", + "raw_stderr": "0.373", + "latency": "22.196", + "cost_per_test": "0.005196", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.373, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.351, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.351", + "raw_stderr": "1.929", + "latency": "15.556", + "cost_per_test": "0.00387", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.929, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.453, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.453", + "raw_stderr": "0.852", + "latency": "17.887", + "cost_per_test": "0.005539", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.852, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.456, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.456", + "raw_stderr": "1.346", + "latency": "14.49", + "cost_per_test": "0.003107", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.346, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-mini-2025-08-07/91b25d5d-d987-4103-b5b6-cd815de99a12.json b/data/vals_ai/openai/gpt-5-mini-2025-08-07/91b25d5d-d987-4103-b5b6-cd815de99a12.json new file mode 100644 index 0000000000000000000000000000000000000000..c81c21dffe4bb964ec41f684db0e160e530cbd1f --- /dev/null +++ b/data/vals_ai/openai/gpt-5-mini-2025-08-07/91b25d5d-d987-4103-b5b6-cd815de99a12.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/openai_gpt-5-mini-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-mini-2025-08-07", + "id": "openai/gpt-5-mini-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-mini-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 64.759, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.759", + "raw_stderr": "1.366", + "latency": "37.434", + "cost_per_test": "0.011528", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.366, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 75.225, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "75.225", + "raw_stderr": "0.847", + "latency": "37.434", + "cost_per_test": "0.011528", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.847, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 85.691, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "85.691", + "raw_stderr": "1.001", + "latency": "37.434", + "cost_per_test": "0.011528", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.001, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-mini-2025-08-07/943f0265-a146-40b2-88f1-947fc3ebefbd.json b/data/vals_ai/openai/gpt-5-mini-2025-08-07/943f0265-a146-40b2-88f1-947fc3ebefbd.json new file mode 100644 index 0000000000000000000000000000000000000000..54f482cc5ca574291120be11393af1be8c69e70b --- /dev/null +++ b/data/vals_ai/openai/gpt-5-mini-2025-08-07/943f0265-a146-40b2-88f1-947fc3ebefbd.json @@ -0,0 +1,444 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/openai_gpt-5-mini-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-mini-2025-08-07", + "id": "openai/gpt-5-mini-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-mini-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.272, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.272", + "raw_stderr": "0.672", + "latency": "11.836", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.672, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.723, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.723", + "raw_stderr": "0.207", + "latency": "7.401", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.207, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.09, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.09", + "raw_stderr": "0.574", + "latency": "6.488", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.574, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.77, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.77", + "raw_stderr": "0.452", + "latency": "9.432", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.452, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.957, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.957", + "raw_stderr": "0.955", + "latency": "4.951", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.955, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 74.806, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "74.806", + "raw_stderr": "1.834", + "latency": "16.482", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.834, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-mini-2025-08-07/b13d6f5b-d9c0-42a1-b352-bfc6f43674c0.json b/data/vals_ai/openai/gpt-5-mini-2025-08-07/b13d6f5b-d9c0-42a1-b352-bfc6f43674c0.json new file mode 100644 index 0000000000000000000000000000000000000000..ebf8daaee51c9ba994c76d9f646dc1958e2c2779 --- /dev/null +++ b/data/vals_ai/openai/gpt-5-mini-2025-08-07/b13d6f5b-d9c0-42a1-b352-bfc6f43674c0.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/openai_gpt-5-mini-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-mini-2025-08-07", + "id": "openai/gpt-5-mini-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-mini-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 58.858, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.858", + "raw_stderr": "1.68", + "latency": "17.494", + "cost_per_test": "0.003881", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.68, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.49, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.49", + "raw_stderr": "1.669", + "latency": "93.929", + "cost_per_test": "0.039253", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.669, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.179, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.179", + "raw_stderr": "0.965", + "latency": "73.788", + "cost_per_test": "0.020976", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.965, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.189, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.189", + "raw_stderr": "1.664", + "latency": "109.941", + "cost_per_test": "0.019795", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.664, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-mini-2025-08-07/b692c600-2a16-4ed1-b220-7bc037f04ace.json b/data/vals_ai/openai/gpt-5-mini-2025-08-07/b692c600-2a16-4ed1-b220-7bc037f04ace.json new file mode 100644 index 0000000000000000000000000000000000000000..b2d086e35f66f862e803d02608a784ca2ea4abfe --- /dev/null +++ b/data/vals_ai/openai/gpt-5-mini-2025-08-07/b692c600-2a16-4ed1-b220-7bc037f04ace.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/openai_gpt-5-mini-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-mini-2025-08-07", + "id": "openai/gpt-5-mini-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-mini-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 6.5, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "6.5", + "raw_stderr": "4.37", + "latency": "272.805", + "cost_per_test": "0.033967", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.37, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "high" + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 7.0, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "7", + "raw_stderr": "4.989", + "latency": "507.95", + "cost_per_test": "0.044368", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.989, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "high" + } + } + }, + { + "evaluation_result_id": "ioi:overall:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 6.75, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "6.75", + "raw_stderr": "3.316", + "latency": "390.378", + "cost_per_test": "0.039168", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.316, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-mini-2025-08-07/b84577fb-3ba4-4c96-868c-e4c637f24464.json b/data/vals_ai/openai/gpt-5-mini-2025-08-07/b84577fb-3ba4-4c96-868c-e4c637f24464.json new file mode 100644 index 0000000000000000000000000000000000000000..1ec066f1030fbf4ffbf87a7aebca5eca9f1ba5e3 --- /dev/null +++ b/data/vals_ai/openai/gpt-5-mini-2025-08-07/b84577fb-3ba4-4c96-868c-e4c637f24464.json @@ -0,0 +1,356 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/openai_gpt-5-mini-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-mini-2025-08-07", + "id": "openai/gpt-5-mini-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-mini-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 36.508, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "36.508", + "raw_stderr": "6.042", + "latency": "44.522", + "cost_per_test": "0.005441", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.042, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:automata:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 29.762, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "29.762", + "raw_stderr": "5.852", + "latency": "33.513", + "cost_per_test": "0.004528", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.852, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:calculus:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 39.815, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "39.815", + "raw_stderr": "6.484", + "latency": "40.17", + "cost_per_test": "0.010142", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.484, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 62.695, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "62.695", + "raw_stderr": "5.432", + "latency": "31.536", + "cost_per_test": "0.005581", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.432, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:overall:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 42.988, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "42.988", + "raw_stderr": "3.338", + "latency": "36.524", + "cost_per_test": "0.005183", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.338, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-mini-2025-08-07/c0cf1a35-b3cd-430a-88ab-adb485daeb47.json b/data/vals_ai/openai/gpt-5-mini-2025-08-07/c0cf1a35-b3cd-430a-88ab-adb485daeb47.json new file mode 100644 index 0000000000000000000000000000000000000000..7893110c05be3491abb0a64ae5b5b33480d45762 --- /dev/null +++ b/data/vals_ai/openai/gpt-5-mini-2025-08-07/c0cf1a35-b3cd-430a-88ab-adb485daeb47.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/openai_gpt-5-mini-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-mini-2025-08-07", + "id": "openai/gpt-5-mini-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-mini-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 78.283, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.283", + "raw_stderr": "2.93", + "latency": "39.938", + "cost_per_test": "0.014157", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.93, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 82.323, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.323", + "raw_stderr": "2.711", + "latency": "94.328", + "cost_per_test": "0.013029", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.711, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 80.303, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.303", + "raw_stderr": "1.996", + "latency": "67.133", + "cost_per_test": "0.013593", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.996, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-mini-2025-08-07/c3991124-33e4-4f9f-9a9e-33693503b9ef.json b/data/vals_ai/openai/gpt-5-mini-2025-08-07/c3991124-33e4-4f9f-9a9e-33693503b9ef.json new file mode 100644 index 0000000000000000000000000000000000000000..4e8be6e9bb3589a74f4a756c5572c9b8b1e13fd6 --- /dev/null +++ b/data/vals_ai/openai/gpt-5-mini-2025-08-07/c3991124-33e4-4f9f-9a9e-33693503b9ef.json @@ -0,0 +1,105 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/openai_gpt-5-mini-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-mini-2025-08-07", + "id": "openai/gpt-5-mini-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-mini-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 78.914, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.914", + "raw_stderr": "0.98", + "latency": "26.703", + "cost_per_test": "0.007152", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.98, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-mini-2025-08-07/d123cfaa-31d9-41e5-bfe1-9878cb80d9cc.json b/data/vals_ai/openai/gpt-5-mini-2025-08-07/d123cfaa-31d9-41e5-bfe1-9878cb80d9cc.json new file mode 100644 index 0000000000000000000000000000000000000000..469ff59eba03366a080fc6ad7f99fa802af1a6ee --- /dev/null +++ b/data/vals_ai/openai/gpt-5-mini-2025-08-07/d123cfaa-31d9-41e5-bfe1-9878cb80d9cc.json @@ -0,0 +1,371 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/openai_gpt-5-mini-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-mini-2025-08-07", + "id": "openai/gpt-5-mini-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-mini-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 16.667, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "16.667", + "raw_stderr": "5.82", + "latency": "251.467", + "cost_per_test": "0.077516", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.82, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 54.789, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "54.789", + "raw_stderr": "3.087", + "latency": "198.869", + "cost_per_test": "0.054211", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.087, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 78.866, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "78.866", + "raw_stderr": "2.939", + "latency": "156.395", + "cost_per_test": "0.039188", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.939, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "252.487", + "cost_per_test": "0.063025", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:overall:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 60.8, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "60.8", + "raw_stderr": "2.185", + "latency": "187.129", + "cost_per_test": "0.050393", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.185, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-mini-2025-08-07/f30030ac-aae3-448d-8ad4-ea82df3ab7d8.json b/data/vals_ai/openai/gpt-5-mini-2025-08-07/f30030ac-aae3-448d-8ad4-ea82df3ab7d8.json new file mode 100644 index 0000000000000000000000000000000000000000..96bf3e043db2885d104eaeadb74e04d4e9db59bc --- /dev/null +++ b/data/vals_ai/openai/gpt-5-mini-2025-08-07/f30030ac-aae3-448d-8ad4-ea82df3ab7d8.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/openai_gpt-5-mini-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-mini-2025-08-07", + "id": "openai/gpt-5-mini-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-mini-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 99.068, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "99.068", + "raw_stderr": "0.535", + "latency": "6.74", + "cost_per_test": "0.002057", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "lcb:hard:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 71.714, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "71.714", + "raw_stderr": "2.407", + "latency": "68.622", + "cost_per_test": "0.024106", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.407, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "lcb:medium:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 89.034, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "89.034", + "raw_stderr": "1.597", + "latency": "25.656", + "cost_per_test": "0.008427", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.597, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "lcb:overall:openai/gpt-5-mini-2025-08-07:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 86.605, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "86.605", + "raw_stderr": "0.979", + "latency": "33.673", + "cost_per_test": "0.01153", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.979, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-nano-2025-08-07/151cd8f1-00c7-4d47-86ae-6997329954e7.json b/data/vals_ai/openai/gpt-5-nano-2025-08-07/151cd8f1-00c7-4d47-86ae-6997329954e7.json new file mode 100644 index 0000000000000000000000000000000000000000..c7adcf953d58527fc34580537d72266c122de08b --- /dev/null +++ b/data/vals_ai/openai/gpt-5-nano-2025-08-07/151cd8f1-00c7-4d47-86ae-6997329954e7.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/openai_gpt-5-nano-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-nano-2025-08-07", + "id": "openai/gpt-5-nano-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-nano-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 12.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "12", + "raw_stderr": "3.25", + "latency": "3729.837", + "cost_per_test": "0.049218", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.25, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-nano-2025-08-07/342e59f8-c3da-4bc3-8746-7137da1b89a4.json b/data/vals_ai/openai/gpt-5-nano-2025-08-07/342e59f8-c3da-4bc3-8746-7137da1b89a4.json new file mode 100644 index 0000000000000000000000000000000000000000..ac29ea5bc7767262c8457e067582102ddae59f8d --- /dev/null +++ b/data/vals_ai/openai/gpt-5-nano-2025-08-07/342e59f8-c3da-4bc3-8746-7137da1b89a4.json @@ -0,0 +1,444 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/openai_gpt-5-nano-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-nano-2025-08-07", + "id": "openai/gpt-5-nano-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-nano-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.278, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.278", + "raw_stderr": "1.124", + "latency": "25.918", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.124, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 47.563, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "47.563", + "raw_stderr": "0.284", + "latency": "23.351", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.284, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 52.874, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "52.874", + "raw_stderr": "0.956", + "latency": "17.937", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.956, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 50.129, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "50.129", + "raw_stderr": "0.472", + "latency": "22.496", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.472, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 30.641, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "30.641", + "raw_stderr": "0.839", + "latency": "14.893", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.839, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 42.29, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "42.29", + "raw_stderr": "1.618", + "latency": "30.381", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.618, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-nano-2025-08-07/46379784-71d3-440e-b56a-9601a1d2efa7.json b/data/vals_ai/openai/gpt-5-nano-2025-08-07/46379784-71d3-440e-b56a-9601a1d2efa7.json new file mode 100644 index 0000000000000000000000000000000000000000..7aa95dd07959db2230d47d2d4a8e67a684944462 --- /dev/null +++ b/data/vals_ai/openai/gpt-5-nano-2025-08-07/46379784-71d3-440e-b56a-9601a1d2efa7.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/openai_gpt-5-nano-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-nano-2025-08-07", + "id": "openai/gpt-5-nano-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-nano-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 72.865, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "72.865", + "raw_stderr": "1.891", + "latency": "112.909", + "cost_per_test": "0.006961", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.891, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-nano-2025-08-07/4bed656f-06a5-4326-b14b-924a66d7c91b.json b/data/vals_ai/openai/gpt-5-nano-2025-08-07/4bed656f-06a5-4326-b14b-924a66d7c91b.json new file mode 100644 index 0000000000000000000000000000000000000000..84f17c9c3fb543b14dd0854d72430a94eefbccad --- /dev/null +++ b/data/vals_ai/openai/gpt-5-nano-2025-08-07/4bed656f-06a5-4326-b14b-924a66d7c91b.json @@ -0,0 +1,1071 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/openai_gpt-5-nano-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-nano-2025-08-07", + "id": "openai/gpt-5-nano-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-nano-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.332, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.332", + "raw_stderr": "1.283", + "latency": "17.004", + "cost_per_test": "0.001283", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.283, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.51, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.51", + "raw_stderr": "1.352", + "latency": "22.853", + "cost_per_test": "0.001792", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.352, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.424, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.424", + "raw_stderr": "1.049", + "latency": "27.441", + "cost_per_test": "0.00228", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.049, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.415, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.415", + "raw_stderr": "1.837", + "latency": "17.478", + "cost_per_test": "0.001542", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.837, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.517, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.517", + "raw_stderr": "1.336", + "latency": "17.711", + "cost_per_test": "0.001325", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.336, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 66.151, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.151", + "raw_stderr": "1.52", + "latency": "52.506", + "cost_per_test": "0.004676", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.52, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.227, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.227", + "raw_stderr": "1.548", + "latency": "30.193", + "cost_per_test": "0.001306", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.548, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 61.942, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "61.942", + "raw_stderr": "2.487", + "latency": "38.253", + "cost_per_test": "0.001643", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.487, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 50.5, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "50.5", + "raw_stderr": "1.507", + "latency": "32.578", + "cost_per_test": "0.003021", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.507, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.82, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.82", + "raw_stderr": "0.702", + "latency": "18.394", + "cost_per_test": "0.001493", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.702, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 72.619, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.619", + "raw_stderr": "1.467", + "latency": "30.014", + "cost_per_test": "0.00129", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.467, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.067, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.067", + "raw_stderr": "0.412", + "latency": "26.107", + "cost_per_test": "0.001887", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.412, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 65.13, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "65.13", + "raw_stderr": "2.133", + "latency": "22.338", + "cost_per_test": "0.001592", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.133, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.527, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.527", + "raw_stderr": "0.978", + "latency": "22.595", + "cost_per_test": "0.002019", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.978, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.82, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.82", + "raw_stderr": "1.471", + "latency": "16.134", + "cost_per_test": "0.001158", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.471, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-nano-2025-08-07/6381f1c7-6071-45db-abc4-79f1fc9723bc.json b/data/vals_ai/openai/gpt-5-nano-2025-08-07/6381f1c7-6071-45db-abc4-79f1fc9723bc.json new file mode 100644 index 0000000000000000000000000000000000000000..88847e81950bdda548b9f6119e87074c5611410e --- /dev/null +++ b/data/vals_ai/openai/gpt-5-nano-2025-08-07/6381f1c7-6071-45db-abc4-79f1fc9723bc.json @@ -0,0 +1,356 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/openai_gpt-5-nano-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-nano-2025-08-07", + "id": "openai/gpt-5-nano-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-nano-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 30.159, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "30.159", + "raw_stderr": "5.759", + "latency": "33.45", + "cost_per_test": "0.001982", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.759, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:automata:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 5.159, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "5.159", + "raw_stderr": "2.825", + "latency": "44.351", + "cost_per_test": "0.002358", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.825, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:calculus:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 7.716, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "7.716", + "raw_stderr": "3.55", + "latency": "78.88", + "cost_per_test": "0.004291", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.55, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 55.813, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "55.813", + "raw_stderr": "5.578", + "latency": "38.342", + "cost_per_test": "0.002429", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.578, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:overall:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 30.377, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "30.377", + "raw_stderr": "2.834", + "latency": "38.714", + "cost_per_test": "0.002256", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.834, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-nano-2025-08-07/7884c97f-7b89-4182-a81c-f684a5a34f7f.json b/data/vals_ai/openai/gpt-5-nano-2025-08-07/7884c97f-7b89-4182-a81c-f684a5a34f7f.json new file mode 100644 index 0000000000000000000000000000000000000000..6b41b7fa777d594edbac58313e9d6ed33b092c1f --- /dev/null +++ b/data/vals_ai/openai/gpt-5-nano-2025-08-07/7884c97f-7b89-4182-a81c-f684a5a34f7f.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/openai_gpt-5-nano-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-nano-2025-08-07", + "id": "openai/gpt-5-nano-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-nano-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 64.141, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "64.141", + "raw_stderr": "3.408", + "latency": "44.156", + "cost_per_test": "0.004136", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.408, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 62.626, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "62.626", + "raw_stderr": "3.535", + "latency": "55.277", + "cost_per_test": "0.00502", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 63.384, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "63.384", + "raw_stderr": "2.455", + "latency": "49.716", + "cost_per_test": "0.004578", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.455, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-nano-2025-08-07/954938a8-eca1-4626-a3d8-6ed0a0a5d48e.json b/data/vals_ai/openai/gpt-5-nano-2025-08-07/954938a8-eca1-4626-a3d8-6ed0a0a5d48e.json new file mode 100644 index 0000000000000000000000000000000000000000..970e2881b3e80aafcff7b404aaf5ce6cfa390486 --- /dev/null +++ b/data/vals_ai/openai/gpt-5-nano-2025-08-07/954938a8-eca1-4626-a3d8-6ed0a0a5d48e.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/openai_gpt-5-nano-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-nano-2025-08-07", + "id": "openai/gpt-5-nano-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-nano-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 54.702, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "54.702", + "raw_stderr": "1.423", + "latency": "69.944", + "cost_per_test": "0.004904", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.423, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 67.376, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.376", + "raw_stderr": "0.913", + "latency": "69.944", + "cost_per_test": "0.004904", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.913, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 80.049, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "80.049", + "raw_stderr": "1.143", + "latency": "69.944", + "cost_per_test": "0.004904", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.143, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-nano-2025-08-07/b687cadd-d38a-4117-bddb-e6386f691883.json b/data/vals_ai/openai/gpt-5-nano-2025-08-07/b687cadd-d38a-4117-bddb-e6386f691883.json new file mode 100644 index 0000000000000000000000000000000000000000..60794ac02a72a23d57e0887bdf47a804b9064e56 --- /dev/null +++ b/data/vals_ai/openai/gpt-5-nano-2025-08-07/b687cadd-d38a-4117-bddb-e6386f691883.json @@ -0,0 +1,105 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/openai_gpt-5-nano-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-nano-2025-08-07", + "id": "openai/gpt-5-nano-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-nano-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 70.942, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.942", + "raw_stderr": "1.091", + "latency": "30.776", + "cost_per_test": "0.002619", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.091, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-nano-2025-08-07/b8a17a6e-7861-46a7-8703-9f8d3775ded2.json b/data/vals_ai/openai/gpt-5-nano-2025-08-07/b8a17a6e-7861-46a7-8703-9f8d3775ded2.json new file mode 100644 index 0000000000000000000000000000000000000000..194952def716afb9017655d114877a8a3ca69326 --- /dev/null +++ b/data/vals_ai/openai/gpt-5-nano-2025-08-07/b8a17a6e-7861-46a7-8703-9f8d3775ded2.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/openai_gpt-5-nano-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-nano-2025-08-07", + "id": "openai/gpt-5-nano-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-nano-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 99.068, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "99.068", + "raw_stderr": "0.535", + "latency": "12.902", + "cost_per_test": "0.001137", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "lcb:hard:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 22.286, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "22.286", + "raw_stderr": "2.224", + "latency": "271.445", + "cost_per_test": "0.00966", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.224, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "lcb:medium:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 89.295, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "89.295", + "raw_stderr": "1.58", + "latency": "124.73", + "cost_per_test": "0.004362", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.58, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "lcb:overall:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 70.216, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "70.216", + "raw_stderr": "0.927", + "latency": "136.359", + "cost_per_test": "0.005053", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.927, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5-nano-2025-08-07/bc5d55b1-d227-4b48-80c7-a59dadd7ce11.json b/data/vals_ai/openai/gpt-5-nano-2025-08-07/bc5d55b1-d227-4b48-80c7-a59dadd7ce11.json new file mode 100644 index 0000000000000000000000000000000000000000..5fa5e459ad3a6da868561257d50d056c1bc8f428 --- /dev/null +++ b/data/vals_ai/openai/gpt-5-nano-2025-08-07/bc5d55b1-d227-4b48-80c7-a59dadd7ce11.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/openai_gpt-5-nano-2025-08-07/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5-nano-2025-08-07", + "id": "openai/gpt-5-nano-2025-08-07", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5-nano-2025-08-07", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 64.865, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.865", + "raw_stderr": "1.346", + "latency": "49.316", + "cost_per_test": "0.003034", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.346, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 53.617, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "53.617", + "raw_stderr": "0.969", + "latency": "33.364", + "cost_per_test": "0.002295", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.969, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:openai/gpt-5-nano-2025-08-07:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 42.369, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "42.369", + "raw_stderr": "1.393", + "latency": "17.413", + "cost_per_test": "0.001556", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "verbosity": "medium", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.393, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high", + "verbosity": "medium" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.1-2025-11-13/27fcc700-ef0e-4ce8-8b27-60958b913723.json b/data/vals_ai/openai/gpt-5.1-2025-11-13/27fcc700-ef0e-4ce8-8b27-60958b913723.json new file mode 100644 index 0000000000000000000000000000000000000000..ed7d0f592c1936ec72f8f51b809e934b73034917 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.1-2025-11-13/27fcc700-ef0e-4ce8-8b27-60958b913723.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/openai_gpt-5.1-2025-11-13/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.1-2025-11-13", + "id": "openai/gpt-5.1-2025-11-13", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.1-2025-11-13", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 63.054, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.054", + "raw_stderr": "1.648", + "latency": "24.022", + "cost_per_test": "0.004427", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.648, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 63.403, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.403", + "raw_stderr": "1.644", + "latency": "48.958", + "cost_per_test": "0.181071", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.644, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 63.831, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.831", + "raw_stderr": "0.947", + "latency": "35.289", + "cost_per_test": "0.091991", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.947, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 65.035, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.035", + "raw_stderr": "1.628", + "latency": "32.888", + "cost_per_test": "0.090475", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.628, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.1-2025-11-13/2bc15588-df39-4217-a838-f46b74e96835.json b/data/vals_ai/openai/gpt-5.1-2025-11-13/2bc15588-df39-4217-a838-f46b74e96835.json new file mode 100644 index 0000000000000000000000000000000000000000..3e9e3f10e117fac46e89c5887c193126a8d3b2e3 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.1-2025-11-13/2bc15588-df39-4217-a838-f46b74e96835.json @@ -0,0 +1,356 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/openai_gpt-5.1-2025-11-13/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.1-2025-11-13", + "id": "openai/gpt-5.1-2025-11-13", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.1-2025-11-13", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 42.857, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "42.857", + "raw_stderr": "6.21", + "latency": "116.648", + "cost_per_test": "0.003056", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.21, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:automata:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 18.254, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "18.254", + "raw_stderr": "4.915", + "latency": "86.624", + "cost_per_test": "0.005948", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.915, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:calculus:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 69.753, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "69.753", + "raw_stderr": "5.652", + "latency": "119.216", + "cost_per_test": "0.014167", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.652, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 68.595, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "68.595", + "raw_stderr": "5.213", + "latency": "91.348", + "cost_per_test": "0.00551", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.213, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:overall:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 43.235, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "43.235", + "raw_stderr": "3.16", + "latency": "98.207", + "cost_per_test": "0.004838", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.16, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.1-2025-11-13/2d8b8ff8-9cdd-4dd7-996e-70a9d964dc60.json b/data/vals_ai/openai/gpt-5.1-2025-11-13/2d8b8ff8-9cdd-4dd7-996e-70a9d964dc60.json new file mode 100644 index 0000000000000000000000000000000000000000..8a445c5d67f8db0ef3bb54611ca75baa5920e717 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.1-2025-11-13/2d8b8ff8-9cdd-4dd7-996e-70a9d964dc60.json @@ -0,0 +1,438 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/openai_gpt-5.1-2025-11-13/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.1-2025-11-13", + "id": "openai/gpt-5.1-2025-11-13", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.1-2025-11-13", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 89.425, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.425", + "raw_stderr": "0.644", + "latency": "7.573", + "cost_per_test": "0.00546", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.644, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.416, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.416", + "raw_stderr": "0.208", + "latency": "6.983", + "cost_per_test": "0.004895", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.208, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 90.011, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.011", + "raw_stderr": "0.566", + "latency": "5.379", + "cost_per_test": "0.004186", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.566, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 85.683, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.683", + "raw_stderr": "0.397", + "latency": "7.048", + "cost_per_test": "0.004796", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.397, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.946, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.946", + "raw_stderr": "0.917", + "latency": "4.484", + "cost_per_test": "0.00244", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.917, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.616, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.616", + "raw_stderr": "1.521", + "latency": "10.822", + "cost_per_test": "0.006998", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.521, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.1-2025-11-13/33cc6e76-a2d6-4ca2-973b-7f3f65d45d89.json b/data/vals_ai/openai/gpt-5.1-2025-11-13/33cc6e76-a2d6-4ca2-973b-7f3f65d45d89.json new file mode 100644 index 0000000000000000000000000000000000000000..4aefbb7f49882e6be7f3328101ddc5e26b6f2334 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.1-2025-11-13/33cc6e76-a2d6-4ca2-973b-7f3f65d45d89.json @@ -0,0 +1,103 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/openai_gpt-5.1-2025-11-13/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.1-2025-11-13", + "id": "openai/gpt-5.1-2025-11-13", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.1-2025-11-13", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 83.179, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.179", + "raw_stderr": "0.899", + "latency": "75.307", + "cost_per_test": "0.002764", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.899, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.1-2025-11-13/5a2dcee4-9764-4525-bb40-82bbc281073c.json b/data/vals_ai/openai/gpt-5.1-2025-11-13/5a2dcee4-9764-4525-bb40-82bbc281073c.json new file mode 100644 index 0000000000000000000000000000000000000000..2f757a68ac57adfba01f93b097afaba8cee98fb4 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.1-2025-11-13/5a2dcee4-9764-4525-bb40-82bbc281073c.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/openai_gpt-5.1-2025-11-13/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.1-2025-11-13", + "id": "openai/gpt-5.1-2025-11-13", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.1-2025-11-13", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 86.364, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.364", + "raw_stderr": "2.439", + "latency": "83.303", + "cost_per_test": "0.077042", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.439, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 86.869, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.869", + "raw_stderr": "2.4", + "latency": "88.114", + "cost_per_test": "0.072084", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.4, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 86.616, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.616", + "raw_stderr": "1.711", + "latency": "85.708", + "cost_per_test": "0.074563", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.711, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.1-2025-11-13/8a7f4dd1-925a-494f-949e-29b101653793.json b/data/vals_ai/openai/gpt-5.1-2025-11-13/8a7f4dd1-925a-494f-949e-29b101653793.json new file mode 100644 index 0000000000000000000000000000000000000000..bb1754fa3749285c58841f4739186ee3b57a58a9 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.1-2025-11-13/8a7f4dd1-925a-494f-949e-29b101653793.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/openai_gpt-5.1-2025-11-13/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.1-2025-11-13", + "id": "openai/gpt-5.1-2025-11-13", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.1-2025-11-13", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.608, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.608", + "raw_stderr": "0.977", + "latency": "17.242", + "cost_per_test": "0.01105", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.977, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.861, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.861", + "raw_stderr": "1.075", + "latency": "19.818", + "cost_per_test": "0.017077", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.075, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.106, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.106", + "raw_stderr": "0.887", + "latency": "28.924", + "cost_per_test": "0.024697", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.887, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.756, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.756", + "raw_stderr": "1.498", + "latency": "21.483", + "cost_per_test": "0.018843", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.498, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.981, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.981", + "raw_stderr": "1.078", + "latency": "18.018", + "cost_per_test": "0.010654", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.078, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.495, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.495", + "raw_stderr": "1.278", + "latency": "60.928", + "cost_per_test": "0.058109", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.278, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.929, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.929", + "raw_stderr": "1.374", + "latency": "14.211", + "cost_per_test": "0.011522", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.374, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.69, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.69", + "raw_stderr": "2.133", + "latency": "22.006", + "cost_per_test": "0.009824", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.133, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.84, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.84", + "raw_stderr": "1.29", + "latency": "33.183", + "cost_per_test": "0.024716", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.29, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.301, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.301", + "raw_stderr": "0.635", + "latency": "20.251", + "cost_per_test": "0.019754", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.635, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.957, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.957", + "raw_stderr": "1.18", + "latency": "12.591", + "cost_per_test": "0.010452", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.18, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.377, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.377", + "raw_stderr": "0.337", + "latency": "23.056", + "cost_per_test": "0.018514", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.337, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.972, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.972", + "raw_stderr": "1.555", + "latency": "18.756", + "cost_per_test": "0.011306", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.555, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.685, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.685", + "raw_stderr": "0.806", + "latency": "23.472", + "cost_per_test": "0.022039", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.806, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.093, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.093", + "raw_stderr": "1.187", + "latency": "11.907", + "cost_per_test": "0.009152", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.187, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.1-2025-11-13/a8a688c3-a3d6-4155-b8e8-999e99800a4b.json b/data/vals_ai/openai/gpt-5.1-2025-11-13/a8a688c3-a3d6-4155-b8e8-999e99800a4b.json new file mode 100644 index 0000000000000000000000000000000000000000..aa2b2a1188155872be7fa85aba36323f17e1c2ad --- /dev/null +++ b/data/vals_ai/openai/gpt-5.1-2025-11-13/a8a688c3-a3d6-4155-b8e8-999e99800a4b.json @@ -0,0 +1,371 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/openai_gpt-5.1-2025-11-13/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.1-2025-11-13", + "id": "openai/gpt-5.1-2025-11-13", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.1-2025-11-13", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 35.714, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "35.714", + "raw_stderr": "7.483", + "latency": "736.071", + "cost_per_test": "1.151571", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.483, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 67.433, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "67.433", + "raw_stderr": "2.906", + "latency": "541.558", + "cost_per_test": "0.793283", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.906, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 81.443, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "81.443", + "raw_stderr": "2.798", + "latency": "392.096", + "cost_per_test": "0.506984", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.798, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "665.812", + "cost_per_test": "0.982603", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:overall:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 69.8, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "69.8", + "raw_stderr": "2.055", + "latency": "500.651", + "cost_per_test": "0.713431", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.055, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.1-2025-11-13/b459ce45-ffbd-4dbe-beb8-c6fe7222610c.json b/data/vals_ai/openai/gpt-5.1-2025-11-13/b459ce45-ffbd-4dbe-beb8-c6fe7222610c.json new file mode 100644 index 0000000000000000000000000000000000000000..e0e99ecc15f0775dc104f06a052cae7f40ddc0a3 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.1-2025-11-13/b459ce45-ffbd-4dbe-beb8-c6fe7222610c.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/openai_gpt-5.1-2025-11-13/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.1-2025-11-13", + "id": "openai/gpt-5.1-2025-11-13", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.1-2025-11-13", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 75.199, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "75.199", + "raw_stderr": "1.218", + "latency": "67.998", + "cost_per_test": "0.03919", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.218, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 61.368, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.368", + "raw_stderr": "0.931", + "latency": "46.254", + "cost_per_test": "0.026408", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.931, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 47.536, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "47.536", + "raw_stderr": "1.408", + "latency": "24.51", + "cost_per_test": "0.013626", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.408, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.1-2025-11-13/bb3b5b63-e342-4287-9ee2-9389934b90ce.json b/data/vals_ai/openai/gpt-5.1-2025-11-13/bb3b5b63-e342-4287-9ee2-9389934b90ce.json new file mode 100644 index 0000000000000000000000000000000000000000..495a122ea4b57fc2d3d1894e2698142f33d56abc --- /dev/null +++ b/data/vals_ai/openai/gpt-5.1-2025-11-13/bb3b5b63-e342-4287-9ee2-9389934b90ce.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/openai_gpt-5.1-2025-11-13/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.1-2025-11-13", + "id": "openai/gpt-5.1-2025-11-13", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.1-2025-11-13", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 24.606, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "24.606", + "raw_stderr": "4.254", + "latency": "1836.188", + "cost_per_test": "2.574926", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.254, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.1-2025-11-13/c25c6619-685d-457f-aba2-9690c906a639.json b/data/vals_ai/openai/gpt-5.1-2025-11-13/c25c6619-685d-457f-aba2-9690c906a639.json new file mode 100644 index 0000000000000000000000000000000000000000..e316ead6cad36760365349beefb4c12dd78fc770 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.1-2025-11-13/c25c6619-685d-457f-aba2-9690c906a639.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/openai_gpt-5.1-2025-11-13/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.1-2025-11-13", + "id": "openai/gpt-5.1-2025-11-13", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.1-2025-11-13", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 11.833, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "11.833", + "raw_stderr": "6.126", + "latency": "1702.974", + "cost_per_test": "0.363369", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.126, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 31.167, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "31.167", + "raw_stderr": "13.332", + "latency": "1797.164", + "cost_per_test": "0.373017", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 13.332, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "ioi:overall:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 21.5, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "21.5", + "raw_stderr": "7.336", + "latency": "1750.069", + "cost_per_test": "0.368193", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.336, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.1-2025-11-13/def7c1df-ec27-49e5-b95c-d59fac1986e6.json b/data/vals_ai/openai/gpt-5.1-2025-11-13/def7c1df-ec27-49e5-b95c-d59fac1986e6.json new file mode 100644 index 0000000000000000000000000000000000000000..d7b26c64fea467d13ad3e2ac2ac61c843e146e5b --- /dev/null +++ b/data/vals_ai/openai/gpt-5.1-2025-11-13/def7c1df-ec27-49e5-b95c-d59fac1986e6.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/openai_gpt-5.1-2025-11-13/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.1-2025-11-13", + "id": "openai/gpt-5.1-2025-11-13", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.1-2025-11-13", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 88.09, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "88.09", + "raw_stderr": "1.942", + "latency": "77.983", + "cost_per_test": "0.096508", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.942, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.1-2025-11-13/f27accde-bae2-4b2e-be97-eb9d3b36f7ca.json b/data/vals_ai/openai/gpt-5.1-2025-11-13/f27accde-bae2-4b2e-be97-eb9d3b36f7ca.json new file mode 100644 index 0000000000000000000000000000000000000000..cd98c387401083d6df2f6086de06852a27f373c9 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.1-2025-11-13/f27accde-bae2-4b2e-be97-eb9d3b36f7ca.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/openai_gpt-5.1-2025-11-13/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.1-2025-11-13", + "id": "openai/gpt-5.1-2025-11-13", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.1-2025-11-13", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.758, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.758", + "raw_stderr": "0.617", + "latency": "18.448", + "cost_per_test": "0.013965", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.617, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:hard:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 71.143, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "71.143", + "raw_stderr": "2.422", + "latency": "311.987", + "cost_per_test": "0.273252", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.422, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:medium:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 89.556, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "89.556", + "raw_stderr": "1.563", + "latency": "93.286", + "cost_per_test": "0.0899", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.563, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:overall:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 86.486, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "86.486", + "raw_stderr": "0.983", + "latency": "141.24", + "cost_per_test": "0.125706", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.983, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.1-2025-11-13/fe119f07-546a-477b-a7ef-8076e07e51af.json b/data/vals_ai/openai/gpt-5.1-2025-11-13/fe119f07-546a-477b-a7ef-8076e07e51af.json new file mode 100644 index 0000000000000000000000000000000000000000..1794aaa57a01006824e78481ad9dcb6dd59dc898 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.1-2025-11-13/fe119f07-546a-477b-a7ef-8076e07e51af.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/openai_gpt-5.1-2025-11-13/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.1-2025-11-13", + "id": "openai/gpt-5.1-2025-11-13", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.1-2025-11-13", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 64.841, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.841", + "raw_stderr": "1.365", + "latency": "44.176", + "cost_per_test": "0.007193", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.365, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 74.857, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "74.857", + "raw_stderr": "0.854", + "latency": "44.176", + "cost_per_test": "0.007193", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.854, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:openai/gpt-5.1-2025-11-13:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 84.873, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "84.873", + "raw_stderr": "1.025", + "latency": "44.176", + "cost_per_test": "0.007193", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.025, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.1-codex-max/73f0887d-d5a1-4483-9fab-1937c9fab6cc.json b/data/vals_ai/openai/gpt-5.1-codex-max/73f0887d-d5a1-4483-9fab-1937c9fab6cc.json new file mode 100644 index 0000000000000000000000000000000000000000..5695bb6b4fade3095c3bad04cfc3b6502c2b672c --- /dev/null +++ b/data/vals_ai/openai/gpt-5.1-codex-max/73f0887d-d5a1-4483-9fab-1937c9fab6cc.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/openai_gpt-5.1-codex-max/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.1-codex-max", + "id": "openai/gpt-5.1-codex-max", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.1-codex-max", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:openai/gpt-5.1-codex-max:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 9.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "9", + "raw_stderr": "2.862", + "latency": "3184.213", + "cost_per_test": "0.527356", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.862, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.1-codex-max/ac98b596-f1c0-4c17-a3e5-b9d4dba2d8c0.json b/data/vals_ai/openai/gpt-5.1-codex-max/ac98b596-f1c0-4c17-a3e5-b9d4dba2d8c0.json new file mode 100644 index 0000000000000000000000000000000000000000..f0b67596c455ad3bdb59d5d1237225df6992d1b1 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.1-codex-max/ac98b596-f1c0-4c17-a3e5-b9d4dba2d8c0.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/openai_gpt-5.1-codex-max/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.1-codex-max", + "id": "openai/gpt-5.1-codex-max", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.1-codex-max", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:openai/gpt-5.1-codex-max:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 97.516, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "97.516", + "raw_stderr": "0.867", + "latency": "9.963", + "cost_per_test": "0.001916", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.867, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:hard:openai/gpt-5.1-codex-max:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 65.429, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "65.429", + "raw_stderr": "2.542", + "latency": "151.6", + "cost_per_test": "-0.000378", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.542, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:medium:openai/gpt-5.1-codex-max:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 87.728, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "87.728", + "raw_stderr": "1.677", + "latency": "37.039", + "cost_per_test": "0.00313", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.677, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:overall:openai/gpt-5.1-codex-max:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 83.558, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "83.558", + "raw_stderr": "1.055", + "latency": "66.201", + "cost_per_test": "0.001556", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.055, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.1-codex-max/cbda25d8-3477-41bc-bfcc-c2d00e9594d6.json b/data/vals_ai/openai/gpt-5.1-codex-max/cbda25d8-3477-41bc-bfcc-c2d00e9594d6.json new file mode 100644 index 0000000000000000000000000000000000000000..48df55961d48e5a5cb6aa85285dd0942b1534f68 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.1-codex-max/cbda25d8-3477-41bc-bfcc-c2d00e9594d6.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/openai_gpt-5.1-codex-max/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.1-codex-max", + "id": "openai/gpt-5.1-codex-max", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.1-codex-max", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:openai/gpt-5.1-codex-max:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 14.5, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "14.5", + "raw_stderr": "9.757", + "latency": "975.205", + "cost_per_test": "0.49262", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 9.757, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:openai/gpt-5.1-codex-max:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 28.333, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "28.333", + "raw_stderr": "10.027", + "latency": "938.178", + "cost_per_test": "0.35267", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 10.027, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "ioi:overall:openai/gpt-5.1-codex-max:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 21.416, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "21.416", + "raw_stderr": "6.995", + "latency": "956.692", + "cost_per_test": "0.422645", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.995, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.1-codex-max/f2d61537-4139-4edd-b13a-1d1e785fe9b6.json b/data/vals_ai/openai/gpt-5.1-codex-max/f2d61537-4139-4edd-b13a-1d1e785fe9b6.json new file mode 100644 index 0000000000000000000000000000000000000000..af4007f830a90e1f8bac5bc991fa45bf05a33da1 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.1-codex-max/f2d61537-4139-4edd-b13a-1d1e785fe9b6.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/openai_gpt-5.1-codex-max/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.1-codex-max", + "id": "openai/gpt-5.1-codex-max", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.1-codex-max", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:openai/gpt-5.1-codex-max:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 22.168, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "22.168", + "raw_stderr": "3.84", + "latency": "31182.938", + "cost_per_test": "7.324105", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.84, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.1-codex/0ed119e5-b380-4b9c-9dac-a4caaaf70739.json b/data/vals_ai/openai/gpt-5.1-codex/0ed119e5-b380-4b9c-9dac-a4caaaf70739.json new file mode 100644 index 0000000000000000000000000000000000000000..4b3841cc9058a6616db7df5147cbcc83600cc94c --- /dev/null +++ b/data/vals_ai/openai/gpt-5.1-codex/0ed119e5-b380-4b9c-9dac-a4caaaf70739.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/openai_gpt-5.1-codex/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.1-codex", + "id": "openai/gpt-5.1-codex", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.1-codex", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:openai/gpt-5.1-codex:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.758, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.758", + "raw_stderr": "0.617", + "latency": "13.338", + "cost_per_test": "0.007355", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.617, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:hard:openai/gpt-5.1-codex:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 68.857, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "68.857", + "raw_stderr": "2.475", + "latency": "569.836", + "cost_per_test": "0.331205", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.475, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:medium:openai/gpt-5.1-codex:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 89.034, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "89.034", + "raw_stderr": "1.597", + "latency": "117.498", + "cost_per_test": "0.073947", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.597, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:overall:openai/gpt-5.1-codex:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 85.55, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "85.55", + "raw_stderr": "1.003", + "latency": "233.557", + "cost_per_test": "0.137502", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.003, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.1-codex/410a558c-77fc-47c4-bc21-2c5e87316aca.json b/data/vals_ai/openai/gpt-5.1-codex/410a558c-77fc-47c4-bc21-2c5e87316aca.json new file mode 100644 index 0000000000000000000000000000000000000000..b7aabb257404a49a459d26d2022cf7b78bf57c7e --- /dev/null +++ b/data/vals_ai/openai/gpt-5.1-codex/410a558c-77fc-47c4-bc21-2c5e87316aca.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/openai_gpt-5.1-codex/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.1-codex", + "id": "openai/gpt-5.1-codex", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.1-codex", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:openai/gpt-5.1-codex:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "2326.487", + "cost_per_test": "0.201958", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:openai/gpt-5.1-codex:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 7.333, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "7.333", + "raw_stderr": "6.694", + "latency": "2128.396", + "cost_per_test": "0.151756", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.694, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "ioi:overall:openai/gpt-5.1-codex:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 3.666, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "3.666", + "raw_stderr": "3.347", + "latency": "2227.442", + "cost_per_test": "0.176857", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.347, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.1-codex/c8ddd919-4a18-4dde-a289-55184ae9bc53.json b/data/vals_ai/openai/gpt-5.1-codex/c8ddd919-4a18-4dde-a289-55184ae9bc53.json new file mode 100644 index 0000000000000000000000000000000000000000..1c3a05d1a482cf03854869fecb0b933f8a978b97 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.1-codex/c8ddd919-4a18-4dde-a289-55184ae9bc53.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/openai_gpt-5.1-codex/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.1-codex", + "id": "openai/gpt-5.1-codex", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.1-codex", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:openai/gpt-5.1-codex:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 13.115, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "13.115", + "raw_stderr": "3.161", + "latency": "3026.069", + "cost_per_test": "3.797914", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.161, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.2-2025-12-11/1ba7d5bc-e01e-4647-af0a-cd2ddd386f30.json b/data/vals_ai/openai/gpt-5.2-2025-12-11/1ba7d5bc-e01e-4647-af0a-cd2ddd386f30.json new file mode 100644 index 0000000000000000000000000000000000000000..5da8a7bbec0bc50b2bc9a4326418dbe6fabb7229 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.2-2025-12-11/1ba7d5bc-e01e-4647-af0a-cd2ddd386f30.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/openai_gpt-5.2-2025-12-11/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.2-2025-12-11", + "id": "openai/gpt-5.2-2025-12-11", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.2-2025-12-11", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 53.499, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "53.499", + "raw_stderr": "5.067", + "latency": "4971.341", + "cost_per_test": "17.745182", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.067, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.2-2025-12-11/365a6bfe-aa69-470d-b202-cb8efdb5f594.json b/data/vals_ai/openai/gpt-5.2-2025-12-11/365a6bfe-aa69-470d-b202-cb8efdb5f594.json new file mode 100644 index 0000000000000000000000000000000000000000..f19ff633fb389cb545db01f081dac5f35377caa6 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.2-2025-12-11/365a6bfe-aa69-470d-b202-cb8efdb5f594.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/openai_gpt-5.2-2025-12-11/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.2-2025-12-11", + "id": "openai/gpt-5.2-2025-12-11", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.2-2025-12-11", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.758, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.758", + "raw_stderr": "0.617", + "latency": "14.231", + "cost_per_test": "0.012439", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.617, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "lcb:hard:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 70.118, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "70.118", + "raw_stderr": "2.49", + "latency": "354.881", + "cost_per_test": "0.30169", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.49, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "lcb:medium:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 87.206, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "87.206", + "raw_stderr": "1.707", + "latency": "85.426", + "cost_per_test": "0.091357", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.707, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "lcb:overall:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 85.361, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "85.361", + "raw_stderr": "1.027", + "latency": "151.513", + "cost_per_test": "0.135162", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.027, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.2-2025-12-11/3fcb2852-b1a8-4b10-9946-53d3d98a6e48.json b/data/vals_ai/openai/gpt-5.2-2025-12-11/3fcb2852-b1a8-4b10-9946-53d3d98a6e48.json new file mode 100644 index 0000000000000000000000000000000000000000..fdbb8dcdf21cdc5df6272bc10f019345b2acd71c --- /dev/null +++ b/data/vals_ai/openai/gpt-5.2-2025-12-11/3fcb2852-b1a8-4b10-9946-53d3d98a6e48.json @@ -0,0 +1,371 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/openai_gpt-5.2-2025-12-11/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.2-2025-12-11", + "id": "openai/gpt-5.2-2025-12-11", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.2-2025-12-11", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 38.095, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "38.095", + "raw_stderr": "7.584", + "latency": "1104.392", + "cost_per_test": "2.845221", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.584, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 72.414, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "72.414", + "raw_stderr": "2.772", + "latency": "701.396", + "cost_per_test": "1.569703", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.772, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 89.175, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "89.175", + "raw_stderr": "2.236", + "latency": "454.049", + "cost_per_test": "0.979118", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.236, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "900.573", + "cost_per_test": "2.515054", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "swebench:overall:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 75.8, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "75.8", + "raw_stderr": "1.917", + "latency": "640.472", + "cost_per_test": "1.453372", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.917, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.2-2025-12-11/4cfadee6-7b64-4d89-90f9-da1153726c57.json b/data/vals_ai/openai/gpt-5.2-2025-12-11/4cfadee6-7b64-4d89-90f9-da1153726c57.json new file mode 100644 index 0000000000000000000000000000000000000000..b34fad47774c66de8fbbd3a15434c6e5640c8bc1 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.2-2025-12-11/4cfadee6-7b64-4d89-90f9-da1153726c57.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/openai_gpt-5.2-2025-12-11/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.2-2025-12-11", + "id": "openai/gpt-5.2-2025-12-11", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.2-2025-12-11", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 68.065, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.065", + "raw_stderr": "1.592", + "latency": "33.05", + "cost_per_test": "0.335055", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.592, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 63.869, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.869", + "raw_stderr": "1.64", + "latency": "14.006", + "cost_per_test": "0.016663", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.64, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 65.889, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.889", + "raw_stderr": "0.934", + "latency": "26.038", + "cost_per_test": "0.153756", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.934, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 65.734, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.734", + "raw_stderr": "1.62", + "latency": "31.057", + "cost_per_test": "0.109551", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.62, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.2-2025-12-11/5ae1e1eb-71f0-42cc-af76-e707394b1b93.json b/data/vals_ai/openai/gpt-5.2-2025-12-11/5ae1e1eb-71f0-42cc-af76-e707394b1b93.json new file mode 100644 index 0000000000000000000000000000000000000000..e885f25b1ae21c71f75e8d361e3894ada85b527c --- /dev/null +++ b/data/vals_ai/openai/gpt-5.2-2025-12-11/5ae1e1eb-71f0-42cc-af76-e707394b1b93.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/openai_gpt-5.2-2025-12-11/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.2-2025-12-11", + "id": "openai/gpt-5.2-2025-12-11", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.2-2025-12-11", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 66.312, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.312", + "raw_stderr": "1.352", + "latency": "60.479", + "cost_per_test": "0.066435", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.352, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 75.756, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "75.756", + "raw_stderr": "0.845", + "latency": "60.479", + "cost_per_test": "0.066435", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.845, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 85.2, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "85.2", + "raw_stderr": "1.015", + "latency": "60.479", + "cost_per_test": "0.066435", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.015, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.2-2025-12-11/71cab120-ac4f-4bb4-b769-282a1a0e48ab.json b/data/vals_ai/openai/gpt-5.2-2025-12-11/71cab120-ac4f-4bb4-b769-282a1a0e48ab.json new file mode 100644 index 0000000000000000000000000000000000000000..bdb1024692847bf426005b6c5258cb65df9fb55b --- /dev/null +++ b/data/vals_ai/openai/gpt-5.2-2025-12-11/71cab120-ac4f-4bb4-b769-282a1a0e48ab.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/openai_gpt-5.2-2025-12-11/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.2-2025-12-11", + "id": "openai/gpt-5.2-2025-12-11", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.2-2025-12-11", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 15.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "15", + "raw_stderr": "3.571", + "latency": "1493.817", + "cost_per_test": "1.059689", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.571, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.2-2025-12-11/809471ec-fe99-402a-ac02-bc24546c7d37.json b/data/vals_ai/openai/gpt-5.2-2025-12-11/809471ec-fe99-402a-ac02-bc24546c7d37.json new file mode 100644 index 0000000000000000000000000000000000000000..71673d9788386b11f6710c93bf391f380f1d74a1 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.2-2025-12-11/809471ec-fe99-402a-ac02-bc24546c7d37.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/openai_gpt-5.2-2025-12-11/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.2-2025-12-11", + "id": "openai/gpt-5.2-2025-12-11", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.2-2025-12-11", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 43.833, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "43.833", + "raw_stderr": "12.379", + "latency": "7378.658", + "cost_per_test": "4.02342", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 12.379, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 65.833, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "65.833", + "raw_stderr": "11.114", + "latency": "6100.567", + "cost_per_test": "3.720564", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 11.114, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "ioi:overall:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 54.833, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "54.833", + "raw_stderr": "8.318", + "latency": "6739.612", + "cost_per_test": "3.871992", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 8.318, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.2-2025-12-11/8458aa0b-62a6-48a1-9ea1-efe62b7fd957.json b/data/vals_ai/openai/gpt-5.2-2025-12-11/8458aa0b-62a6-48a1-9ea1-efe62b7fd957.json new file mode 100644 index 0000000000000000000000000000000000000000..c6dd343d207bed01023a01027ac693c1851becd6 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.2-2025-12-11/8458aa0b-62a6-48a1-9ea1-efe62b7fd957.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/openai_gpt-5.2-2025-12-11/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.2-2025-12-11", + "id": "openai/gpt-5.2-2025-12-11", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.2-2025-12-11", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 91.919, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.919", + "raw_stderr": "2.513", + "latency": "170.268", + "cost_per_test": "0.149071", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.513, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 91.414, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.414", + "raw_stderr": "2.68", + "latency": "176.161", + "cost_per_test": "0.149003", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.68, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 91.666, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.666", + "raw_stderr": "1.837", + "latency": "173.214", + "cost_per_test": "0.149037", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.837, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.2-2025-12-11/8fc65c6b-1431-47e2-a37a-30852cf1d0f8.json b/data/vals_ai/openai/gpt-5.2-2025-12-11/8fc65c6b-1431-47e2-a37a-30852cf1d0f8.json new file mode 100644 index 0000000000000000000000000000000000000000..e1bfe030d7180fed1045c44a2d08b002c69bf7c9 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.2-2025-12-11/8fc65c6b-1431-47e2-a37a-30852cf1d0f8.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/openai_gpt-5.2-2025-12-11/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.2-2025-12-11", + "id": "openai/gpt-5.2-2025-12-11", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.2-2025-12-11", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 84.387, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "84.387", + "raw_stderr": "1.856", + "latency": "126.205", + "cost_per_test": "0.115422", + "max_output_tokens": "30000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.856, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.2-2025-12-11/92bd9305-854b-46cb-8193-6e69ef938b60.json b/data/vals_ai/openai/gpt-5.2-2025-12-11/92bd9305-854b-46cb-8193-6e69ef938b60.json new file mode 100644 index 0000000000000000000000000000000000000000..6e7df4aed5eb519517a133d90c8066c6f8c8fc14 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.2-2025-12-11/92bd9305-854b-46cb-8193-6e69ef938b60.json @@ -0,0 +1,356 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/openai_gpt-5.2-2025-12-11/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.2-2025-12-11", + "id": "openai/gpt-5.2-2025-12-11", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.2-2025-12-11", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 46.032, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "46.032", + "raw_stderr": "6.255", + "latency": "191.629", + "cost_per_test": "0.124815", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.255, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "sage:automata:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 32.143, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "32.143", + "raw_stderr": "5.952", + "latency": "217.3", + "cost_per_test": "0.134564", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.952, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "sage:calculus:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 62.037, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "62.037", + "raw_stderr": "6.102", + "latency": "174.449", + "cost_per_test": "0.137909", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.102, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 69.636, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "69.636", + "raw_stderr": "5.165", + "latency": "127.733", + "cost_per_test": "0.088042", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.165, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "sage:overall:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 49.27, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "49.27", + "raw_stderr": "3.354", + "latency": "178.887", + "cost_per_test": "0.115807", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.354, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.2-2025-12-11/a5e7f4ef-da7f-482c-a9e6-ea1b1130e980.json b/data/vals_ai/openai/gpt-5.2-2025-12-11/a5e7f4ef-da7f-482c-a9e6-ea1b1130e980.json new file mode 100644 index 0000000000000000000000000000000000000000..87aa3f573fd13f0a80f2afd65edbca16257cf0c8 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.2-2025-12-11/a5e7f4ef-da7f-482c-a9e6-ea1b1130e980.json @@ -0,0 +1,103 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/openai_gpt-5.2-2025-12-11/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.2-2025-12-11", + "id": "openai/gpt-5.2-2025-12-11", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.2-2025-12-11", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 86.667, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.667", + "raw_stderr": "0.822", + "latency": "96.324", + "cost_per_test": "0.061121", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.822, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.2-2025-12-11/bdeb5778-1d8d-4a72-8294-352117a3dfa5.json b/data/vals_ai/openai/gpt-5.2-2025-12-11/bdeb5778-1d8d-4a72-8294-352117a3dfa5.json new file mode 100644 index 0000000000000000000000000000000000000000..4c0459d7f7a35dea4cfd04c041aa46d8ef32057f --- /dev/null +++ b/data/vals_ai/openai/gpt-5.2-2025-12-11/bdeb5778-1d8d-4a72-8294-352117a3dfa5.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/openai_gpt-5.2-2025-12-11/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.2-2025-12-11", + "id": "openai/gpt-5.2-2025-12-11", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.2-2025-12-11", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.608, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.608", + "raw_stderr": "1.018", + "latency": "16.66", + "cost_per_test": "0.016803", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.018, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.635, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.635", + "raw_stderr": "1.012", + "latency": "34.755", + "cost_per_test": "0.038321", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.012, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.724, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.724", + "raw_stderr": "0.873", + "latency": "35.206", + "cost_per_test": "0.041726", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.873, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.293, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.293", + "raw_stderr": "1.588", + "latency": "33.394", + "cost_per_test": "0.034376", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.588, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.573, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.573", + "raw_stderr": "1.062", + "latency": "20.197", + "cost_per_test": "0.020471", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.062, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.972, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.972", + "raw_stderr": "1.21", + "latency": "116.232", + "cost_per_test": "0.1194", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.21, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.174, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.174", + "raw_stderr": "1.367", + "latency": "21.417", + "cost_per_test": "0.022852", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.367, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.378, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.378", + "raw_stderr": "2.201", + "latency": "21.72", + "cost_per_test": "0.020553", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.201, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.479, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.479", + "raw_stderr": "1.35", + "latency": "72.826", + "cost_per_test": "0.067295", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.35, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 95.041, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "95.041", + "raw_stderr": "0.599", + "latency": "26.339", + "cost_per_test": "0.031514", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.599, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.359, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.359", + "raw_stderr": "1.278", + "latency": "23.394", + "cost_per_test": "0.024895", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.278, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.234, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.234", + "raw_stderr": "0.344", + "latency": "35.413", + "cost_per_test": "0.03698", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.344, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.571, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.571", + "raw_stderr": "1.609", + "latency": "24.271", + "cost_per_test": "0.024711", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.609, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.378, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.378", + "raw_stderr": "0.779", + "latency": "34.187", + "cost_per_test": "0.039451", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.779, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.09, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.09", + "raw_stderr": "1.248", + "latency": "15.183", + "cost_per_test": "0.015346", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.248, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.2-2025-12-11/e199c428-83ec-4dbf-997c-c5b16006cbf1.json b/data/vals_ai/openai/gpt-5.2-2025-12-11/e199c428-83ec-4dbf-997c-c5b16006cbf1.json new file mode 100644 index 0000000000000000000000000000000000000000..959156d981f31213a2042a3e43f8ec5ec81e1a2f --- /dev/null +++ b/data/vals_ai/openai/gpt-5.2-2025-12-11/e199c428-83ec-4dbf-997c-c5b16006cbf1.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/openai_gpt-5.2-2025-12-11/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.2-2025-12-11", + "id": "openai/gpt-5.2-2025-12-11", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.2-2025-12-11", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 76.55, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "76.55", + "raw_stderr": "1.195", + "latency": "105.239", + "cost_per_test": "0.070212", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.195, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 67.13, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.13", + "raw_stderr": "0.918", + "latency": "70.666", + "cost_per_test": "0.050478", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.918, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 57.711, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.711", + "raw_stderr": "1.393", + "latency": "36.093", + "cost_per_test": "0.030743", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.393, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.2-2025-12-11/f75dbbb4-c848-429a-b5c8-2000af983e34.json b/data/vals_ai/openai/gpt-5.2-2025-12-11/f75dbbb4-c848-429a-b5c8-2000af983e34.json new file mode 100644 index 0000000000000000000000000000000000000000..494ac0ebe505a6788047193deb283f58efe45326 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.2-2025-12-11/f75dbbb4-c848-429a-b5c8-2000af983e34.json @@ -0,0 +1,438 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/openai_gpt-5.2-2025-12-11/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.2-2025-12-11", + "id": "openai/gpt-5.2-2025-12-11", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.2-2025-12-11", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 90.168, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.168", + "raw_stderr": "0.608", + "latency": "15.108", + "cost_per_test": "0.013817", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.608, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.756, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.756", + "raw_stderr": "0.194", + "latency": "12.853", + "cost_per_test": "0.013416", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.194, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 88.305, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.305", + "raw_stderr": "0.591", + "latency": "6.548", + "cost_per_test": "0.007325", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.591, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.764, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.764", + "raw_stderr": "0.395", + "latency": "17.138", + "cost_per_test": "0.017566", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.395, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.872, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.872", + "raw_stderr": "0.916", + "latency": "4.813", + "cost_per_test": "0.005559", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.916, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:openai/gpt-5.2-2025-12-11:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 69.717, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "69.717", + "raw_stderr": "1.518", + "latency": "46.366", + "cost_per_test": "0.047712", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.518, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.2-codex/34dc6565-6d38-4462-b2da-78aa261eb44b.json b/data/vals_ai/openai/gpt-5.2-codex/34dc6565-6d38-4462-b2da-78aa261eb44b.json new file mode 100644 index 0000000000000000000000000000000000000000..906b35f1eaef49ed54e4f80f75ecc81fb766d326 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.2-codex/34dc6565-6d38-4462-b2da-78aa261eb44b.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/openai_gpt-5.2-codex/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.2-codex", + "id": "openai/gpt-5.2-codex", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.2-codex", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:openai/gpt-5.2-codex:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 37.912, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "37.912", + "raw_stderr": "4.576", + "latency": "1934.792", + "cost_per_test": "4.152495", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.576, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.2-codex/72fdceba-93ee-41c0-9548-94ab4750dc17.json b/data/vals_ai/openai/gpt-5.2-codex/72fdceba-93ee-41c0-9548-94ab4750dc17.json new file mode 100644 index 0000000000000000000000000000000000000000..afc32ce3470a9ce5ec00821daa909bc456004873 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.2-codex/72fdceba-93ee-41c0-9548-94ab4750dc17.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/openai_gpt-5.2-codex/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.2-codex", + "id": "openai/gpt-5.2-codex", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.2-codex", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:openai/gpt-5.2-codex:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 99.068, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "99.068", + "raw_stderr": "0.535", + "latency": "10.612", + "cost_per_test": "0.007545", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:hard:openai/gpt-5.2-codex:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 74.571, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "74.571", + "raw_stderr": "2.328", + "latency": "409.148", + "cost_per_test": "0.279114", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.328, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:medium:openai/gpt-5.2-codex:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 90.339, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "90.339", + "raw_stderr": "1.51", + "latency": "85.052", + "cost_per_test": "0.063351", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.51, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:overall:openai/gpt-5.2-codex:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 87.993, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "87.993", + "raw_stderr": "0.942", + "latency": "168.271", + "cost_per_test": "0.11667", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.942, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.2-codex/79b0d518-44dd-4b13-8216-1ea0fa47477f.json b/data/vals_ai/openai/gpt-5.2-codex/79b0d518-44dd-4b13-8216-1ea0fa47477f.json new file mode 100644 index 0000000000000000000000000000000000000000..36467fc34881618188b110558bf1fc454ac54e74 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.2-codex/79b0d518-44dd-4b13-8216-1ea0fa47477f.json @@ -0,0 +1,371 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/openai_gpt-5.2-codex/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.2-codex", + "id": "openai/gpt-5.2-codex", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.2-codex", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:openai/gpt-5.2-codex:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 28.571, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "28.571", + "raw_stderr": "7.055", + "latency": "362.318", + "cost_per_test": "0.668577", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.055, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:openai/gpt-5.2-codex:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 68.199, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "68.199", + "raw_stderr": "2.888", + "latency": "206.383", + "cost_per_test": "0.33597", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.888, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:openai/gpt-5.2-codex:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 88.144, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "88.144", + "raw_stderr": "2.327", + "latency": "125.331", + "cost_per_test": "0.198269", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.327, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:openai/gpt-5.2-codex:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "268.266", + "cost_per_test": "0.510369", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:overall:openai/gpt-5.2-codex:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 72.4, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "72.4", + "raw_stderr": "2.001", + "latency": "188.405", + "cost_per_test": "0.311527", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.001, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.3-codex/09a897be-fba5-4a94-9efe-47ab99f7589e.json b/data/vals_ai/openai/gpt-5.3-codex/09a897be-fba5-4a94-9efe-47ab99f7589e.json new file mode 100644 index 0000000000000000000000000000000000000000..b35fbc36054ce944effe1d6b1b4eadac97ed3510 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.3-codex/09a897be-fba5-4a94-9efe-47ab99f7589e.json @@ -0,0 +1,371 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/openai_gpt-5.3-codex/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.3-codex", + "id": "openai/gpt-5.3-codex", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.3-codex", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:openai/gpt-5.3-codex:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 54.762, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "54.762", + "raw_stderr": "7.773", + "latency": "405.865", + "cost_per_test": "0.818585", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.773, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:openai/gpt-5.3-codex:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 73.18, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "73.18", + "raw_stderr": "2.748", + "latency": "264.025", + "cost_per_test": "0.479692", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.748, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:openai/gpt-5.3-codex:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 90.206, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "90.206", + "raw_stderr": "2.14", + "latency": "185.93", + "cost_per_test": "0.340336", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.14, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:openai/gpt-5.3-codex:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "412.097", + "cost_per_test": "0.69549", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "swebench:overall:openai/gpt-5.3-codex:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 78.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "78", + "raw_stderr": "1.854", + "latency": "246.527", + "cost_per_test": "0.455384", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.854, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.3-codex/655ccb47-b729-4b00-807e-b11dfaad7d47.json b/data/vals_ai/openai/gpt-5.3-codex/655ccb47-b729-4b00-807e-b11dfaad7d47.json new file mode 100644 index 0000000000000000000000000000000000000000..dc8d1176860112f418902118ae0fd42436702a08 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.3-codex/655ccb47-b729-4b00-807e-b11dfaad7d47.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/openai_gpt-5.3-codex/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.3-codex", + "id": "openai/gpt-5.3-codex", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.3-codex", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:openai/gpt-5.3-codex:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 99.068, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "99.068", + "raw_stderr": "0.535", + "latency": "17.469", + "cost_per_test": "0.013703", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "lcb:hard:openai/gpt-5.3-codex:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 75.143, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "75.143", + "raw_stderr": "2.31", + "latency": "306.379", + "cost_per_test": "0.209533", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.31, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "lcb:medium:openai/gpt-5.3-codex:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 87.728, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "87.728", + "raw_stderr": "1.677", + "latency": "74.045", + "cost_per_test": "0.051518", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.677, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "lcb:overall:openai/gpt-5.3-codex:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 87.313, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "87.313", + "raw_stderr": "0.968", + "latency": "132.631", + "cost_per_test": "0.091585", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.968, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.3-codex/73c2a7b5-8a57-485b-8997-ddd42a8ef562.json b/data/vals_ai/openai/gpt-5.3-codex/73c2a7b5-8a57-485b-8997-ddd42a8ef562.json new file mode 100644 index 0000000000000000000000000000000000000000..57ca658e902f13eee1410f33c51b40ffd3cab8b1 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.3-codex/73c2a7b5-8a57-485b-8997-ddd42a8ef562.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/openai_gpt-5.3-codex/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.3-codex", + "id": "openai/gpt-5.3-codex", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.3-codex", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:openai/gpt-5.3-codex:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 61.767, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "61.767", + "raw_stderr": "4.714", + "latency": "4548.115", + "cost_per_test": "11.914536", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.714, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.3-codex/c28b447f-e02e-44c6-9a4b-4c8b4444019a.json b/data/vals_ai/openai/gpt-5.3-codex/c28b447f-e02e-44c6-9a4b-4c8b4444019a.json new file mode 100644 index 0000000000000000000000000000000000000000..1d21afbc4c3945d0109f8fd660b9d03a896432a5 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.3-codex/c28b447f-e02e-44c6-9a4b-4c8b4444019a.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/openai_gpt-5.3-codex/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.3-codex", + "id": "openai/gpt-5.3-codex", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.3-codex", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:openai/gpt-5.3-codex:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 58.167, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "58.167", + "raw_stderr": "14.481", + "latency": "2631.97", + "cost_per_test": "1.818037", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 14.481, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:openai/gpt-5.3-codex:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 29.5, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "29.5", + "raw_stderr": "5.854", + "latency": "4516.544", + "cost_per_test": "2.222013", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.854, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "ioi:overall:openai/gpt-5.3-codex:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 43.834, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "43.834", + "raw_stderr": "7.81", + "latency": "3574.257", + "cost_per_test": "2.020025", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.81, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-2026-03-05-high/2ee1ada0-95c9-4527-8a03-ba03df4acdae.json b/data/vals_ai/openai/gpt-5.4-2026-03-05-high/2ee1ada0-95c9-4527-8a03-ba03df4acdae.json new file mode 100644 index 0000000000000000000000000000000000000000..e80950a26ca33f3e8c4dc028a5c9f692fbf6474a --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-2026-03-05-high/2ee1ada0-95c9-4527-8a03-ba03df4acdae.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/programbench/openai_gpt-5.4-2026-03-05-high/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProgramBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-2026-03-05-high", + "id": "openai/gpt-5.4-2026-03-05-high", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-2026-03-05-high", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "programbench:almost:openai/gpt-5.4-2026-03-05-high:score", + "evaluation_name": "vals_ai.programbench.almost", + "source_data": { + "dataset_name": "ProgramBench - Almost Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "almost", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Almost Resolved).", + "metric_id": "vals_ai.programbench.almost.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 3.5, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "almost", + "task_name": "Almost Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "3.5", + "raw_stderr": "1.303", + "latency": "958.495", + "cost_per_test": "2.27429", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.303, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "programbench:overall:openai/gpt-5.4-2026-03-05-high:score", + "evaluation_name": "vals_ai.programbench.overall", + "source_data": { + "dataset_name": "ProgramBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Overall).", + "metric_id": "vals_ai.programbench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.5, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0.5", + "raw_stderr": "0.5", + "latency": "958.495", + "cost_per_test": "2.27429", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.5, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "programbench:partial:openai/gpt-5.4-2026-03-05-high:score", + "evaluation_name": "vals_ai.programbench.partial", + "source_data": { + "dataset_name": "ProgramBench - Raw Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "partial", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Raw Pass Rate).", + "metric_id": "vals_ai.programbench.partial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 61.764, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "partial", + "task_name": "Raw Pass Rate", + "dataset_type": "public", + "industry": "coding", + "raw_score": "61.764", + "raw_stderr": "1.791", + "latency": "958.495", + "cost_per_test": "2.27429", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.791, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "programbench:strict:openai/gpt-5.4-2026-03-05-high:score", + "evaluation_name": "vals_ai.programbench.strict", + "source_data": { + "dataset_name": "ProgramBench - Fully Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "strict", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Fully Resolved).", + "metric_id": "vals_ai.programbench.strict.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.5, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "strict", + "task_name": "Fully Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0.5", + "raw_stderr": "0.5", + "latency": "958.495", + "cost_per_test": "2.27429", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.5, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-2026-03-05/09b459e2-f823-4bff-8fa8-26a4278d4dce.json b/data/vals_ai/openai/gpt-5.4-2026-03-05/09b459e2-f823-4bff-8fa8-26a4278d4dce.json new file mode 100644 index 0000000000000000000000000000000000000000..134d65c551dea825fde86ad62431b54b97450b8e --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-2026-03-05/09b459e2-f823-4bff-8fa8-26a4278d4dce.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/openai_gpt-5.4-2026-03-05/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-2026-03-05", + "id": "openai/gpt-5.4-2026-03-05", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-2026-03-05", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 67.421, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "67.421", + "raw_stderr": "4.843", + "latency": "5174.848", + "cost_per_test": "10.686976", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.843, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-2026-03-05/18fb8969-6bb4-46e1-8122-f01daee00628.json b/data/vals_ai/openai/gpt-5.4-2026-03-05/18fb8969-6bb4-46e1-8122-f01daee00628.json new file mode 100644 index 0000000000000000000000000000000000000000..81b19651c3962a0179fbb4545997f6d9aac45c81 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-2026-03-05/18fb8969-6bb4-46e1-8122-f01daee00628.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/openai_gpt-5.4-2026-03-05/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-2026-03-05", + "id": "openai/gpt-5.4-2026-03-05", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-2026-03-05", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 64.103, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.103", + "raw_stderr": "1.638", + "latency": "51.952", + "cost_per_test": "0.025446", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.638, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 64.918, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.918", + "raw_stderr": "1.629", + "latency": "72.418", + "cost_per_test": "0.483226", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.629, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 65.268, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.268", + "raw_stderr": "0.938", + "latency": "82.34", + "cost_per_test": "0.235266", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.938, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 66.783, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.783", + "raw_stderr": "1.608", + "latency": "122.651", + "cost_per_test": "0.197126", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.608, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-2026-03-05/226f2c70-31cf-4aa8-bc46-38b6ad9625b9.json b/data/vals_ai/openai/gpt-5.4-2026-03-05/226f2c70-31cf-4aa8-bc46-38b6ad9625b9.json new file mode 100644 index 0000000000000000000000000000000000000000..71a9385a25383ea0a9b7abe13de1640e451b1dc4 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-2026-03-05/226f2c70-31cf-4aa8-bc46-38b6ad9625b9.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/openai_gpt-5.4-2026-03-05/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-2026-03-05", + "id": "openai/gpt-5.4-2026-03-05", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-2026-03-05", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.887, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.887", + "raw_stderr": "1.53", + "latency": "14.15", + "cost_per_test": "0.027279", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.53, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.762, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.762", + "raw_stderr": "1.441", + "latency": "22.267", + "cost_per_test": "0.045262", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.441, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.078, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.078", + "raw_stderr": "1.082", + "latency": "21.94", + "cost_per_test": "0.048378", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.082, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.756, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.756", + "raw_stderr": "2.274", + "latency": "22.64", + "cost_per_test": "0.045736", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.274, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.692, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.692", + "raw_stderr": "1.356", + "latency": "15.82", + "cost_per_test": "0.030999", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.356, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.243, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.243", + "raw_stderr": "1.381", + "latency": "55.333", + "cost_per_test": "0.123866", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.381, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.396, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.396", + "raw_stderr": "1.598", + "latency": "38.712", + "cost_per_test": "0.045974", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.598, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.265, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.265", + "raw_stderr": "2.4", + "latency": "20.393", + "cost_per_test": "0.040285", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.4, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.477, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.477", + "raw_stderr": "1.456", + "latency": "45.188", + "cost_per_test": "0.093825", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.456, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.671, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.671", + "raw_stderr": "0.805", + "latency": "16.611", + "cost_per_test": "0.037007", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.805, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.472, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.472", + "raw_stderr": "1.564", + "latency": "26.512", + "cost_per_test": "0.051891", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.564, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.482, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.482", + "raw_stderr": "0.421", + "latency": "28.875", + "cost_per_test": "0.050888", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.421, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.974, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.974", + "raw_stderr": "1.908", + "latency": "50.281", + "cost_per_test": "0.047208", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.908, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.224, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.224", + "raw_stderr": "0.931", + "latency": "19.87", + "cost_per_test": "0.041984", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.931, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.845, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.845", + "raw_stderr": "1.474", + "latency": "34.537", + "cost_per_test": "0.032734", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.474, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-2026-03-05/29b8caf1-2f5b-475e-804b-ac3d69b64f1f.json b/data/vals_ai/openai/gpt-5.4-2026-03-05/29b8caf1-2f5b-475e-804b-ac3d69b64f1f.json new file mode 100644 index 0000000000000000000000000000000000000000..3b209b4ef985ee80a7106d28007f655cfe4e87f0 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-2026-03-05/29b8caf1-2f5b-475e-804b-ac3d69b64f1f.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/openai_gpt-5.4-2026-03-05/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-2026-03-05", + "id": "openai/gpt-5.4-2026-03-05", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-2026-03-05", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 67.167, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "67.167", + "raw_stderr": "14.878", + "latency": "2241.026", + "cost_per_test": "3.371664", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 14.878, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 68.5, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "68.5", + "raw_stderr": "12.965", + "latency": "3080.746", + "cost_per_test": "3.571001", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 12.965, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "ioi:overall:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 67.834, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "67.834", + "raw_stderr": "9.867", + "latency": "2660.886", + "cost_per_test": "3.471332", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 9.867, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-2026-03-05/489da6c0-cf4d-4b04-bdc5-88dc04c25240.json b/data/vals_ai/openai/gpt-5.4-2026-03-05/489da6c0-cf4d-4b04-bdc5-88dc04c25240.json new file mode 100644 index 0000000000000000000000000000000000000000..4297fca5c1f5b98507c13c91d8bad12b53a02e42 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-2026-03-05/489da6c0-cf4d-4b04-bdc5-88dc04c25240.json @@ -0,0 +1,195 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/cyber/openai_gpt-5.4-2026-03-05/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CyberBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "beta", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-2026-03-05", + "id": "openai/gpt-5.4-2026-03-05", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-2026-03-05", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "cyber:overall:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.cyber.overall", + "source_data": { + "dataset_name": "CyberBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (Overall).", + "metric_id": "vals_ai.cyber.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 72.881, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "beta", + "raw_score": "72.881", + "raw_stderr": "5.652", + "latency": "1109.034", + "cost_per_test": "2.945283", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.652, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "cyber:patch:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.cyber.patch", + "source_data": { + "dataset_name": "CyberBench - Patch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "patch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (Patch).", + "metric_id": "vals_ai.cyber.patch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 81.356, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "patch", + "task_name": "Patch", + "dataset_type": "private", + "industry": "beta", + "raw_score": "81.356", + "raw_stderr": "5.07", + "latency": "1024.183", + "cost_per_test": "0.836845", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.07, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "cyber:poc:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.cyber.poc", + "source_data": { + "dataset_name": "CyberBench - PoC", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "poc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (PoC).", + "metric_id": "vals_ai.cyber.poc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 64.407, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "poc", + "task_name": "PoC", + "dataset_type": "private", + "industry": "beta", + "raw_score": "64.407", + "raw_stderr": "6.233", + "latency": "1193.886", + "cost_per_test": "5.053722", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.233, + "method": "vals_reported" + } + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-2026-03-05/5093bd20-6564-44f0-be43-3725bdcbbb48.json b/data/vals_ai/openai/gpt-5.4-2026-03-05/5093bd20-6564-44f0-be43-3725bdcbbb48.json new file mode 100644 index 0000000000000000000000000000000000000000..a599a04e5d68e3063db6367acb116121ab8a82e5 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-2026-03-05/5093bd20-6564-44f0-be43-3725bdcbbb48.json @@ -0,0 +1,356 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/openai_gpt-5.4-2026-03-05/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-2026-03-05", + "id": "openai/gpt-5.4-2026-03-05", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-2026-03-05", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 39.683, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "39.683", + "raw_stderr": "6.148", + "latency": "166.337", + "cost_per_test": "0.10951", + "max_output_tokens": "30000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.148, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "sage:automata:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 21.428, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "21.428", + "raw_stderr": "4.715", + "latency": "189.656", + "cost_per_test": "0.112253", + "max_output_tokens": "30000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.715, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "sage:calculus:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 70.988, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "70.988", + "raw_stderr": "5.601", + "latency": "150.536", + "cost_per_test": "0.105882", + "max_output_tokens": "30000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.601, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 68.826, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "68.826", + "raw_stderr": "5.226", + "latency": "101.22", + "cost_per_test": "0.07285", + "max_output_tokens": "30000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.226, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "sage:overall:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 43.312, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "43.312", + "raw_stderr": "3.115", + "latency": "152.404", + "cost_per_test": "0.098204", + "max_output_tokens": "30000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.115, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-2026-03-05/57f918cd-b975-4853-bd01-970478a75378.json b/data/vals_ai/openai/gpt-5.4-2026-03-05/57f918cd-b975-4853-bd01-970478a75378.json new file mode 100644 index 0000000000000000000000000000000000000000..b9834c9d968f6a404e2f4a9cc36e09146e0ca2e5 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-2026-03-05/57f918cd-b975-4853-bd01-970478a75378.json @@ -0,0 +1,3386 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/hlab/openai_gpt-5.4-2026-03-05/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Harvey's Legal Agent Benchmark", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-2026-03-05", + "id": "openai/gpt-5.4-2026-03-05", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-2026-03-05", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "hlab:antitrust_competition:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "809.935", + "cost_per_test": "2.734005", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:arbitration_international_dispute_resolution:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "940.904", + "cost_per_test": "4.937446", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:banking_finance:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "781.3", + "cost_per_test": "2.221578", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:bankruptcy_restructuring:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "940.396", + "cost_per_test": "4.66934", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:capital_markets:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "678.406", + "cost_per_test": "3.277627", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:corporate_governance:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "807.582", + "cost_per_test": "3.489705", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:corporate_ma:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "901.917", + "cost_per_test": "2.075068", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Criteria Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Criteria Pass Rate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 74.497, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate", + "task_name": "Criteria Pass Rate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "74.497", + "raw_stderr": "0.524", + "latency": "876.223", + "cost_per_test": "3.598304", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.524, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__antitrust_competition:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 76.449, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "76.449", + "raw_stderr": "2.559", + "latency": "809.935", + "cost_per_test": "2.734005", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.559, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__arbitration_international_dispute_resolution:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "66.667", + "raw_stderr": "2.827", + "latency": "940.904", + "cost_per_test": "4.937446", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.827, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__banking_finance:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 80.297, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.297", + "raw_stderr": "2.43", + "latency": "781.3", + "cost_per_test": "2.221578", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.43, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__bankruptcy_restructuring:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 68.512, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "68.512", + "raw_stderr": "2.737", + "latency": "940.396", + "cost_per_test": "4.66934", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.737, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__capital_markets:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 81.569, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.569", + "raw_stderr": "2.433", + "latency": "678.406", + "cost_per_test": "3.277627", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.433, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_governance:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 72.945, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "72.945", + "raw_stderr": "2.604", + "latency": "807.582", + "cost_per_test": "3.489705", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.604, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_ma:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 81.731, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.731", + "raw_stderr": "2.191", + "latency": "901.917", + "cost_per_test": "2.075068", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.191, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__data_privacy_cybersecurity:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 79.152, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.152", + "raw_stderr": "2.419", + "latency": "812.538", + "cost_per_test": "2.616626", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.419, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__emerging_companies_venture_capital:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 69.625, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "69.625", + "raw_stderr": "2.691", + "latency": "967.826", + "cost_per_test": "2.63225", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.691, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__employment_labor:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 64.794, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "64.794", + "raw_stderr": "2.928", + "latency": "790.096", + "cost_per_test": "1.501786", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.928, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__energy_natural_resources:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 87.459, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.459", + "raw_stderr": "1.906", + "latency": "1088.193", + "cost_per_test": "9.258809", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.906, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__environmental_esg:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 72.401, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "72.401", + "raw_stderr": "2.681", + "latency": "882.074", + "cost_per_test": "2.235795", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.681, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__funds_asset_management:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 71.137, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "71.137", + "raw_stderr": "2.45", + "latency": "974.287", + "cost_per_test": "3.406867", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.45, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__healthcare_life_sciences:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 74.257, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "74.257", + "raw_stderr": "2.516", + "latency": "934.798", + "cost_per_test": "2.130891", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.516, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__immigration:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 72.364, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "72.364", + "raw_stderr": "2.702", + "latency": "771.27", + "cost_per_test": "2.396939", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.702, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__insurance:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 73.26, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "73.26", + "raw_stderr": "2.684", + "latency": "708.141", + "cost_per_test": "1.663014", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.684, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__intellectual_property:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 82.332, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.332", + "raw_stderr": "2.271", + "latency": "903.214", + "cost_per_test": "3.326913", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.271, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__international_trade_sanctions:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 74.603, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "74.603", + "raw_stderr": "2.747", + "latency": "750.936", + "cost_per_test": "1.851192", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.747, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__litigation_dispute_resolution:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 76.779, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "76.779", + "raw_stderr": "2.589", + "latency": "941.832", + "cost_per_test": "9.61742", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.589, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__real_estate:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 80.758, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.758", + "raw_stderr": "2.132", + "latency": "1012.236", + "cost_per_test": "4.677799", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.132, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__structured_finance_securitization:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 70.794, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "70.794", + "raw_stderr": "2.566", + "latency": "842.82", + "cost_per_test": "2.506224", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.566, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__tax:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 57.525, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "57.525", + "raw_stderr": "2.863", + "latency": "913.987", + "cost_per_test": "3.459839", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.863, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__trusts_estates_private_client:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 80.534, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.534", + "raw_stderr": "2.451", + "latency": "1068.535", + "cost_per_test": "5.065086", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.451, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__white_collar_defense_investigations:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 72.391, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "72.391", + "raw_stderr": "2.599", + "latency": "806.12", + "cost_per_test": "4.607069", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.599, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:data_privacy_cybersecurity:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "812.538", + "cost_per_test": "2.616626", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:emerging_companies_venture_capital:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "967.826", + "cost_per_test": "2.63225", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:employment_labor:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "790.096", + "cost_per_test": "1.501786", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:energy_natural_resources:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1088.193", + "cost_per_test": "9.258809", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:environmental_esg:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "882.074", + "cost_per_test": "2.235795", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:funds_asset_management:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "974.287", + "cost_per_test": "3.406867", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:healthcare_life_sciences:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "934.798", + "cost_per_test": "2.130891", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:immigration:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "771.27", + "cost_per_test": "2.396939", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:insurance:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "708.141", + "cost_per_test": "1.663014", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:intellectual_property:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "903.214", + "cost_per_test": "3.326913", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:international_trade_sanctions:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "750.936", + "cost_per_test": "1.851192", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:litigation_dispute_resolution:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "941.832", + "cost_per_test": "9.61742", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:overall:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.overall", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Overall).", + "metric_id": "vals_ai.hlab.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "876.223", + "cost_per_test": "3.598304", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:real_estate:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1012.236", + "cost_per_test": "4.677799", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:structured_finance_securitization:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "842.82", + "cost_per_test": "2.506224", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:tax:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "913.987", + "cost_per_test": "3.459839", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:trusts_estates_private_client:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1068.535", + "cost_per_test": "5.065086", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:white_collar_defense_investigations:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.hlab.white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "806.12", + "cost_per_test": "4.607069", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-2026-03-05/7498ef18-2509-4cfd-8214-9fab0e8778e0.json b/data/vals_ai/openai/gpt-5.4-2026-03-05/7498ef18-2509-4cfd-8214-9fab0e8778e0.json new file mode 100644 index 0000000000000000000000000000000000000000..44e8e556eb3ce28cf8f3a841b4216cbf46cfddae --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-2026-03-05/7498ef18-2509-4cfd-8214-9fab0e8778e0.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/openai_gpt-5.4-2026-03-05/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-2026-03-05", + "id": "openai/gpt-5.4-2026-03-05", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-2026-03-05", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 91.414, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.414", + "raw_stderr": "3.127", + "latency": "186.909", + "cost_per_test": "0.196199", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.127, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 91.919, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.919", + "raw_stderr": "2.188", + "latency": "218.997", + "cost_per_test": "0.222481", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.188, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 91.666, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.666", + "raw_stderr": "1.908", + "latency": "202.953", + "cost_per_test": "0.20934", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.908, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-2026-03-05/81b85237-3fe2-446c-8efb-d97818f83b1a.json b/data/vals_ai/openai/gpt-5.4-2026-03-05/81b85237-3fe2-446c-8efb-d97818f83b1a.json new file mode 100644 index 0000000000000000000000000000000000000000..6ebc23e61edb846430bea518a71234071bda7718 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-2026-03-05/81b85237-3fe2-446c-8efb-d97818f83b1a.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/openai_gpt-5.4-2026-03-05/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-2026-03-05", + "id": "openai/gpt-5.4-2026-03-05", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-2026-03-05", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 64.922, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.922", + "raw_stderr": "1.365", + "latency": "78.717", + "cost_per_test": "0.157344", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.365, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 73.958, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "73.958", + "raw_stderr": "0.868", + "latency": "78.717", + "cost_per_test": "0.157344", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.868, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 82.993, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "82.993", + "raw_stderr": "1.074", + "latency": "78.717", + "cost_per_test": "0.157344", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.074, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-2026-03-05/8ef80961-1787-4424-97f5-715aad5023b0.json b/data/vals_ai/openai/gpt-5.4-2026-03-05/8ef80961-1787-4424-97f5-715aad5023b0.json new file mode 100644 index 0000000000000000000000000000000000000000..20dd5dbd8c3b14f7de39a5032c95f705ec8be6ad --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-2026-03-05/8ef80961-1787-4424-97f5-715aad5023b0.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/openai_gpt-5.4-2026-03-05/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-2026-03-05", + "id": "openai/gpt-5.4-2026-03-05", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-2026-03-05", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 56.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "56", + "raw_stderr": "4.964", + "latency": "2751.526", + "cost_per_test": "3.202015", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.964, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-2026-03-05/8f2edac2-27d6-490d-a2f2-3263b43b0a34.json b/data/vals_ai/openai/gpt-5.4-2026-03-05/8f2edac2-27d6-490d-a2f2-3263b43b0a34.json new file mode 100644 index 0000000000000000000000000000000000000000..598d46d57fb35e7ad180db819c8705a0bc833741 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-2026-03-05/8f2edac2-27d6-490d-a2f2-3263b43b0a34.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/openai_gpt-5.4-2026-03-05/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-2026-03-05", + "id": "openai/gpt-5.4-2026-03-05", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-2026-03-05", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 77.549, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "77.549", + "raw_stderr": "3.316", + "latency": "282.914", + "cost_per_test": "0.639282", + "max_output_tokens": "30000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.316, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-2026-03-05/95eeb16f-da2e-467b-bb28-f54c45b6f69f.json b/data/vals_ai/openai/gpt-5.4-2026-03-05/95eeb16f-da2e-467b-bb28-f54c45b6f69f.json new file mode 100644 index 0000000000000000000000000000000000000000..3e86850e0208596a7e6aa650ffccf10d7e4bf8a2 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-2026-03-05/95eeb16f-da2e-467b-bb28-f54c45b6f69f.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/openai_gpt-5.4-2026-03-05/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-2026-03-05", + "id": "openai/gpt-5.4-2026-03-05", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-2026-03-05", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 99.068, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "99.068", + "raw_stderr": "0.535", + "latency": "46.546", + "cost_per_test": "0.032344", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "lcb:hard:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 67.714, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "67.714", + "raw_stderr": "2.499", + "latency": "421.687", + "cost_per_test": "0.743287", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.499, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "lcb:medium:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 85.64, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "85.64", + "raw_stderr": "1.792", + "latency": "160.04", + "cost_per_test": "0.212107", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.792, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "lcb:overall:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 84.141, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "84.141", + "raw_stderr": "1.04", + "latency": "209.424", + "cost_per_test": "0.329246", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.04, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-2026-03-05/af24b023-8f43-4244-8629-9ef5b729ce47.json b/data/vals_ai/openai/gpt-5.4-2026-03-05/af24b023-8f43-4244-8629-9ef5b729ce47.json new file mode 100644 index 0000000000000000000000000000000000000000..d9baf934fbc596327a9455d03e6569d7daa476e1 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-2026-03-05/af24b023-8f43-4244-8629-9ef5b729ce47.json @@ -0,0 +1,371 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/openai_gpt-5.4-2026-03-05/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-2026-03-05", + "id": "openai/gpt-5.4-2026-03-05", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-2026-03-05", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 50.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "50", + "raw_stderr": "7.809", + "latency": "479.149", + "cost_per_test": "1.371538", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.809, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 76.245, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "76.245", + "raw_stderr": "2.639", + "latency": "326.944", + "cost_per_test": "0.856764", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.639, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 88.144, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "88.144", + "raw_stderr": "2.327", + "latency": "237.096", + "cost_per_test": "0.589203", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.327, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "702.037", + "cost_per_test": "1.680539", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "swebench:overall:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 78.2, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "78.2", + "raw_stderr": "1.848", + "latency": "307.118", + "cost_per_test": "0.801134", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.848, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-2026-03-05/cefd392d-ca9d-46a5-a64e-fb1b23d44f4c.json b/data/vals_ai/openai/gpt-5.4-2026-03-05/cefd392d-ca9d-46a5-a64e-fb1b23d44f4c.json new file mode 100644 index 0000000000000000000000000000000000000000..ada0a5bf534b90834758f77fb6ce61a282300d8d --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-2026-03-05/cefd392d-ca9d-46a5-a64e-fb1b23d44f4c.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/code-migration/openai_gpt-5.4-2026-03-05/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Code Migration", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-2026-03-05", + "id": "openai/gpt-5.4-2026-03-05", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-2026-03-05", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "code-migration:cli:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.code-migration.cli", + "source_data": { + "dataset_name": "Code Migration - CLI", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cli", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (CLI).", + "metric_id": "vals_ai.code-migration.cli.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 23.788, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cli", + "task_name": "CLI", + "dataset_type": "private", + "industry": "coding", + "raw_score": "23.788", + "raw_stderr": "2.23", + "latency": "2507.905", + "cost_per_test": "9.677001", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.23, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "code-migration:cobol:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.code-migration.cobol", + "source_data": { + "dataset_name": "Code Migration - COBOL", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cobol", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (COBOL).", + "metric_id": "vals_ai.code-migration.cobol.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 68.571, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cobol", + "task_name": "COBOL", + "dataset_type": "private", + "industry": "coding", + "raw_score": "68.571", + "raw_stderr": "15.028", + "latency": "1061.3", + "cost_per_test": "1.99524", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 15.028, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "code-migration:code_quality:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.code-migration.code_quality", + "source_data": { + "dataset_name": "Code Migration - Code Quality", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "code_quality", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Code Quality).", + "metric_id": "vals_ai.code-migration.code_quality.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 3.632, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "code_quality", + "task_name": "Code Quality", + "dataset_type": "private", + "industry": "coding", + "raw_score": "3.632", + "raw_stderr": "0.085", + "latency": "2146.253", + "cost_per_test": "7.756561", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.085, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "code-migration:overall:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.code-migration.overall", + "source_data": { + "dataset_name": "Code Migration - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Overall).", + "metric_id": "vals_ai.code-migration.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 34.984, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "34.984", + "raw_stderr": "4.113", + "latency": "2146.253", + "cost_per_test": "7.756561", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.113, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-2026-03-05/d8ab853a-12f9-4061-8db7-0225f39abe60.json b/data/vals_ai/openai/gpt-5.4-2026-03-05/d8ab853a-12f9-4061-8db7-0225f39abe60.json new file mode 100644 index 0000000000000000000000000000000000000000..4cbef1b954d06a6690ff4a3be015ec345015bbaf --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-2026-03-05/d8ab853a-12f9-4061-8db7-0225f39abe60.json @@ -0,0 +1,103 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/openai_gpt-5.4-2026-03-05/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-2026-03-05", + "id": "openai/gpt-5.4-2026-03-05", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-2026-03-05", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 87.514, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.514", + "raw_stderr": "0.795", + "latency": "65.565", + "cost_per_test": "0.047083", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.795, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-2026-03-05/e0bfd7b0-ee77-4eb2-99e7-827d90d3d6bb.json b/data/vals_ai/openai/gpt-5.4-2026-03-05/e0bfd7b0-ee77-4eb2-99e7-827d90d3d6bb.json new file mode 100644 index 0000000000000000000000000000000000000000..2ebb9d55bef7f85c4df401fb3c32042e7b4a1ce4 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-2026-03-05/e0bfd7b0-ee77-4eb2-99e7-827d90d3d6bb.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/openai_gpt-5.4-2026-03-05/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-2026-03-05", + "id": "openai/gpt-5.4-2026-03-05", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-2026-03-05", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 75.994, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "75.994", + "raw_stderr": "1.204", + "latency": "56.849", + "cost_per_test": "0.130934", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.204, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 68.323, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.323", + "raw_stderr": "0.915", + "latency": "45.002", + "cost_per_test": "0.103445", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.915, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 60.652, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.652", + "raw_stderr": "1.377", + "latency": "33.154", + "cost_per_test": "0.075956", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.377, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-2026-03-05/ea40698b-1b38-4c57-a81a-80e449710ae0.json b/data/vals_ai/openai/gpt-5.4-2026-03-05/ea40698b-1b38-4c57-a81a-80e449710ae0.json new file mode 100644 index 0000000000000000000000000000000000000000..bd7aed37a42590f154c2b5a5de645df70409e785 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-2026-03-05/ea40698b-1b38-4c57-a81a-80e449710ae0.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/programbench/openai_gpt-5.4-2026-03-05/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProgramBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-2026-03-05", + "id": "openai/gpt-5.4-2026-03-05", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-2026-03-05", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "programbench:almost:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.programbench.almost", + "source_data": { + "dataset_name": "ProgramBench - Almost Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "almost", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Almost Resolved).", + "metric_id": "vals_ai.programbench.almost.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 7.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "almost", + "task_name": "Almost Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "7", + "raw_stderr": "1.809", + "latency": "2386.411", + "cost_per_test": "7.379154", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.809, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "programbench:overall:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.programbench.overall", + "source_data": { + "dataset_name": "ProgramBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Overall).", + "metric_id": "vals_ai.programbench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "2386.411", + "cost_per_test": "7.379154", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "programbench:partial:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.programbench.partial", + "source_data": { + "dataset_name": "ProgramBench - Raw Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "partial", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Raw Pass Rate).", + "metric_id": "vals_ai.programbench.partial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 68.596, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "partial", + "task_name": "Raw Pass Rate", + "dataset_type": "public", + "industry": "coding", + "raw_score": "68.596", + "raw_stderr": "1.792", + "latency": "2386.411", + "cost_per_test": "7.379154", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.792, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "programbench:strict:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.programbench.strict", + "source_data": { + "dataset_name": "ProgramBench - Fully Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "strict", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Fully Resolved).", + "metric_id": "vals_ai.programbench.strict.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "strict", + "task_name": "Fully Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "2386.411", + "cost_per_test": "7.379154", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-2026-03-05/edd3a1c1-a5c2-4ed2-a379-cd44fc6b2358.json b/data/vals_ai/openai/gpt-5.4-2026-03-05/edd3a1c1-a5c2-4ed2-a379-cd44fc6b2358.json new file mode 100644 index 0000000000000000000000000000000000000000..664ae5c7b82b33043866354bf92c7b5be979c0cb --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-2026-03-05/edd3a1c1-a5c2-4ed2-a379-cd44fc6b2358.json @@ -0,0 +1,438 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/openai_gpt-5.4-2026-03-05/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-2026-03-05", + "id": "openai/gpt-5.4-2026-03-05", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-2026-03-05", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 88.939, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.939", + "raw_stderr": "0.711", + "latency": "18.611", + "cost_per_test": "0.015054", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.711, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.921, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.921", + "raw_stderr": "0.199", + "latency": "22.065", + "cost_per_test": "0.019498", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.199, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 90.526, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.526", + "raw_stderr": "0.569", + "latency": "9.338", + "cost_per_test": "0.008692", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.569, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 86.044, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.044", + "raw_stderr": "0.414", + "latency": "27.787", + "cost_per_test": "0.023053", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.414, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.824, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.824", + "raw_stderr": "1.304", + "latency": "14.027", + "cost_per_test": "0.012138", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.304, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:openai/gpt-5.4-2026-03-05:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 84.008, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.008", + "raw_stderr": "1.306", + "latency": "74.892", + "cost_per_test": "0.059881", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.306, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/047f75c4-4b1a-48b7-a117-3557133a47d0.json b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/047f75c4-4b1a-48b7-a117-3557133a47d0.json new file mode 100644 index 0000000000000000000000000000000000000000..6de1adf3f3f490267daebbd6e1d16e633d7d8079 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/047f75c4-4b1a-48b7-a117-3557133a47d0.json @@ -0,0 +1,371 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/openai_gpt-5.4-mini-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-mini-2026-03-17", + "id": "openai/gpt-5.4-mini-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-mini-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "7.362", + "latency": "506.724", + "cost_per_test": "0.954575", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.362, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 70.881, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "70.881", + "raw_stderr": "2.818", + "latency": "356.811", + "cost_per_test": "0.555325", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.818, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 85.567, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "85.567", + "raw_stderr": "2.53", + "latency": "246.086", + "cost_per_test": "0.353731", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.53, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "359.782", + "cost_per_test": "0.623433", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "swebench:overall:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 73.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "73", + "raw_stderr": "1.987", + "latency": "326.46", + "cost_per_test": "0.511052", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.987, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/1692bb6b-389b-4d59-b8e2-5c163e1a472d.json b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/1692bb6b-389b-4d59-b8e2-5c163e1a472d.json new file mode 100644 index 0000000000000000000000000000000000000000..945634d4b5b1fd8ba436b8719c7c4ba2bc99e72f --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/1692bb6b-389b-4d59-b8e2-5c163e1a472d.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/openai_gpt-5.4-mini-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-mini-2026-03-17", + "id": "openai/gpt-5.4-mini-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-mini-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 21.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "21", + "raw_stderr": "4.073", + "latency": "1410.671", + "cost_per_test": "0.249441", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.073, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/290123fb-f79a-411a-a662-c0adca3fbaeb.json b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/290123fb-f79a-411a-a662-c0adca3fbaeb.json new file mode 100644 index 0000000000000000000000000000000000000000..c16e155faa156a6b3911bd11e4dacae2c72c0451 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/290123fb-f79a-411a-a662-c0adca3fbaeb.json @@ -0,0 +1,103 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/openai_gpt-5.4-mini-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-mini-2026-03-17", + "id": "openai/gpt-5.4-mini-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-mini-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 79.249, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.249", + "raw_stderr": "0.975", + "latency": "50.813", + "cost_per_test": "0.007061", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.975, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/2b19c6d9-5da5-430d-ae6b-81fae68169c8.json b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/2b19c6d9-5da5-430d-ae6b-81fae68169c8.json new file mode 100644 index 0000000000000000000000000000000000000000..af2d55e377a00444dde5c24a630d95a555451b8d --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/2b19c6d9-5da5-430d-ae6b-81fae68169c8.json @@ -0,0 +1,676 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/emb/openai_gpt-5.4-mini-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - EMB", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-mini-2026-03-17", + "id": "openai/gpt-5.4-mini-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-mini-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "emb:comps:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.emb.comps", + "source_data": { + "dataset_name": "EMB - Comps", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "comps", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Comps).", + "metric_id": "vals_ai.emb.comps.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 40.488, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "comps", + "task_name": "Comps", + "dataset_type": "private", + "industry": "finance", + "raw_score": "40.488", + "raw_stderr": "8.948", + "latency": "3904.48", + "cost_per_test": "3.183265", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 8.948, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:dataroom:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.emb.dataroom", + "source_data": { + "dataset_name": "EMB - Dataroom", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dataroom", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Dataroom).", + "metric_id": "vals_ai.emb.dataroom.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 48.286, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dataroom", + "task_name": "Dataroom", + "dataset_type": "private", + "industry": "finance", + "raw_score": "48.286", + "raw_stderr": "9.764", + "latency": "1783.343", + "cost_per_test": "1.595869", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 9.764, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:dcf:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.emb.dcf", + "source_data": { + "dataset_name": "EMB - DCF", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dcf", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (DCF).", + "metric_id": "vals_ai.emb.dcf.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 40.605, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dcf", + "task_name": "DCF", + "dataset_type": "private", + "industry": "finance", + "raw_score": "40.605", + "raw_stderr": "8.924", + "latency": "2573.977", + "cost_per_test": "2.145553", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 8.924, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:lbo:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.emb.lbo", + "source_data": { + "dataset_name": "EMB - LBO", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "lbo", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (LBO).", + "metric_id": "vals_ai.emb.lbo.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 41.608, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "lbo", + "task_name": "LBO", + "dataset_type": "private", + "industry": "finance", + "raw_score": "41.608", + "raw_stderr": "5.198", + "latency": "3185.685", + "cost_per_test": "2.327975", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.198, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:ma:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.emb.ma", + "source_data": { + "dataset_name": "EMB - M&A", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "ma", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (M&A).", + "metric_id": "vals_ai.emb.ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 49.814, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "ma", + "task_name": "M&A", + "dataset_type": "private", + "industry": "finance", + "raw_score": "49.814", + "raw_stderr": "9.217", + "latency": "2013.006", + "cost_per_test": "1.621479", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 9.217, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:misc:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.emb.misc", + "source_data": { + "dataset_name": "EMB - Misc", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "misc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Misc).", + "metric_id": "vals_ai.emb.misc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 46.346, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "misc", + "task_name": "Misc", + "dataset_type": "private", + "industry": "finance", + "raw_score": "46.346", + "raw_stderr": "8.643", + "latency": "2474", + "cost_per_test": "2.013106", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 8.643, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:operating:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.emb.operating", + "source_data": { + "dataset_name": "EMB - Operating", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "operating", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Operating).", + "metric_id": "vals_ai.emb.operating.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 50.834, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "operating", + "task_name": "Operating", + "dataset_type": "private", + "industry": "finance", + "raw_score": "50.834", + "raw_stderr": "8.241", + "latency": "2063.042", + "cost_per_test": "1.973419", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 8.241, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:overall:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.emb.overall", + "source_data": { + "dataset_name": "EMB - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Overall).", + "metric_id": "vals_ai.emb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 45.426, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "45.426", + "raw_stderr": "3.645", + "latency": "2601.215", + "cost_per_test": "2.146984", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.645, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:scratch:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.emb.scratch", + "source_data": { + "dataset_name": "EMB - Scratch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "scratch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Scratch).", + "metric_id": "vals_ai.emb.scratch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 54.67, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "scratch", + "task_name": "Scratch", + "dataset_type": "private", + "industry": "finance", + "raw_score": "54.67", + "raw_stderr": "3.462", + "latency": "2533.883", + "cost_per_test": "2.010804", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.462, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:template:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.emb.template", + "source_data": { + "dataset_name": "EMB - Template", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "template", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Template).", + "metric_id": "vals_ai.emb.template.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 36.014, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "template", + "task_name": "Template", + "dataset_type": "private", + "industry": "finance", + "raw_score": "36.014", + "raw_stderr": "4.913", + "latency": "2668.546", + "cost_per_test": "2.283165", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.913, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/329d941d-38e2-411a-90ac-35823b6f72bc.json b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/329d941d-38e2-411a-90ac-35823b6f72bc.json new file mode 100644 index 0000000000000000000000000000000000000000..3c0dc7057cf9f8df73f00b303a9666d74b699a3c --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/329d941d-38e2-411a-90ac-35823b6f72bc.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/openai_gpt-5.4-mini-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-mini-2026-03-17", + "id": "openai/gpt-5.4-mini-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-mini-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.538, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.538", + "raw_stderr": "1.661", + "latency": "44.668", + "cost_per_test": "0.014516", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.661, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.956, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.956", + "raw_stderr": "1.665", + "latency": "66.943", + "cost_per_test": "0.146241", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.665, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.917, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.917", + "raw_stderr": "0.962", + "latency": "52.297", + "cost_per_test": "0.076368", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.962, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.256, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.256", + "raw_stderr": "1.671", + "latency": "45.279", + "cost_per_test": "0.068348", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.671, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/3d48e77d-2568-4e70-9755-e66ab48db4c3.json b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/3d48e77d-2568-4e70-9755-e66ab48db4c3.json new file mode 100644 index 0000000000000000000000000000000000000000..4ae2df935722d022d9b1cd8eba48a149dd6e39b1 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/3d48e77d-2568-4e70-9755-e66ab48db4c3.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/openai_gpt-5.4-mini-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-mini-2026-03-17", + "id": "openai/gpt-5.4-mini-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-mini-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 6.667, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "6.667", + "raw_stderr": "4.513", + "latency": "1876.864", + "cost_per_test": "0.384694", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.513, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 6.167, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "6.167", + "raw_stderr": "5.629", + "latency": "1916.743", + "cost_per_test": "0.316446", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.629, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "ioi:overall:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 6.417, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "6.417", + "raw_stderr": "3.607", + "latency": "1896.804", + "cost_per_test": "0.35057", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.607, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/50d88802-6704-4a1c-b572-c3ac18f26a74.json b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/50d88802-6704-4a1c-b572-c3ac18f26a74.json new file mode 100644 index 0000000000000000000000000000000000000000..b4acbcd03e61b5a880a25ca45f4c1154868bc3b8 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/50d88802-6704-4a1c-b572-c3ac18f26a74.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/openai_gpt-5.4-mini-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-mini-2026-03-17", + "id": "openai/gpt-5.4-mini-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-mini-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.353, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.353", + "raw_stderr": "1.08", + "latency": "14.45", + "cost_per_test": "0.002026", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.08, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.734, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.734", + "raw_stderr": "1.098", + "latency": "20.098", + "cost_per_test": "0.003267", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.098, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.134, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.134", + "raw_stderr": "0.928", + "latency": "15.401", + "cost_per_test": "0.003744", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.928, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.073, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.073", + "raw_stderr": "1.683", + "latency": "24.636", + "cost_per_test": "0.003661", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.683, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.744, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.744", + "raw_stderr": "1.093", + "latency": "15.064", + "cost_per_test": "0.00222", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.093, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.043, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.043", + "raw_stderr": "1.239", + "latency": "51.562", + "cost_per_test": "0.010098", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.239, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.073, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.073", + "raw_stderr": "1.406", + "latency": "16.235", + "cost_per_test": "0.002732", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.406, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.328, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.328", + "raw_stderr": "2.209", + "latency": "15.341", + "cost_per_test": "0.002869", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.209, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 70.118, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.118", + "raw_stderr": "1.384", + "latency": "30.578", + "cost_per_test": "0.006384", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.384, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.449, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.449", + "raw_stderr": "0.638", + "latency": "15.14", + "cost_per_test": "0.00289", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.638, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.411, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.411", + "raw_stderr": "1.322", + "latency": "16.063", + "cost_per_test": "0.002935", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.322, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.554, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.554", + "raw_stderr": "0.359", + "latency": "20.001", + "cost_per_test": "0.00368", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.359, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.561, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.561", + "raw_stderr": "1.785", + "latency": "15.236", + "cost_per_test": "0.003351", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.785, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.531, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.531", + "raw_stderr": "0.812", + "latency": "16.319", + "cost_per_test": "0.003314", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.812, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.211, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.211", + "raw_stderr": "1.299", + "latency": "13.894", + "cost_per_test": "0.002023", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.299, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/5363849a-d428-445f-9cb8-5f144ea8d177.json b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/5363849a-d428-445f-9cb8-5f144ea8d177.json new file mode 100644 index 0000000000000000000000000000000000000000..545322fe8954a3b70e9313c3cc1e8a8a15173210 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/5363849a-d428-445f-9cb8-5f144ea8d177.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/openai_gpt-5.4-mini-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-mini-2026-03-17", + "id": "openai/gpt-5.4-mini-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-mini-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 79.412, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "79.412", + "raw_stderr": "1.14", + "latency": "148.71", + "cost_per_test": "0.052812", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.14, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 63.514, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.514", + "raw_stderr": "0.906", + "latency": "126.454", + "cost_per_test": "0.044769", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.906, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 47.615, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "47.615", + "raw_stderr": "1.408", + "latency": "104.197", + "cost_per_test": "0.036726", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.408, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/6527b790-01dc-4481-ba74-1c329437077c.json b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/6527b790-01dc-4481-ba74-1c329437077c.json new file mode 100644 index 0000000000000000000000000000000000000000..e70993a4f1f05fe9902a9523a22633c3c3aca869 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/6527b790-01dc-4481-ba74-1c329437077c.json @@ -0,0 +1,1188 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_research/openai_gpt-5.4-mini-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Legal Research Bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-mini-2026-03-17", + "id": "openai/gpt-5.4-mini-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-mini-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_research:administrative_regulatory:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.legal_research.administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 15.152, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "15.152", + "raw_stderr": "4.447", + "latency": "1961.008", + "cost_per_test": "1.456649", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.447, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:business_commercial:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.legal_research.business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 11.628, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "11.628", + "raw_stderr": "4.946", + "latency": "2506.739", + "cost_per_test": "2.128528", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.946, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:civil_litigation:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.legal_research.civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 14.815, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "14.815", + "raw_stderr": "6.967", + "latency": "2120.497", + "cost_per_test": "1.875076", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.967, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:constitutional_civil_rights:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.legal_research.constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 13.158, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "13.158", + "raw_stderr": "5.557", + "latency": "1433.795", + "cost_per_test": "1.467472", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.557, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:criminal:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.legal_research.criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 5.263, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "5.263", + "raw_stderr": "2.984", + "latency": "2330.292", + "cost_per_test": "2.024329", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.984, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:family:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.legal_research.family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "4205.69", + "cost_per_test": "2.933958", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:health:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.legal_research.health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 16.667, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "16.667", + "raw_stderr": "6.92", + "latency": "1671.994", + "cost_per_test": "1.329631", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.92, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:immigration:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.legal_research.immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 18.182, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "18.182", + "raw_stderr": "12.197", + "latency": "2027.42", + "cost_per_test": "1.631269", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 12.197, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:overall:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.legal_research.overall", + "source_data": { + "dataset_name": "Legal Research Bench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Overall).", + "metric_id": "vals_ai.legal_research.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 12.5, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "legal", + "raw_score": "12.5", + "raw_stderr": "2.299", + "latency": "2256.339", + "cost_per_test": "1.864673", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.299, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass", + "source_data": { + "dataset_name": "Legal Research Bench - Weighted Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Weighted Pass).", + "metric_id": "vals_ai.legal_research.weighted_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 67.326, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass", + "task_name": "Weighted Pass", + "dataset_type": "private", + "industry": "legal", + "raw_score": "67.326", + "raw_stderr": "1.837", + "latency": "2256.339", + "cost_per_test": "1.864673", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.837, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__administrative_regulatory:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.weighted_pass__administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 69.72, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "69.72", + "raw_stderr": "3.312", + "latency": "1961.008", + "cost_per_test": "1.456649", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.312, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__business_commercial:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.weighted_pass__business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 71.172, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "71.172", + "raw_stderr": "3.437", + "latency": "2506.739", + "cost_per_test": "2.128528", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.437, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__civil_litigation:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.weighted_pass__civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 67.267, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "67.267", + "raw_stderr": "5.492", + "latency": "2120.497", + "cost_per_test": "1.875076", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.492, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__constitutional_civil_rights:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 74.113, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "74.113", + "raw_stderr": "3.484", + "latency": "1433.795", + "cost_per_test": "1.467472", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.484, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__criminal:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.weighted_pass__criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 64.389, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "64.389", + "raw_stderr": "3.396", + "latency": "2330.292", + "cost_per_test": "2.024329", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.396, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__family:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.weighted_pass__family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 51.57, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "51.57", + "raw_stderr": "5.858", + "latency": "4205.69", + "cost_per_test": "2.933958", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.858, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__health:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.weighted_pass__health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 73.772, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "73.772", + "raw_stderr": "4.948", + "latency": "1671.994", + "cost_per_test": "1.329631", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.948, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__immigration:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.weighted_pass__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 68.382, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "68.382", + "raw_stderr": "8.841", + "latency": "2027.42", + "cost_per_test": "1.631269", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 8.841, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/7d796c25-1691-4fc6-a6a2-3234550afbb7.json b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/7d796c25-1691-4fc6-a6a2-3234550afbb7.json new file mode 100644 index 0000000000000000000000000000000000000000..3c9f99a40a248907d35bae807521e09336d990ee --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/7d796c25-1691-4fc6-a6a2-3234550afbb7.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/programbench/openai_gpt-5.4-mini-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProgramBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-mini-2026-03-17", + "id": "openai/gpt-5.4-mini-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-mini-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "programbench:almost:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.programbench.almost", + "source_data": { + "dataset_name": "ProgramBench - Almost Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "almost", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Almost Resolved).", + "metric_id": "vals_ai.programbench.almost.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 3.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "almost", + "task_name": "Almost Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "3", + "raw_stderr": "1.209", + "latency": "3225.344", + "cost_per_test": "6.386444", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.209, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "programbench:overall:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.programbench.overall", + "source_data": { + "dataset_name": "ProgramBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Overall).", + "metric_id": "vals_ai.programbench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "3225.344", + "cost_per_test": "6.386444", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "programbench:partial:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.programbench.partial", + "source_data": { + "dataset_name": "ProgramBench - Raw Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "partial", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Raw Pass Rate).", + "metric_id": "vals_ai.programbench.partial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 54.375, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "partial", + "task_name": "Raw Pass Rate", + "dataset_type": "public", + "industry": "coding", + "raw_score": "54.375", + "raw_stderr": "1.954", + "latency": "3225.344", + "cost_per_test": "6.386444", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.954, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "programbench:strict:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.programbench.strict", + "source_data": { + "dataset_name": "ProgramBench - Fully Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "strict", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Fully Resolved).", + "metric_id": "vals_ai.programbench.strict.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "strict", + "task_name": "Fully Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "3225.344", + "cost_per_test": "6.386444", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/90e44ee2-6caa-453d-a3ad-b1aaa245427c.json b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/90e44ee2-6caa-453d-a3ad-b1aaa245427c.json new file mode 100644 index 0000000000000000000000000000000000000000..b0eaf9718acc008385ed0eee5e6df2f304f0ad65 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/90e44ee2-6caa-453d-a3ad-b1aaa245427c.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/openai_gpt-5.4-mini-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-mini-2026-03-17", + "id": "openai/gpt-5.4-mini-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-mini-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 84.848, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.848", + "raw_stderr": "3.55", + "latency": "60.078", + "cost_per_test": "0.012277", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.55, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 81.313, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.313", + "raw_stderr": "3.397", + "latency": "66.141", + "cost_per_test": "0.013209", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.397, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 83.08, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.08", + "raw_stderr": "2.457", + "latency": "63.11", + "cost_per_test": "0.012743", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.457, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/959c45cb-b72b-466c-bc99-c85829de0739.json b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/959c45cb-b72b-466c-bc99-c85829de0739.json new file mode 100644 index 0000000000000000000000000000000000000000..741dd425b3f82c8b612f1b60b078039f25bd1d6d --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/959c45cb-b72b-466c-bc99-c85829de0739.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/code-migration/openai_gpt-5.4-mini-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Code Migration", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-mini-2026-03-17", + "id": "openai/gpt-5.4-mini-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-mini-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "code-migration:cli:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.code-migration.cli", + "source_data": { + "dataset_name": "Code Migration - CLI", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cli", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (CLI).", + "metric_id": "vals_ai.code-migration.cli.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 8.2, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cli", + "task_name": "CLI", + "dataset_type": "private", + "industry": "coding", + "raw_score": "8.2", + "raw_stderr": "1.295", + "latency": "1956.927", + "cost_per_test": "2.31844", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.295, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "code-migration:cobol:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.code-migration.cobol", + "source_data": { + "dataset_name": "Code Migration - COBOL", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cobol", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (COBOL).", + "metric_id": "vals_ai.code-migration.cobol.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 27.143, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cobol", + "task_name": "COBOL", + "dataset_type": "private", + "industry": "coding", + "raw_score": "27.143", + "raw_stderr": "14.038", + "latency": "1063.103", + "cost_per_test": "0.857472", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 14.038, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "code-migration:code_quality:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.code-migration.code_quality", + "source_data": { + "dataset_name": "Code Migration - Code Quality", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "code_quality", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Code Quality).", + "metric_id": "vals_ai.code-migration.code_quality.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 3.053, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "code_quality", + "task_name": "Code Quality", + "dataset_type": "private", + "industry": "coding", + "raw_score": "3.053", + "raw_stderr": "0.076", + "latency": "1733.471", + "cost_per_test": "1.953198", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.076, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "code-migration:overall:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.code-migration.overall", + "source_data": { + "dataset_name": "Code Migration - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Overall).", + "metric_id": "vals_ai.code-migration.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 12.936, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "12.936", + "raw_stderr": "3.641", + "latency": "1733.471", + "cost_per_test": "1.953198", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.641, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/b1cc3582-72b4-4474-bade-a4d4ec1eb16c.json b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/b1cc3582-72b4-4474-bade-a4d4ec1eb16c.json new file mode 100644 index 0000000000000000000000000000000000000000..3c34d4397acb804ebccc0980e3618584787886aa --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/b1cc3582-72b4-4474-bade-a4d4ec1eb16c.json @@ -0,0 +1,356 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/openai_gpt-5.4-mini-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-mini-2026-03-17", + "id": "openai/gpt-5.4-mini-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-mini-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 39.683, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "39.683", + "raw_stderr": "6.148", + "latency": "133.32", + "cost_per_test": "0.078775", + "max_output_tokens": "30000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.148, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "sage:automata:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 41.27, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "41.27", + "raw_stderr": "6.284", + "latency": "195.033", + "cost_per_test": "0.093136", + "max_output_tokens": "30000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.284, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "sage:calculus:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 59.259, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "59.259", + "raw_stderr": "6.36", + "latency": "134.807", + "cost_per_test": "0.082637", + "max_output_tokens": "30000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.36, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 71.486, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "71.486", + "raw_stderr": "5.189", + "latency": "94.844", + "cost_per_test": "0.055846", + "max_output_tokens": "30000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.189, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "sage:overall:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 50.813, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "50.813", + "raw_stderr": "3.403", + "latency": "141.066", + "cost_per_test": "0.075919", + "max_output_tokens": "30000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.403, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/b4420f25-9eef-4e2b-9d43-8392ab5e954d.json b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/b4420f25-9eef-4e2b-9d43-8392ab5e954d.json new file mode 100644 index 0000000000000000000000000000000000000000..ec8571731151b4cf4c8b3192b7c9f2c22442951e --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/b4420f25-9eef-4e2b-9d43-8392ab5e954d.json @@ -0,0 +1,536 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_multimodal_index/openai_gpt-5.4-mini-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Multimodal Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-mini-2026-03-17", + "id": "openai/gpt-5.4-mini-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-mini-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_multimodal_index:corp_fin_v2:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.vals_multimodal_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Multimodal Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (CorpFin v2).", + "metric_id": "vals_ai.vals_multimodal_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 60.256, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "60.256", + "raw_stderr": "1.671", + "latency": "45.279", + "cost_per_test": "0.068348", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.671, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:finance_agent:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.vals_multimodal_index.finance_agent", + "source_data": { + "dataset_name": "Vals Multimodal Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_multimodal_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 45.407, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "45.407", + "raw_stderr": "1.406", + "latency": "1918.861", + "cost_per_test": "1.037185", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.406, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:mortgage_tax:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.vals_multimodal_index.mortgage_tax", + "source_data": { + "dataset_name": "Vals Multimodal Index - Mortgage Tax", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "mortgage_tax", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Mortgage Tax).", + "metric_id": "vals_ai.vals_multimodal_index.mortgage_tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 63.514, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "mortgage_tax", + "task_name": "Mortgage Tax", + "dataset_type": "private", + "industry": "index", + "raw_score": "63.514", + "raw_stderr": "0.906", + "latency": "126.454", + "cost_per_test": "0.044769", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.906, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:overall:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.vals_multimodal_index.overall", + "source_data": { + "dataset_name": "Vals Multimodal Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Overall).", + "metric_id": "vals_ai.vals_multimodal_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 54.219, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "54.219", + "raw_stderr": "1.863", + "latency": "759.345", + "cost_per_test": "0.470551", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.863, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:sage:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.vals_multimodal_index.sage", + "source_data": { + "dataset_name": "Vals Multimodal Index - SAGE", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "sage", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SAGE).", + "metric_id": "vals_ai.vals_multimodal_index.sage.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 50.813, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "sage", + "task_name": "SAGE", + "dataset_type": "private", + "industry": "index", + "raw_score": "50.813", + "raw_stderr": "3.403", + "latency": "141.066", + "cost_per_test": "0.075919", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.403, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:swebench:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.vals_multimodal_index.swebench", + "source_data": { + "dataset_name": "Vals Multimodal Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SWE-bench).", + "metric_id": "vals_ai.vals_multimodal_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 67.647, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "67.647", + "raw_stderr": "4.632", + "latency": "355.599", + "cost_per_test": "0.589598" + }, + "uncertainty": { + "standard_error": { + "value": 4.632, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:terminal_bench_2_1:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.vals_multimodal_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Multimodal Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_multimodal_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 54.682, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "54.682", + "raw_stderr": "4.556", + "latency": "673.036", + "cost_per_test": "0.286739", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.556, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:vibe_code_bench:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.vals_multimodal_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Multimodal Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_multimodal_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 42.523, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "42.523", + "raw_stderr": "8.901", + "latency": "2055.117", + "cost_per_test": "1.191299", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 8.901, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/cec48b03-5e42-43ca-ada4-15f28a210024.json b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/cec48b03-5e42-43ca-ada4-15f28a210024.json new file mode 100644 index 0000000000000000000000000000000000000000..9ff74f77a5806ea9542a45299a311c8cb2d53553 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/cec48b03-5e42-43ca-ada4-15f28a210024.json @@ -0,0 +1,740 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/openai_gpt-5.4-mini-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-mini-2026-03-17", + "id": "openai/gpt-5.4-mini-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-mini-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 45.366, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "45.366", + "raw_stderr": "1.159", + "latency": "2520.252", + "cost_per_test": "1.132987", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.159, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 32.369, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "32.369", + "raw_stderr": "0.155", + "latency": "2153.503", + "cost_per_test": "1.201815", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.155, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 40.273, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "40.273", + "raw_stderr": "2.567", + "latency": "2683.277", + "cost_per_test": "1.509603", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.567, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 48.338, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "48.338", + "raw_stderr": "0.778", + "latency": "2307.807", + "cost_per_test": "1.372803", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.778, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 65.294, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.294", + "raw_stderr": "0.426", + "latency": "1636.237", + "cost_per_test": "0.754578", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.426, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 19.617, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "19.617", + "raw_stderr": "1.636", + "latency": "2159.906", + "cost_per_test": "1.074981", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.636, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 46.366, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "46.366", + "raw_stderr": "0.966", + "latency": "2096.642", + "cost_per_test": "1.273272", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.966, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 64.025, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.025", + "raw_stderr": "2.039", + "latency": "1661.07", + "cost_per_test": "1.07807", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.039, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 56.85, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.85", + "raw_stderr": "1.068", + "latency": "1289.162", + "cost_per_test": "0.809499", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.068, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:overall:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 45.36, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "45.36", + "raw_stderr": "0.447", + "latency": "2153.503", + "cost_per_test": "1.201815", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.447, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 22.111, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "22.111", + "raw_stderr": "2.142", + "latency": "3206.023", + "cost_per_test": "1.890767", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.142, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/d01b61d6-28fc-409d-ab45-d6e1e2639b07.json b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/d01b61d6-28fc-409d-ab45-d6e1e2639b07.json new file mode 100644 index 0000000000000000000000000000000000000000..3ab020a22130692bb558f5cf633bb65fe5577921 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/d01b61d6-28fc-409d-ab45-d6e1e2639b07.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/openai_gpt-5.4-mini-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-mini-2026-03-17", + "id": "openai/gpt-5.4-mini-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-mini-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 97.826, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "97.826", + "raw_stderr": "0.813", + "latency": "11.686", + "cost_per_test": "0.001914", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.813, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "lcb:hard:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 61.714, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "61.714", + "raw_stderr": "2.598", + "latency": "442.662", + "cost_per_test": "0.076308", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.598, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "lcb:medium:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 84.856, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "84.856", + "raw_stderr": "1.832", + "latency": "111.923", + "cost_per_test": "0.020582", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.832, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "lcb:overall:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 81.465, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "81.465", + "raw_stderr": "1.094", + "latency": "188.757", + "cost_per_test": "0.032935", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.094, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/e1e3e61d-2718-4846-95c8-b74c7c58b597.json b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/e1e3e61d-2718-4846-95c8-b74c7c58b597.json new file mode 100644 index 0000000000000000000000000000000000000000..0cddd912495e27baf024b5337dea75653eef1b71 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/e1e3e61d-2718-4846-95c8-b74c7c58b597.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/openai_gpt-5.4-mini-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-mini-2026-03-17", + "id": "openai/gpt-5.4-mini-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-mini-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 63.042, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.042", + "raw_stderr": "1.38", + "latency": "44.578", + "cost_per_test": "0.00677", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.38, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 71.218, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.218", + "raw_stderr": "0.9", + "latency": "44.578", + "cost_per_test": "0.00677", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.9, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 79.395, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "79.395", + "raw_stderr": "1.157", + "latency": "44.578", + "cost_per_test": "0.00677", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.157, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/e6cf0c46-7615-45b0-b4ef-423c1e9e0eaa.json b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/e6cf0c46-7615-45b0-b4ef-423c1e9e0eaa.json new file mode 100644 index 0000000000000000000000000000000000000000..bfead5675cbf8914da2ebfa2c2398b2e24c3eb29 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/e6cf0c46-7615-45b0-b4ef-423c1e9e0eaa.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/openai_gpt-5.4-mini-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-mini-2026-03-17", + "id": "openai/gpt-5.4-mini-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-mini-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 47.969, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "47.969", + "raw_stderr": "5.606", + "latency": "2055.117", + "cost_per_test": "1.191299", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.606, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/ee9c54d1-e105-433c-836b-c9d4f6c534a8.json b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/ee9c54d1-e105-433c-836b-c9d4f6c534a8.json new file mode 100644 index 0000000000000000000000000000000000000000..c968f05fcb80686d8eb8d373eaed39574ce76fb6 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/ee9c54d1-e105-433c-836b-c9d4f6c534a8.json @@ -0,0 +1,408 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/openai_gpt-5.4-mini-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-mini-2026-03-17", + "id": "openai/gpt-5.4-mini-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-mini-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 60.256, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "60.256", + "raw_stderr": "1.671", + "latency": "45.279", + "cost_per_test": "0.068348", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.671, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 45.407, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "45.407", + "raw_stderr": "1.406", + "latency": "1918.861", + "cost_per_test": "1.037185", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.406, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_index:overall:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 52.425, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "52.425", + "raw_stderr": "2.054", + "latency": "1009.578", + "cost_per_test": "0.634634", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.054, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 67.647, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "67.647", + "raw_stderr": "4.632", + "latency": "355.599", + "cost_per_test": "0.589598" + }, + "uncertainty": { + "standard_error": { + "value": 4.632, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 54.682, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "54.682", + "raw_stderr": "4.556", + "latency": "673.036", + "cost_per_test": "0.286739", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.556, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 42.523, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "42.523", + "raw_stderr": "8.901", + "latency": "2055.117", + "cost_per_test": "1.191299", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 8.901, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/fadf48a3-f72c-4f31-a6f0-359d9bc944b2.json b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/fadf48a3-f72c-4f31-a6f0-359d9bc944b2.json new file mode 100644 index 0000000000000000000000000000000000000000..194416f2de8e1df33c7bff920b81d8d79e8afdde --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-mini-2026-03-17/fadf48a3-f72c-4f31-a6f0-359d9bc944b2.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/openai_gpt-5.4-mini-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-mini-2026-03-17", + "id": "openai/gpt-5.4-mini-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-mini-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 75.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "75", + "raw_stderr": "0", + "latency": "14975.056", + "cost_per_test": "6.379954", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 34.444, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "34.444", + "raw_stderr": "6.759", + "latency": "1996.674", + "cost_per_test": "0.85066", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.759, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 64.242, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "64.242", + "raw_stderr": "3.687", + "latency": "1089.095", + "cost_per_test": "0.463997", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.687, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:openai/gpt-5.4-mini-2026-03-17:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 54.682, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "54.682", + "raw_stderr": "4.556", + "latency": "673.036", + "cost_per_test": "0.286739", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.556, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/037effae-7152-44b9-adc5-b01a3f1716d4.json b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/037effae-7152-44b9-adc5-b01a3f1716d4.json new file mode 100644 index 0000000000000000000000000000000000000000..9631598dda57a88a21ee1b653285db5508727685 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/037effae-7152-44b9-adc5-b01a3f1716d4.json @@ -0,0 +1,536 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_multimodal_index/openai_gpt-5.4-nano-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Multimodal Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-nano-2026-03-17", + "id": "openai/gpt-5.4-nano-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-nano-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_multimodal_index:corp_fin_v2:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.vals_multimodal_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Multimodal Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (CorpFin v2).", + "metric_id": "vals_ai.vals_multimodal_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 61.072, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "61.072", + "raw_stderr": "1.665", + "latency": "7.363", + "cost_per_test": "0.01612", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.665, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:finance_agent:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.vals_multimodal_index.finance_agent", + "source_data": { + "dataset_name": "Vals Multimodal Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_multimodal_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 38.236, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "38.236", + "raw_stderr": "0.381", + "latency": "326.257", + "cost_per_test": "0.152321", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.381, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:mortgage_tax:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.vals_multimodal_index.mortgage_tax", + "source_data": { + "dataset_name": "Vals Multimodal Index - Mortgage Tax", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "mortgage_tax", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Mortgage Tax).", + "metric_id": "vals_ai.vals_multimodal_index.mortgage_tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 59.102, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "mortgage_tax", + "task_name": "Mortgage Tax", + "dataset_type": "private", + "industry": "index", + "raw_score": "59.102", + "raw_stderr": "0.964", + "latency": "14.526", + "cost_per_test": "0.002585", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.964, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:overall:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.vals_multimodal_index.overall", + "source_data": { + "dataset_name": "Vals Multimodal Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Overall).", + "metric_id": "vals_ai.vals_multimodal_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 47.643, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "47.643", + "raw_stderr": "1.707", + "latency": "671.046", + "cost_per_test": "0.233578", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.707, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:sage:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.vals_multimodal_index.sage", + "source_data": { + "dataset_name": "Vals Multimodal Index - SAGE", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "sage", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SAGE).", + "metric_id": "vals_ai.vals_multimodal_index.sage.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 38.081, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "sage", + "task_name": "SAGE", + "dataset_type": "private", + "industry": "index", + "raw_score": "38.081", + "raw_stderr": "3.087", + "latency": "34.334", + "cost_per_test": "0.006061", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.087, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:swebench:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.vals_multimodal_index.swebench", + "source_data": { + "dataset_name": "Vals Multimodal Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SWE-bench).", + "metric_id": "vals_ai.vals_multimodal_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 67.647, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "67.647", + "raw_stderr": "4.632", + "latency": "281.133", + "cost_per_test": "0.107316" + }, + "uncertainty": { + "standard_error": { + "value": 4.632, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:terminal_bench_2_1:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.vals_multimodal_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Multimodal Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_multimodal_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 41.573, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "41.573", + "raw_stderr": "2.973", + "latency": "757.485", + "cost_per_test": "0.072321", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.973, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:vibe_code_bench:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.vals_multimodal_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Multimodal Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_multimodal_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 30.032, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "30.032", + "raw_stderr": "8.282", + "latency": "3276.224", + "cost_per_test": "1.277298", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 8.282, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/081bf614-f2b7-4969-aa3e-72404052a832.json b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/081bf614-f2b7-4969-aa3e-72404052a832.json new file mode 100644 index 0000000000000000000000000000000000000000..ec83d00e9263cb285a12f935d097451e14664115 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/081bf614-f2b7-4969-aa3e-72404052a832.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/openai_gpt-5.4-nano-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-nano-2026-03-17", + "id": "openai/gpt-5.4-nano-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-nano-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 83.333, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "83.333", + "raw_stderr": "16.667", + "latency": "16854.051", + "cost_per_test": "1.609147", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 16.667, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 25.556, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "25.556", + "raw_stderr": "4.843", + "latency": "2247.207", + "cost_per_test": "0.214553", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.843, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 47.273, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "47.273", + "raw_stderr": "1.05", + "latency": "1225.749", + "cost_per_test": "0.117029", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.05, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 41.573, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "41.573", + "raw_stderr": "2.973", + "latency": "757.485", + "cost_per_test": "0.072321", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.973, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/0f90aabb-2d97-4934-8bb3-7df7e7af7ea9.json b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/0f90aabb-2d97-4934-8bb3-7df7e7af7ea9.json new file mode 100644 index 0000000000000000000000000000000000000000..11b8b7ae95a7729f290f1562135bfebb554db20c --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/0f90aabb-2d97-4934-8bb3-7df7e7af7ea9.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/openai_gpt-5.4-nano-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-nano-2026-03-17", + "id": "openai/gpt-5.4-nano-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-nano-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 14.167, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "14.167", + "raw_stderr": "6.269", + "latency": "967.635", + "cost_per_test": "0.035235", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.269, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 16.333, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "16.333", + "raw_stderr": "11.246", + "latency": "1373.348", + "cost_per_test": "0.050431", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 11.246, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "ioi:overall:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 15.25, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "15.25", + "raw_stderr": "6.438", + "latency": "1170.492", + "cost_per_test": "0.042833", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.438, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/1b91efc3-e217-4fd2-8dda-366612c2fa2d.json b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/1b91efc3-e217-4fd2-8dda-366612c2fa2d.json new file mode 100644 index 0000000000000000000000000000000000000000..a1119a3598bd48db4a84736c0ee54cfe3543566e --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/1b91efc3-e217-4fd2-8dda-366612c2fa2d.json @@ -0,0 +1,676 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/emb/openai_gpt-5.4-nano-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - EMB", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-nano-2026-03-17", + "id": "openai/gpt-5.4-nano-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-nano-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "emb:comps:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.emb.comps", + "source_data": { + "dataset_name": "EMB - Comps", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "comps", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Comps).", + "metric_id": "vals_ai.emb.comps.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 46.949, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "comps", + "task_name": "Comps", + "dataset_type": "private", + "industry": "finance", + "raw_score": "46.949", + "raw_stderr": "6.637", + "latency": "1561.313", + "cost_per_test": "0.585614", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.637, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:dataroom:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.emb.dataroom", + "source_data": { + "dataset_name": "EMB - Dataroom", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dataroom", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Dataroom).", + "metric_id": "vals_ai.emb.dataroom.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 71.966, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dataroom", + "task_name": "Dataroom", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.966", + "raw_stderr": "8.9", + "latency": "1605.292", + "cost_per_test": "0.479352", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 8.9, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:dcf:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.emb.dcf", + "source_data": { + "dataset_name": "EMB - DCF", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dcf", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (DCF).", + "metric_id": "vals_ai.emb.dcf.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 37.537, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dcf", + "task_name": "DCF", + "dataset_type": "private", + "industry": "finance", + "raw_score": "37.537", + "raw_stderr": "7.789", + "latency": "1984.318", + "cost_per_test": "0.64353", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.789, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:lbo:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.emb.lbo", + "source_data": { + "dataset_name": "EMB - LBO", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "lbo", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (LBO).", + "metric_id": "vals_ai.emb.lbo.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 32.592, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "lbo", + "task_name": "LBO", + "dataset_type": "private", + "industry": "finance", + "raw_score": "32.592", + "raw_stderr": "4.258", + "latency": "2103.732", + "cost_per_test": "0.681064", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.258, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:ma:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.emb.ma", + "source_data": { + "dataset_name": "EMB - M&A", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "ma", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (M&A).", + "metric_id": "vals_ai.emb.ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 44.407, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "ma", + "task_name": "M&A", + "dataset_type": "private", + "industry": "finance", + "raw_score": "44.407", + "raw_stderr": "8.687", + "latency": "1683.669", + "cost_per_test": "0.595991", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 8.687, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:misc:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.emb.misc", + "source_data": { + "dataset_name": "EMB - Misc", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "misc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Misc).", + "metric_id": "vals_ai.emb.misc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 43.153, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "misc", + "task_name": "Misc", + "dataset_type": "private", + "industry": "finance", + "raw_score": "43.153", + "raw_stderr": "7.697", + "latency": "1656.003", + "cost_per_test": "0.452349", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.697, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:operating:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.emb.operating", + "source_data": { + "dataset_name": "EMB - Operating", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "operating", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Operating).", + "metric_id": "vals_ai.emb.operating.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 36.655, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "operating", + "task_name": "Operating", + "dataset_type": "private", + "industry": "finance", + "raw_score": "36.655", + "raw_stderr": "7.534", + "latency": "1743.737", + "cost_per_test": "0.543688", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.534, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:overall:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.emb.overall", + "source_data": { + "dataset_name": "EMB - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Overall).", + "metric_id": "vals_ai.emb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 44.751, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "44.751", + "raw_stderr": "3.121", + "latency": "1767.043", + "cost_per_test": "0.57312", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.121, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:scratch:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.emb.scratch", + "source_data": { + "dataset_name": "EMB - Scratch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "scratch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Scratch).", + "metric_id": "vals_ai.emb.scratch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 43.805, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "scratch", + "task_name": "Scratch", + "dataset_type": "private", + "industry": "finance", + "raw_score": "43.805", + "raw_stderr": "3.691", + "latency": "1507.678", + "cost_per_test": "0.484426", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.691, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:template:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.emb.template", + "source_data": { + "dataset_name": "EMB - Template", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "template", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Template).", + "metric_id": "vals_ai.emb.template.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 45.052, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "template", + "task_name": "Template", + "dataset_type": "private", + "industry": "finance", + "raw_score": "45.052", + "raw_stderr": "4.661", + "latency": "2026.409", + "cost_per_test": "0.661814", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.661, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/2de45219-24f3-4cc8-a612-7efc5bff6fdc.json b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/2de45219-24f3-4cc8-a612-7efc5bff6fdc.json new file mode 100644 index 0000000000000000000000000000000000000000..87094c963669bbc89aa9b203af39a8b4bb08f234 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/2de45219-24f3-4cc8-a612-7efc5bff6fdc.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/openai_gpt-5.4-nano-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-nano-2026-03-17", + "id": "openai/gpt-5.4-nano-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-nano-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.373, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.373", + "raw_stderr": "1.67", + "latency": "8.67", + "cost_per_test": "0.000735", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.67, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 62.121, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.121", + "raw_stderr": "1.656", + "latency": "12.158", + "cost_per_test": "0.035271", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.656, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.189, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.189", + "raw_stderr": "0.961", + "latency": "9.397", + "cost_per_test": "0.01612", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.961, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.072, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.072", + "raw_stderr": "1.665", + "latency": "7.363", + "cost_per_test": "0.012353", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.665, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/3169aa2d-e128-4a13-bc6e-3da899deffe5.json b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/3169aa2d-e128-4a13-bc6e-3da899deffe5.json new file mode 100644 index 0000000000000000000000000000000000000000..270170fe16a930e48550a1bd730b7ea06bc26ced --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/3169aa2d-e128-4a13-bc6e-3da899deffe5.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/openai_gpt-5.4-nano-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-nano-2026-03-17", + "id": "openai/gpt-5.4-nano-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-nano-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 57.482, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.482", + "raw_stderr": "1.414", + "latency": "15.898", + "cost_per_test": "0.000638", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.414, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 67.416, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.416", + "raw_stderr": "0.926", + "latency": "15.898", + "cost_per_test": "0.000638", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.926, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 77.351, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "77.351", + "raw_stderr": "1.197", + "latency": "15.898", + "cost_per_test": "0.000638", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.197, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/371976e4-d30f-4585-b380-d4d1802f783b.json b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/371976e4-d30f-4585-b380-d4d1802f783b.json new file mode 100644 index 0000000000000000000000000000000000000000..9ada7cde67ac4c9eefba94afae86d969a84e1aa3 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/371976e4-d30f-4585-b380-d4d1802f783b.json @@ -0,0 +1,408 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/openai_gpt-5.4-nano-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-nano-2026-03-17", + "id": "openai/gpt-5.4-nano-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-nano-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 61.072, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "61.072", + "raw_stderr": "1.665", + "latency": "7.363", + "cost_per_test": "0.01612", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.665, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 38.236, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "38.236", + "raw_stderr": "0.381", + "latency": "326.257", + "cost_per_test": "0.152321", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.381, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:overall:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 46.635, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "46.635", + "raw_stderr": "1.866", + "latency": "929.692", + "cost_per_test": "0.325075", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.866, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 67.647, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "67.647", + "raw_stderr": "4.632", + "latency": "281.133", + "cost_per_test": "0.107316" + }, + "uncertainty": { + "standard_error": { + "value": 4.632, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 41.573, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "41.573", + "raw_stderr": "2.973", + "latency": "757.485", + "cost_per_test": "0.072321", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.973, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 30.032, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "30.032", + "raw_stderr": "8.282", + "latency": "3276.224", + "cost_per_test": "1.277298", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 8.282, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/3c2a5d2c-1382-43d1-a857-dd8a3a747176.json b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/3c2a5d2c-1382-43d1-a857-dd8a3a747176.json new file mode 100644 index 0000000000000000000000000000000000000000..5b79f8ee8dae9850e43a3cb078d51b8938965e37 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/3c2a5d2c-1382-43d1-a857-dd8a3a747176.json @@ -0,0 +1,103 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/openai_gpt-5.4-nano-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-nano-2026-03-17", + "id": "openai/gpt-5.4-nano-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-nano-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 73.584, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.584", + "raw_stderr": "1.06", + "latency": "21.907", + "cost_per_test": "0.000844", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.06, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/5402751b-9afc-4601-b30b-38cebeab762c.json b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/5402751b-9afc-4601-b30b-38cebeab762c.json new file mode 100644 index 0000000000000000000000000000000000000000..d3e5ae6ccdf6278264b66dab9051ecd85e6a57cf --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/5402751b-9afc-4601-b30b-38cebeab762c.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/openai_gpt-5.4-nano-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-nano-2026-03-17", + "id": "openai/gpt-5.4-nano-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-nano-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 68.124, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.124", + "raw_stderr": "1.314", + "latency": "18.86", + "cost_per_test": "0.003092", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.314, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 59.102, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.102", + "raw_stderr": "0.964", + "latency": "14.526", + "cost_per_test": "0.002585", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.964, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 50.079, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "50.079", + "raw_stderr": "1.41", + "latency": "10.193", + "cost_per_test": "0.002079", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.41, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/5c435efe-c59f-4478-9e8f-52b39d755f9b.json b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/5c435efe-c59f-4478-9e8f-52b39d755f9b.json new file mode 100644 index 0000000000000000000000000000000000000000..aa288783ca1b1d07c0c93fd10b0fd2124d441545 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/5c435efe-c59f-4478-9e8f-52b39d755f9b.json @@ -0,0 +1,356 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/openai_gpt-5.4-nano-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-nano-2026-03-17", + "id": "openai/gpt-5.4-nano-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-nano-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 30.159, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "30.159", + "raw_stderr": "5.772", + "latency": "37.818", + "cost_per_test": "0.007019", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.772, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:automata:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 23.412, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "23.412", + "raw_stderr": "4.736", + "latency": "38.242", + "cost_per_test": "0.004953", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.736, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:calculus:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 20.988, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "20.988", + "raw_stderr": "4.183", + "latency": "45.07", + "cost_per_test": "0.002745", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.183, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 60.671, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "60.671", + "raw_stderr": "5.477", + "latency": "26.941", + "cost_per_test": "0.005836", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.477, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:overall:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 38.081, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "38.081", + "raw_stderr": "3.087", + "latency": "34.334", + "cost_per_test": "0.005936", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.087, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/86da2806-d863-4d48-ab04-c2c843a6cdf7.json b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/86da2806-d863-4d48-ab04-c2c843a6cdf7.json new file mode 100644 index 0000000000000000000000000000000000000000..235ee124e181afd7d77b1b71fe536df14f9b6c78 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/86da2806-d863-4d48-ab04-c2c843a6cdf7.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/openai_gpt-5.4-nano-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-nano-2026-03-17", + "id": "openai/gpt-5.4-nano-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-nano-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 5.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "5", + "raw_stderr": "2.179", + "latency": "525.432", + "cost_per_test": "0.015894", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.179, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/92a4baae-09ed-4c22-aefa-c4d3931a5340.json b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/92a4baae-09ed-4c22-aefa-c4d3931a5340.json new file mode 100644 index 0000000000000000000000000000000000000000..7b39b0f9f7de19110e681265e583d211df4f926b --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/92a4baae-09ed-4c22-aefa-c4d3931a5340.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/openai_gpt-5.4-nano-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-nano-2026-03-17", + "id": "openai/gpt-5.4-nano-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-nano-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 97.516, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "97.516", + "raw_stderr": "0.867", + "latency": "8.932", + "cost_per_test": "0.000242", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.867, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:hard:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 66.0, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66", + "raw_stderr": "2.532", + "latency": "132.769", + "cost_per_test": "0.00563", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.532, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:medium:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 88.512, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "88.512", + "raw_stderr": "1.629", + "latency": "38.567", + "cost_per_test": "0.001531", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.629, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:overall:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 84.009, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "84.009", + "raw_stderr": "1.044", + "latency": "60.089", + "cost_per_test": "0.002468", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.044, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/c451ed74-79f9-4250-8428-1ed0641818b5.json b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/c451ed74-79f9-4250-8428-1ed0641818b5.json new file mode 100644 index 0000000000000000000000000000000000000000..8a2dabce73b5ffed43822a69b7e6166244be5b64 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/c451ed74-79f9-4250-8428-1ed0641818b5.json @@ -0,0 +1,371 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/openai_gpt-5.4-nano-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-nano-2026-03-17", + "id": "openai/gpt-5.4-nano-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-nano-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "7.362", + "latency": "365.183", + "cost_per_test": "0.154669", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.362, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 67.816, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "67.816", + "raw_stderr": "2.897", + "latency": "282.043", + "cost_per_test": "0.106039", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.897, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 80.928, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "80.928", + "raw_stderr": "2.828", + "latency": "213.858", + "cost_per_test": "0.07376", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.828, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "278.84", + "cost_per_test": "0.124824", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:overall:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 69.8, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "69.8", + "raw_stderr": "2.055", + "latency": "262.552", + "cost_per_test": "0.097712", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.055, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/cf4b2ed6-96fd-4991-a183-3032e8bd1582.json b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/cf4b2ed6-96fd-4991-a183-3032e8bd1582.json new file mode 100644 index 0000000000000000000000000000000000000000..0dda1812cf3f8571ecf5645c8008a1489ffbf32f --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/cf4b2ed6-96fd-4991-a183-3032e8bd1582.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/openai_gpt-5.4-nano-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-nano-2026-03-17", + "id": "openai/gpt-5.4-nano-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-nano-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.89, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.89", + "raw_stderr": "1.601", + "latency": "5.551", + "cost_per_test": "0.000242", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.601, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.678, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.678", + "raw_stderr": "1.445", + "latency": "9.108", + "cost_per_test": "0.000415", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.445, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.014, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.014", + "raw_stderr": "1.122", + "latency": "9.297", + "cost_per_test": "0.00054", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.122, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.634, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.634", + "raw_stderr": "1.899", + "latency": "6.365", + "cost_per_test": "0.000369", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.899, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.161, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.161", + "raw_stderr": "1.413", + "latency": "6.062", + "cost_per_test": "0.000222", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.413, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.271, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.271", + "raw_stderr": "1.465", + "latency": "26.91", + "cost_per_test": "0.001399", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.465, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 72.127, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.127", + "raw_stderr": "1.674", + "latency": "5.585", + "cost_per_test": "0.000201", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.674, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 59.318, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "59.318", + "raw_stderr": "2.54", + "latency": "4.698", + "cost_per_test": "0.000241", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.54, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 58.856, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "58.856", + "raw_stderr": "1.493", + "latency": "7.031", + "cost_per_test": "0.000369", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.493, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.896, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.896", + "raw_stderr": "0.886", + "latency": "9.175", + "cost_per_test": "0.000463", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.886, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.861, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.861", + "raw_stderr": "1.557", + "latency": "5.742", + "cost_per_test": "0.000192", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.557, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.172, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.172", + "raw_stderr": "0.433", + "latency": "8.21", + "cost_per_test": "0.000396", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.433, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 66.132, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.132", + "raw_stderr": "2.172", + "latency": "5.847", + "cost_per_test": "0.00023", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.172, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.76, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.76", + "raw_stderr": "0.974", + "latency": "10.012", + "cost_per_test": "0.000492", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.974, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.812, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.812", + "raw_stderr": "1.629", + "latency": "3.55", + "cost_per_test": "0.000169", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.629, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/d6c12f09-1ec3-41e1-9547-3a00170a7f0f.json b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/d6c12f09-1ec3-41e1-9547-3a00170a7f0f.json new file mode 100644 index 0000000000000000000000000000000000000000..4cfa8a7e5c15bcfcb0b95e0040f1dbb8b7d66cf2 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/d6c12f09-1ec3-41e1-9547-3a00170a7f0f.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/openai_gpt-5.4-nano-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-nano-2026-03-17", + "id": "openai/gpt-5.4-nano-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-nano-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 76.768, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.768", + "raw_stderr": "3.408", + "latency": "31.524", + "cost_per_test": "0.001658", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.408, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 78.283, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.283", + "raw_stderr": "3.429", + "latency": "27.473", + "cost_per_test": "0.001472", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.429, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 77.526, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.526", + "raw_stderr": "2.417", + "latency": "29.498", + "cost_per_test": "0.001565", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.417, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/e0a38db1-9521-4cd5-a853-3b541171b827.json b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/e0a38db1-9521-4cd5-a853-3b541171b827.json new file mode 100644 index 0000000000000000000000000000000000000000..c51a50dfb198e9ed885425e267c6169e7307ebd6 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/e0a38db1-9521-4cd5-a853-3b541171b827.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/openai_gpt-5.4-nano-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-nano-2026-03-17", + "id": "openai/gpt-5.4-nano-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-nano-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 77.09, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "77.09", + "raw_stderr": "1.891", + "latency": "20.527", + "cost_per_test": "0.0018", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.891, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/e2ce95d3-f223-4398-beeb-968948445f16.json b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/e2ce95d3-f223-4398-beeb-968948445f16.json new file mode 100644 index 0000000000000000000000000000000000000000..bf603fd6b0d1f0126c9fdf722569349a38913050 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/e2ce95d3-f223-4398-beeb-968948445f16.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/openai_gpt-5.4-nano-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-nano-2026-03-17", + "id": "openai/gpt-5.4-nano-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-nano-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 26.097, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "26.097", + "raw_stderr": "5.075", + "latency": "3276.224", + "cost_per_test": "1.277298", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.075, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/ec91cf91-8ba6-4321-90ee-e3a14a79c74f.json b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/ec91cf91-8ba6-4321-90ee-e3a14a79c74f.json new file mode 100644 index 0000000000000000000000000000000000000000..d43489e43bf9b929b392dd39f25581023b9b3823 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/ec91cf91-8ba6-4321-90ee-e3a14a79c74f.json @@ -0,0 +1,438 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/openai_gpt-5.4-nano-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-nano-2026-03-17", + "id": "openai/gpt-5.4-nano-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-nano-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 84.203, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.203", + "raw_stderr": "0.838", + "latency": "2.749", + "cost_per_test": "0.000121", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.838, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 69.242, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "69.242", + "raw_stderr": "0.266", + "latency": "2.539", + "cost_per_test": "0.00011", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.266, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.807, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.807", + "raw_stderr": "0.654", + "latency": "2.274", + "cost_per_test": "0.000139", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.654, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.92, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.92", + "raw_stderr": "0.427", + "latency": "2.697", + "cost_per_test": "0.000114", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.427, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.085, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.085", + "raw_stderr": "0.925", + "latency": "2.308", + "cost_per_test": "8.8e-05", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.925, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 76.263, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "76.263", + "raw_stderr": "1.585", + "latency": "3.614", + "cost_per_test": "0.000114", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.585, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/f21d0f7c-f47a-4f25-9943-178c9d3c49ba.json b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/f21d0f7c-f47a-4f25-9943-178c9d3c49ba.json new file mode 100644 index 0000000000000000000000000000000000000000..9ab1d7d05a966145b85c8cdc97d2148e6c5c52ee --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4-nano-2026-03-17/f21d0f7c-f47a-4f25-9943-178c9d3c49ba.json @@ -0,0 +1,740 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/openai_gpt-5.4-nano-2026-03-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4-nano-2026-03-17", + "id": "openai/gpt-5.4-nano-2026-03-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4-nano-2026-03-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 32.97, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "32.97", + "raw_stderr": "0.667", + "latency": "360.491", + "cost_per_test": "0.168389", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.667, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 26.022, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "26.022", + "raw_stderr": "0.989", + "latency": "334.538", + "cost_per_test": "0.15925", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.989, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 37.173, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "37.173", + "raw_stderr": "0.672", + "latency": "366.547", + "cost_per_test": "0.207015", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.672, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 42.83, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "42.83", + "raw_stderr": "3.444", + "latency": "356.213", + "cost_per_test": "0.183174", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.444, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 45.604, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "45.604", + "raw_stderr": "1.629", + "latency": "239.547", + "cost_per_test": "0.07704", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.629, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 13.49, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "13.49", + "raw_stderr": "1.455", + "latency": "368.43", + "cost_per_test": "0.125942", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.455, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 46.286, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "46.286", + "raw_stderr": "2.19", + "latency": "344.238", + "cost_per_test": "0.160073", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.19, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 58.313, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.313", + "raw_stderr": "1.54", + "latency": "268.192", + "cost_per_test": "0.156651", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.54, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 46.461, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "46.461", + "raw_stderr": "3.042", + "latency": "237.62", + "cost_per_test": "0.088647", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.042, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:overall:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 38.217, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "38.217", + "raw_stderr": "1.185", + "latency": "334.538", + "cost_per_test": "0.15925", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.185, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:openai/gpt-5.4-nano-2026-03-17:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 20.83, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "20.83", + "raw_stderr": "1.51", + "latency": "532.115", + "cost_per_test": "0.276849", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.51, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.4/ad030b40-250e-496b-a6ef-0492417a66f9.json b/data/vals_ai/openai/gpt-5.4/ad030b40-250e-496b-a6ef-0492417a66f9.json new file mode 100644 index 0000000000000000000000000000000000000000..e6df1ba4b741e92a206f3cd95de76cf0bf38f893 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.4/ad030b40-250e-496b-a6ef-0492417a66f9.json @@ -0,0 +1,89 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/openai_gpt-5.4/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.4", + "id": "openai/gpt-5.4", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.4", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:openai/gpt-5.4:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 48.474, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "48.474", + "raw_stderr": "5.07", + "latency": "1075.396", + "cost_per_test": "7.50518", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.07, + "method": "vals_reported" + } + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.5-codex/4e8326b1-6d30-4835-9103-3a5fa770cd14.json b/data/vals_ai/openai/gpt-5.5-codex/4e8326b1-6d30-4835-9103-3a5fa770cd14.json new file mode 100644 index 0000000000000000000000000000000000000000..4f92f33fca589f3ef495f18c64b552e0838aa2f7 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.5-codex/4e8326b1-6d30-4835-9103-3a5fa770cd14.json @@ -0,0 +1,316 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/openai_gpt-5.5-codex/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.5-codex", + "id": "openai/gpt-5.5-codex", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.5-codex", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:openai/gpt-5.5-codex:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 50.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "50", + "raw_stderr": "7.809", + "latency": "170.104", + "cost_per_test": "1.035674", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.809, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:openai/gpt-5.5-codex:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 74.33, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "74.33", + "raw_stderr": "2.709", + "latency": "117.239", + "cost_per_test": "0.684023", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.709, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:openai/gpt-5.5-codex:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 85.567, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "85.567", + "raw_stderr": "2.53", + "latency": "82.224", + "cost_per_test": "0.520143", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.53, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:openai/gpt-5.5-codex:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "192.303", + "cost_per_test": "1.076622", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "swebench:overall:openai/gpt-5.5-codex:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 76.4, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "76.4", + "raw_stderr": "1.901", + "latency": "108.544", + "cost_per_test": "0.652332", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.901, + "method": "vals_reported" + } + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.5-codex/71fd5d0b-4647-48ff-972e-ae957d595db0.json b/data/vals_ai/openai/gpt-5.5-codex/71fd5d0b-4647-48ff-972e-ae957d595db0.json new file mode 100644 index 0000000000000000000000000000000000000000..a639780f2b0414cba3456f007df181ead57cfdf2 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.5-codex/71fd5d0b-4647-48ff-972e-ae957d595db0.json @@ -0,0 +1,260 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/openai_gpt-5.5-codex/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.5-codex", + "id": "openai/gpt-5.5-codex", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.5-codex", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:openai/gpt-5.5-codex:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 75.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "75", + "raw_stderr": "25", + "latency": "4504.93", + "cost_per_test": "14.356089", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 25.0, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:openai/gpt-5.5-codex:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 43.333, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "43.333", + "raw_stderr": "9.202", + "latency": "600.657", + "cost_per_test": "1.914145", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 9.202, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:openai/gpt-5.5-codex:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 63.636, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "63.636", + "raw_stderr": "6.546", + "latency": "327.631", + "cost_per_test": "1.044079", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.546, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:openai/gpt-5.5-codex:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 57.303, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "57.303", + "raw_stderr": "5.273", + "latency": "202.469", + "cost_per_test": "0.645217", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.273, + "method": "vals_reported" + } + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.5-codex/d3ceaad1-b232-4c44-a042-b50afccc9f67.json b/data/vals_ai/openai/gpt-5.5-codex/d3ceaad1-b232-4c44-a042-b50afccc9f67.json new file mode 100644 index 0000000000000000000000000000000000000000..117b1b9852e89ba2f2158e299db7a48b9fa9c64f --- /dev/null +++ b/data/vals_ai/openai/gpt-5.5-codex/d3ceaad1-b232-4c44-a042-b50afccc9f67.json @@ -0,0 +1,89 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/openai_gpt-5.5-codex/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.5-codex", + "id": "openai/gpt-5.5-codex", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.5-codex", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:openai/gpt-5.5-codex:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 58.194, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "58.194", + "raw_stderr": "4.933", + "latency": "708.51", + "cost_per_test": "3.6264", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.933, + "method": "vals_reported" + } + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.5-factory/431a3ded-8ecb-4e4a-8bf8-42852f01d515.json b/data/vals_ai/openai/gpt-5.5-factory/431a3ded-8ecb-4e4a-8bf8-42852f01d515.json new file mode 100644 index 0000000000000000000000000000000000000000..631225721d65080b70c3f6b32d23c37132dd3f97 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.5-factory/431a3ded-8ecb-4e4a-8bf8-42852f01d515.json @@ -0,0 +1,89 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/openai_gpt-5.5-factory/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.5-factory", + "id": "openai/gpt-5.5-factory", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.5-factory", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:openai/gpt-5.5-factory:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 67.391, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "67.391", + "raw_stderr": "4.75", + "latency": "1177.43", + "cost_per_test": "5.9268", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.75, + "method": "vals_reported" + } + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.5-factory/8fbe8be3-4bfa-4cd3-8992-5c48b00ce79c.json b/data/vals_ai/openai/gpt-5.5-factory/8fbe8be3-4bfa-4cd3-8992-5c48b00ce79c.json new file mode 100644 index 0000000000000000000000000000000000000000..991980a8e4b2d49506db46a31556f863106747ba --- /dev/null +++ b/data/vals_ai/openai/gpt-5.5-factory/8fbe8be3-4bfa-4cd3-8992-5c48b00ce79c.json @@ -0,0 +1,316 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/openai_gpt-5.5-factory/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.5-factory", + "id": "openai/gpt-5.5-factory", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.5-factory", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:openai/gpt-5.5-factory:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 47.619, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "47.619", + "raw_stderr": "7.8", + "latency": "529.492", + "cost_per_test": "1.845651", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.8, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:openai/gpt-5.5-factory:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 73.946, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "73.946", + "raw_stderr": "2.722", + "latency": "421.746", + "cost_per_test": "1.14672", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.722, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:openai/gpt-5.5-factory:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 86.598, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "86.598", + "raw_stderr": "2.452", + "latency": "367.734", + "cost_per_test": "0.94393", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.452, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:openai/gpt-5.5-factory:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "473.047", + "cost_per_test": "1.569327", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "swebench:overall:openai/gpt-5.5-factory:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 76.2, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "76.2", + "raw_stderr": "1.906", + "latency": "410.148", + "cost_per_test": "1.129284", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.906, + "method": "vals_reported" + } + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.5-factory/976aceb9-ca00-4cac-ac7d-f2f847642703.json b/data/vals_ai/openai/gpt-5.5-factory/976aceb9-ca00-4cac-ac7d-f2f847642703.json new file mode 100644 index 0000000000000000000000000000000000000000..cb47f84055321bd8920a2a57de1319ae73caae4e --- /dev/null +++ b/data/vals_ai/openai/gpt-5.5-factory/976aceb9-ca00-4cac-ac7d-f2f847642703.json @@ -0,0 +1,260 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/openai_gpt-5.5-factory/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.5-factory", + "id": "openai/gpt-5.5-factory", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.5-factory", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:openai/gpt-5.5-factory:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 75.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "75", + "raw_stderr": "25", + "latency": "8661.154", + "cost_per_test": "16.721574", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 25.0, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:openai/gpt-5.5-factory:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 40.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "40", + "raw_stderr": "9.097", + "latency": "1154.821", + "cost_per_test": "2.229543", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 9.097, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:openai/gpt-5.5-factory:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 65.455, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "65.455", + "raw_stderr": "6.471", + "latency": "629.902", + "cost_per_test": "1.216114", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.471, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:openai/gpt-5.5-factory:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 57.303, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "57.303", + "raw_stderr": "5.273", + "latency": "389.265", + "cost_per_test": "0.751531", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.273, + "method": "vals_reported" + } + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.5/13a7873f-9d85-4725-906f-09fe7826c796.json b/data/vals_ai/openai/gpt-5.5/13a7873f-9d85-4725-906f-09fe7826c796.json new file mode 100644 index 0000000000000000000000000000000000000000..dc9576dc472b255d88a8081d2f36e3f1897c656e --- /dev/null +++ b/data/vals_ai/openai/gpt-5.5/13a7873f-9d85-4725-906f-09fe7826c796.json @@ -0,0 +1,371 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/openai_gpt-5.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.5", + "id": "openai/gpt-5.5", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.5", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 50.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "50", + "raw_stderr": "7.809", + "latency": "650.107", + "cost_per_test": "1.502703", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.809, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 81.226, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "81.226", + "raw_stderr": "2.422", + "latency": "454.903", + "cost_per_test": "0.868948", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.422, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 91.753, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "91.753", + "raw_stderr": "1.98", + "latency": "338.869", + "cost_per_test": "0.588097", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.98, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66.667", + "raw_stderr": "33.333", + "latency": "479.329", + "cost_per_test": "1.509878", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "swebench:overall:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 82.6, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "82.6", + "raw_stderr": "1.695", + "latency": "426.425", + "cost_per_test": "1.361765", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.695, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.5/278d220a-75b3-4ce8-bf21-6e28abb7ce68.json b/data/vals_ai/openai/gpt-5.5/278d220a-75b3-4ce8-bf21-6e28abb7ce68.json new file mode 100644 index 0000000000000000000000000000000000000000..baa6b51166865f73041300cbae260c8c22514450 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.5/278d220a-75b3-4ce8-bf21-6e28abb7ce68.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/public-benefits-bench/openai_gpt-5.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Public Benefits Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "social mobility", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.5", + "id": "openai/gpt-5.5", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.5", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "public-benefits-bench:overall:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.public-benefits-bench.overall", + "source_data": { + "dataset_name": "Public Benefits Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Public Benefits Bench v1.1 (Overall).", + "metric_id": "vals_ai.public-benefits-bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "score_details": { + "score": 60.893, + "details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "social mobility", + "raw_score": "60.893", + "raw_stderr": "1.269", + "latency": "2512.168", + "cost_per_test": "3.971165", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.269, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.5/32ff65c0-1719-463d-a2e4-6c587550d70f.json b/data/vals_ai/openai/gpt-5.5/32ff65c0-1719-463d-a2e4-6c587550d70f.json new file mode 100644 index 0000000000000000000000000000000000000000..b21fdf6cf0cb73b16f775be4a2da7a38000d50f6 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.5/32ff65c0-1719-463d-a2e4-6c587550d70f.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/openai_gpt-5.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.5", + "id": "openai/gpt-5.5", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.5", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.003, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.003", + "raw_stderr": "0.887", + "latency": "24.475", + "cost_per_test": "0.018287", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.887, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.015, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.015", + "raw_stderr": "0.965", + "latency": "35.966", + "cost_per_test": "0.021982", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.965, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.254, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.254", + "raw_stderr": "0.84", + "latency": "54.56", + "cost_per_test": "0.028489", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.84, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.756, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.756", + "raw_stderr": "1.498", + "latency": "38.434", + "cost_per_test": "0.027059", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.498, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.929, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.929", + "raw_stderr": "1.036", + "latency": "28.846", + "cost_per_test": "0.01703", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.036, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.83, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.83", + "raw_stderr": "1.152", + "latency": "79.599", + "cost_per_test": "0.062984", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.152, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.352, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.352", + "raw_stderr": "1.27", + "latency": "35.101", + "cost_per_test": "0.026769", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.27, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.627, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.627", + "raw_stderr": "1.984", + "latency": "58.094", + "cost_per_test": "0.025591", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.984, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.93, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.93", + "raw_stderr": "1.27", + "latency": "68.6", + "cost_per_test": "0.053088", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.27, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.745, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.745", + "raw_stderr": "0.607", + "latency": "33.027", + "cost_per_test": "0.017893", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.607, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.797, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.797", + "raw_stderr": "1.114", + "latency": "33.861", + "cost_per_test": "0.023843", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.114, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.144, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.144", + "raw_stderr": "0.318", + "latency": "42.114", + "cost_per_test": "0.027891", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.318, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.776, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.776", + "raw_stderr": "1.466", + "latency": "34.673", + "cost_per_test": "0.025903", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.466, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.532, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.532", + "raw_stderr": "0.772", + "latency": "40.216", + "cost_per_test": "0.024257", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.772, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.471, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.471", + "raw_stderr": "1.131", + "latency": "24.149", + "cost_per_test": "0.017305", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.131, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.5/358307e5-b14c-4165-bbbd-77db59dda4fd.json b/data/vals_ai/openai/gpt-5.5/358307e5-b14c-4165-bbbd-77db59dda4fd.json new file mode 100644 index 0000000000000000000000000000000000000000..4878b3b3d9bcd8fe64998db1b718a7d0241eef77 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.5/358307e5-b14c-4165-bbbd-77db59dda4fd.json @@ -0,0 +1,740 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/openai_gpt-5.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.5", + "id": "openai/gpt-5.5", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.5", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 45.74, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "45.74", + "raw_stderr": "1.121", + "latency": "651.473", + "cost_per_test": "3.890026", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.121, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 39.562, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "39.562", + "raw_stderr": "0.865", + "latency": "662.269", + "cost_per_test": "4.149702", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.865, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 43.653, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "43.653", + "raw_stderr": "0.5", + "latency": "798.364", + "cost_per_test": "5.684233", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.5, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 58.909, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.909", + "raw_stderr": "0.871", + "latency": "792.143", + "cost_per_test": "4.884586", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.871, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 73.158, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "73.158", + "raw_stderr": "0", + "latency": "381.734", + "cost_per_test": "1.633725", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 22.93, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "22.93", + "raw_stderr": "2.11", + "latency": "592.855", + "cost_per_test": "3.178322", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.11, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 65.559, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.559", + "raw_stderr": "1.39", + "latency": "717.08", + "cost_per_test": "4.530392", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.39, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 73.714, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "73.714", + "raw_stderr": "1.754", + "latency": "485.345", + "cost_per_test": "3.745478", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.754, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 60.288, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.288", + "raw_stderr": "1.614", + "latency": "498.931", + "cost_per_test": "2.721406", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.614, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:overall:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 51.76, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "51.76", + "raw_stderr": "0.55", + "latency": "662.269", + "cost_per_test": "4.149702", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.55, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 21.894, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "21.894", + "raw_stderr": "1.895", + "latency": "1110.441", + "cost_per_test": "7.463278", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.895, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.5/36899f22-3219-47f8-b908-17968f260541.json b/data/vals_ai/openai/gpt-5.5/36899f22-3219-47f8-b908-17968f260541.json new file mode 100644 index 0000000000000000000000000000000000000000..02e9198db8c899a350fbb12a51a9974c2196f49a --- /dev/null +++ b/data/vals_ai/openai/gpt-5.5/36899f22-3219-47f8-b908-17968f260541.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/openai_gpt-5.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.5", + "id": "openai/gpt-5.5", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.5", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 83.333, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "83.333", + "raw_stderr": "8.333", + "latency": "9509.922", + "cost_per_test": "16.506894", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 8.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66.667", + "raw_stderr": "1.925", + "latency": "1267.99", + "cost_per_test": "2.200919", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.925, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 81.212, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "81.212", + "raw_stderr": "0.606", + "latency": "691.631", + "cost_per_test": "1.200501", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.606, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 76.404, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "76.404", + "raw_stderr": "0.649", + "latency": "427.412", + "cost_per_test": "0.741883", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.649, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.5/4797d264-43ba-4a4e-8298-ddee27eb1dd9.json b/data/vals_ai/openai/gpt-5.5/4797d264-43ba-4a4e-8298-ddee27eb1dd9.json new file mode 100644 index 0000000000000000000000000000000000000000..d6ac3ff362f6f5a7dde4868b5e605a880a6dc183 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.5/4797d264-43ba-4a4e-8298-ddee27eb1dd9.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/openai_gpt-5.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.5", + "id": "openai/gpt-5.5", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.5", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 50.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "50", + "raw_stderr": "5", + "latency": "2953.443", + "cost_per_test": "1.68179", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.5/54a6ee76-1d34-49ce-8003-cd189fed82a1.json b/data/vals_ai/openai/gpt-5.5/54a6ee76-1d34-49ce-8003-cd189fed82a1.json new file mode 100644 index 0000000000000000000000000000000000000000..e597e03a4410babbcc6babe7a730f6d58678a88e --- /dev/null +++ b/data/vals_ai/openai/gpt-5.5/54a6ee76-1d34-49ce-8003-cd189fed82a1.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/openai_gpt-5.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.5", + "id": "openai/gpt-5.5", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.5", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.447, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.447", + "raw_stderr": "0.689", + "latency": "48.393", + "cost_per_test": "0.012472", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.689, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "lcb:hard:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 69.714, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "69.714", + "raw_stderr": "2.456", + "latency": "353.98", + "cost_per_test": "0.151373", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.456, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "lcb:medium:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 87.728, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "87.728", + "raw_stderr": "1.677", + "latency": "98.163", + "cost_per_test": "0.035956", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.677, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "lcb:overall:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 85.296, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "85.296", + "raw_stderr": "1.018", + "latency": "166.845", + "cost_per_test": "0.0666", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.018, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.5/64956078-fd4e-4f4f-868c-659c7659c5bd.json b/data/vals_ai/openai/gpt-5.5/64956078-fd4e-4f4f-868c-659c7659c5bd.json new file mode 100644 index 0000000000000000000000000000000000000000..86fc9b5d6f9fe7316dc8cd83e1c9f74146b756ee --- /dev/null +++ b/data/vals_ai/openai/gpt-5.5/64956078-fd4e-4f4f-868c-659c7659c5bd.json @@ -0,0 +1,408 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/openai_gpt-5.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.5", + "id": "openai/gpt-5.5", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.5", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 69.58, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "69.58", + "raw_stderr": "1.571", + "latency": "26.764", + "cost_per_test": "0.43166", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.571, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 51.834, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "51.834", + "raw_stderr": "0.461", + "latency": "616.56", + "cost_per_test": "3.784562", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.461, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_index:overall:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 67.951, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "67.951", + "raw_stderr": "1.536", + "latency": "681.746", + "cost_per_test": "4.596356", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.536, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 82.6, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "82.6", + "raw_stderr": "1.695", + "latency": "426.425", + "cost_per_test": "1.361765" + }, + "uncertainty": { + "standard_error": { + "value": 1.695, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 76.404, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "76.404", + "raw_stderr": "0.649", + "latency": "427.412", + "cost_per_test": "0.741883", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.649, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 77.099, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "77.099", + "raw_stderr": "7.028", + "latency": "1911.567", + "cost_per_test": "16.661911", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.028, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.5/6819edf1-4f67-4219-bb6d-835b820a10e3.json b/data/vals_ai/openai/gpt-5.5/6819edf1-4f67-4219-bb6d-835b820a10e3.json new file mode 100644 index 0000000000000000000000000000000000000000..c31f28e09306b046a29b241932067be6d9dca27b --- /dev/null +++ b/data/vals_ai/openai/gpt-5.5/6819edf1-4f67-4219-bb6d-835b820a10e3.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/openai_gpt-5.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.5", + "id": "openai/gpt-5.5", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.5", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 69.847, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "69.847", + "raw_stderr": "4.535", + "latency": "1911.567", + "cost_per_test": "16.661911", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.5/74ce4041-822d-4000-947e-47e1db113ebf.json b/data/vals_ai/openai/gpt-5.5/74ce4041-822d-4000-947e-47e1db113ebf.json new file mode 100644 index 0000000000000000000000000000000000000000..e277401736207747fb7cb1e0e1c7da5881166d77 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.5/74ce4041-822d-4000-947e-47e1db113ebf.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/openai_gpt-5.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.5", + "id": "openai/gpt-5.5", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.5", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 65.618, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.618", + "raw_stderr": "1.622", + "latency": "17.821", + "cost_per_test": "0.02963", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.622, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 70.047, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "70.047", + "raw_stderr": "1.564", + "latency": "25.327", + "cost_per_test": "0.902658", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.564, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 68.415, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.415", + "raw_stderr": "0.916", + "latency": "23.304", + "cost_per_test": "0.43166", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.916, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 69.58, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "69.58", + "raw_stderr": "1.571", + "latency": "26.764", + "cost_per_test": "0.362691", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.571, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.5/7514e5b6-0c27-4127-9717-68dff19bb042.json b/data/vals_ai/openai/gpt-5.5/7514e5b6-0c27-4127-9717-68dff19bb042.json new file mode 100644 index 0000000000000000000000000000000000000000..035f3dbafc46367dadee2cdbc748ba31f656b6bf --- /dev/null +++ b/data/vals_ai/openai/gpt-5.5/7514e5b6-0c27-4127-9717-68dff19bb042.json @@ -0,0 +1,3386 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/hlab/openai_gpt-5.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Harvey's Legal Agent Benchmark", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.5", + "id": "openai/gpt-5.5", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.5", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "hlab:antitrust_competition:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "669.614", + "cost_per_test": "3.748755", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:arbitration_international_dispute_resolution:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "679.785", + "cost_per_test": "7.105465", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:banking_finance:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "632.111", + "cost_per_test": "3.453057", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:bankruptcy_restructuring:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "738.645", + "cost_per_test": "4.430016", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:capital_markets:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "578.203", + "cost_per_test": "3.800605", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:corporate_governance:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "757.959", + "cost_per_test": "6.425704", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:corporate_ma:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "803.617", + "cost_per_test": "3.395472", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Criteria Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Criteria Pass Rate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 79.983, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate", + "task_name": "Criteria Pass Rate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.983", + "raw_stderr": "0.481", + "latency": "733.852", + "cost_per_test": "4.597894", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.481, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__antitrust_competition:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 82.246, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.246", + "raw_stderr": "2.304", + "latency": "669.614", + "cost_per_test": "3.748755", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.304, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__arbitration_international_dispute_resolution:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 78.495, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.495", + "raw_stderr": "2.464", + "latency": "679.785", + "cost_per_test": "7.105465", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.464, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__banking_finance:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 86.245, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.245", + "raw_stderr": "2.104", + "latency": "632.111", + "cost_per_test": "3.453057", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.104, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__bankruptcy_restructuring:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 76.125, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "76.125", + "raw_stderr": "2.512", + "latency": "738.645", + "cost_per_test": "4.430016", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.512, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__capital_markets:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 73.726, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "73.726", + "raw_stderr": "2.762", + "latency": "578.203", + "cost_per_test": "3.800605", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.762, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_governance:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 78.425, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.425", + "raw_stderr": "2.411", + "latency": "757.959", + "cost_per_test": "6.425704", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.411, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_ma:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 87.5, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.5", + "raw_stderr": "1.875", + "latency": "803.617", + "cost_per_test": "3.395472", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.875, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__data_privacy_cybersecurity:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 92.226, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "92.226", + "raw_stderr": "1.594", + "latency": "641.518", + "cost_per_test": "3.983816", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.594, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__emerging_companies_venture_capital:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 61.775, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "61.775", + "raw_stderr": "2.844", + "latency": "662.519", + "cost_per_test": "2.900873", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.844, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__employment_labor:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 76.404, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "76.404", + "raw_stderr": "2.603", + "latency": "554.41", + "cost_per_test": "2.284177", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.603, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__energy_natural_resources:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 93.729, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "93.729", + "raw_stderr": "1.395", + "latency": "1415.764", + "cost_per_test": "11.82685", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.395, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__environmental_esg:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 82.079, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.079", + "raw_stderr": "2.3", + "latency": "762.873", + "cost_per_test": "3.081603", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.3, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__funds_asset_management:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 77.551, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.551", + "raw_stderr": "2.256", + "latency": "1206.377", + "cost_per_test": "5.020577", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.256, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__healthcare_life_sciences:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 81.848, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.848", + "raw_stderr": "2.218", + "latency": "812.897", + "cost_per_test": "4.010513", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.218, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__immigration:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 81.818, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.818", + "raw_stderr": "2.33", + "latency": "682.579", + "cost_per_test": "2.985935", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.33, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__insurance:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 82.784, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.784", + "raw_stderr": "2.289", + "latency": "631.639", + "cost_per_test": "2.417604", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.289, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__intellectual_property:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 86.572, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.572", + "raw_stderr": "2.03", + "latency": "628.011", + "cost_per_test": "3.636776", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.03, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__international_trade_sanctions:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 80.952, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.952", + "raw_stderr": "2.479", + "latency": "636.4", + "cost_per_test": "2.359423", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.479, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__litigation_dispute_resolution:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 81.648, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.648", + "raw_stderr": "2.373", + "latency": "596.275", + "cost_per_test": "6.932258", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.373, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__real_estate:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 92.711, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "92.711", + "raw_stderr": "1.406", + "latency": "799.058", + "cost_per_test": "5.565429", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.406, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__structured_finance_securitization:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 76.508, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "76.508", + "raw_stderr": "2.392", + "latency": "794.392", + "cost_per_test": "7.609913", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.392, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__tax:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 72.575, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "72.575", + "raw_stderr": "2.584", + "latency": "671.392", + "cost_per_test": "4.83132", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.584, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__trusts_estates_private_client:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 67.557, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "67.557", + "raw_stderr": "2.898", + "latency": "684.418", + "cost_per_test": "4.928295", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.898, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__white_collar_defense_investigations:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 65.32, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "65.32", + "raw_stderr": "2.766", + "latency": "571.985", + "cost_per_test": "3.615019", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.766, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:data_privacy_cybersecurity:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "641.518", + "cost_per_test": "3.983816", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:emerging_companies_venture_capital:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "662.519", + "cost_per_test": "2.900873", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:employment_labor:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "554.41", + "cost_per_test": "2.284177", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:energy_natural_resources:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "1415.764", + "cost_per_test": "11.82685", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:environmental_esg:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "762.873", + "cost_per_test": "3.081603", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:funds_asset_management:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1206.377", + "cost_per_test": "5.020577", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:healthcare_life_sciences:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "812.897", + "cost_per_test": "4.010513", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:immigration:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "682.579", + "cost_per_test": "2.985935", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:insurance:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "631.639", + "cost_per_test": "2.417604", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:intellectual_property:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "628.011", + "cost_per_test": "3.636776", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:international_trade_sanctions:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "636.4", + "cost_per_test": "2.359423", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:litigation_dispute_resolution:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "596.275", + "cost_per_test": "6.932258", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:overall:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.overall", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Overall).", + "metric_id": "vals_ai.hlab.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 3.75, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "3.75", + "raw_stderr": "1.174", + "latency": "733.852", + "cost_per_test": "4.597894", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.174, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:real_estate:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "799.058", + "cost_per_test": "5.565429", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:structured_finance_securitization:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "794.392", + "cost_per_test": "7.609913", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:tax:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "671.392", + "cost_per_test": "4.83132", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:trusts_estates_private_client:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "684.418", + "cost_per_test": "4.928295", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:white_collar_defense_investigations:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.hlab.white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "571.985", + "cost_per_test": "3.615019", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.5/77b98655-d646-4910-86fc-e429cf0ad693.json b/data/vals_ai/openai/gpt-5.5/77b98655-d646-4910-86fc-e429cf0ad693.json new file mode 100644 index 0000000000000000000000000000000000000000..de822cf9651ce03716b7cad59689b4deb30e78fa --- /dev/null +++ b/data/vals_ai/openai/gpt-5.5/77b98655-d646-4910-86fc-e429cf0ad693.json @@ -0,0 +1,103 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/openai_gpt-5.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.5", + "id": "openai/gpt-5.5", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.5", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 88.266, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.266", + "raw_stderr": "0.774", + "latency": "54.147", + "cost_per_test": "0.043319", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.774, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.5/783fa969-cb86-4ad9-9108-9981b9a76082.json b/data/vals_ai/openai/gpt-5.5/783fa969-cb86-4ad9-9108-9981b9a76082.json new file mode 100644 index 0000000000000000000000000000000000000000..dcd1707c107cc4569fe26d8817f657bab2f5a7a8 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.5/783fa969-cb86-4ad9-9108-9981b9a76082.json @@ -0,0 +1,195 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/cyber/openai_gpt-5.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CyberBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "beta", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.5", + "id": "openai/gpt-5.5", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.5", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "cyber:overall:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.cyber.overall", + "source_data": { + "dataset_name": "CyberBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (Overall).", + "metric_id": "vals_ai.cyber.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 80.508, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "beta", + "raw_score": "80.508", + "raw_stderr": "5.155", + "latency": "680.048", + "cost_per_test": "2.807459", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.155, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "cyber:patch:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.cyber.patch", + "source_data": { + "dataset_name": "CyberBench - Patch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "patch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (Patch).", + "metric_id": "vals_ai.cyber.patch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 81.356, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "patch", + "task_name": "Patch", + "dataset_type": "private", + "industry": "beta", + "raw_score": "81.356", + "raw_stderr": "5.07", + "latency": "942.319", + "cost_per_test": "1.416184", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.07, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "cyber:poc:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.cyber.poc", + "source_data": { + "dataset_name": "CyberBench - PoC", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "poc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (PoC).", + "metric_id": "vals_ai.cyber.poc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 79.661, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "poc", + "task_name": "PoC", + "dataset_type": "private", + "industry": "beta", + "raw_score": "79.661", + "raw_stderr": "5.24", + "latency": "417.778", + "cost_per_test": "4.198733", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.24, + "method": "vals_reported" + } + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.5/817bc395-894e-43f7-9ea5-7cb2a5f5bebc.json b/data/vals_ai/openai/gpt-5.5/817bc395-894e-43f7-9ea5-7cb2a5f5bebc.json new file mode 100644 index 0000000000000000000000000000000000000000..f8d8f7661071e14e62fc6a4acc195acf4a469cf5 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.5/817bc395-894e-43f7-9ea5-7cb2a5f5bebc.json @@ -0,0 +1,676 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/emb/openai_gpt-5.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - EMB", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.5", + "id": "openai/gpt-5.5", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.5", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "emb:comps:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.emb.comps", + "source_data": { + "dataset_name": "EMB - Comps", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "comps", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Comps).", + "metric_id": "vals_ai.emb.comps.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 64.681, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "comps", + "task_name": "Comps", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.681", + "raw_stderr": "5.533", + "latency": "925.371", + "cost_per_test": "6.052587", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.533, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:dataroom:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.emb.dataroom", + "source_data": { + "dataset_name": "EMB - Dataroom", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dataroom", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Dataroom).", + "metric_id": "vals_ai.emb.dataroom.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 81.227, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dataroom", + "task_name": "Dataroom", + "dataset_type": "private", + "industry": "finance", + "raw_score": "81.227", + "raw_stderr": "4.694", + "latency": "818.421", + "cost_per_test": "2.424232", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.694, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:dcf:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.emb.dcf", + "source_data": { + "dataset_name": "EMB - DCF", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dcf", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (DCF).", + "metric_id": "vals_ai.emb.dcf.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 56.736, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dcf", + "task_name": "DCF", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.736", + "raw_stderr": "7.836", + "latency": "883.688", + "cost_per_test": "2.839702", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.836, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:lbo:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.emb.lbo", + "source_data": { + "dataset_name": "EMB - LBO", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "lbo", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (LBO).", + "metric_id": "vals_ai.emb.lbo.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 53.222, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "lbo", + "task_name": "LBO", + "dataset_type": "private", + "industry": "finance", + "raw_score": "53.222", + "raw_stderr": "5.136", + "latency": "1094.17", + "cost_per_test": "3.548798", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.136, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:ma:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.emb.ma", + "source_data": { + "dataset_name": "EMB - M&A", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "ma", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (M&A).", + "metric_id": "vals_ai.emb.ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 66.813, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "ma", + "task_name": "M&A", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.813", + "raw_stderr": "8.09", + "latency": "866.362", + "cost_per_test": "2.601886", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 8.09, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:misc:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.emb.misc", + "source_data": { + "dataset_name": "EMB - Misc", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "misc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Misc).", + "metric_id": "vals_ai.emb.misc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 64.647, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "misc", + "task_name": "Misc", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.647", + "raw_stderr": "6.895", + "latency": "985.394", + "cost_per_test": "2.912134", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.895, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:operating:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.emb.operating", + "source_data": { + "dataset_name": "EMB - Operating", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "operating", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Operating).", + "metric_id": "vals_ai.emb.operating.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 64.478, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "operating", + "task_name": "Operating", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.478", + "raw_stderr": "6.329", + "latency": "798.917", + "cost_per_test": "2.199544", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.329, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:overall:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.emb.overall", + "source_data": { + "dataset_name": "EMB - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Overall).", + "metric_id": "vals_ai.emb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 64.543, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.543", + "raw_stderr": "2.869", + "latency": "910.575", + "cost_per_test": "3.273352", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.869, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:scratch:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.emb.scratch", + "source_data": { + "dataset_name": "EMB - Scratch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "scratch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Scratch).", + "metric_id": "vals_ai.emb.scratch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 66.833, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "scratch", + "task_name": "Scratch", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.833", + "raw_stderr": "2.465", + "latency": "947.624", + "cost_per_test": "3.412393", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.465, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:template:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.emb.template", + "source_data": { + "dataset_name": "EMB - Template", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "template", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Template).", + "metric_id": "vals_ai.emb.template.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 61.809, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "template", + "task_name": "Template", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.809", + "raw_stderr": "4.311", + "latency": "873.526", + "cost_per_test": "3.134312", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.311, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.5/83579cdf-30f5-40d3-ae38-8c151fe7a6ec.json b/data/vals_ai/openai/gpt-5.5/83579cdf-30f5-40d3-ae38-8c151fe7a6ec.json new file mode 100644 index 0000000000000000000000000000000000000000..0653463e12e3924e76e1164f8eda9b6f00e5cf23 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.5/83579cdf-30f5-40d3-ae38-8c151fe7a6ec.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/programbench/openai_gpt-5.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProgramBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.5", + "id": "openai/gpt-5.5", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.5", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "programbench:almost:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.programbench.almost", + "source_data": { + "dataset_name": "ProgramBench - Almost Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "almost", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Almost Resolved).", + "metric_id": "vals_ai.programbench.almost.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 9.5, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "almost", + "task_name": "Almost Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "9.5", + "raw_stderr": "2.079", + "latency": "1322.959", + "cost_per_test": "6.95454", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.079, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "programbench:overall:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.programbench.overall", + "source_data": { + "dataset_name": "ProgramBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Overall).", + "metric_id": "vals_ai.programbench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.5, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0.5", + "raw_stderr": "0.5", + "latency": "1322.959", + "cost_per_test": "6.95454", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.5, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "programbench:partial:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.programbench.partial", + "source_data": { + "dataset_name": "ProgramBench - Raw Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "partial", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Raw Pass Rate).", + "metric_id": "vals_ai.programbench.partial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 70.775, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "partial", + "task_name": "Raw Pass Rate", + "dataset_type": "public", + "industry": "coding", + "raw_score": "70.775", + "raw_stderr": "1.822", + "latency": "1322.959", + "cost_per_test": "6.95454", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.822, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "programbench:strict:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.programbench.strict", + "source_data": { + "dataset_name": "ProgramBench - Fully Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "strict", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Fully Resolved).", + "metric_id": "vals_ai.programbench.strict.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.5, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "strict", + "task_name": "Fully Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0.5", + "raw_stderr": "0.5", + "latency": "1322.959", + "cost_per_test": "6.95454", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.5, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.5/8aedca3e-e707-4fc5-86ae-7238af2bf0c3.json b/data/vals_ai/openai/gpt-5.5/8aedca3e-e707-4fc5-86ae-7238af2bf0c3.json new file mode 100644 index 0000000000000000000000000000000000000000..257e41c4fcbed67bba5bc604f4bc447a893ba310 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.5/8aedca3e-e707-4fc5-86ae-7238af2bf0c3.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/openai_gpt-5.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.5", + "id": "openai/gpt-5.5", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.5", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 86.868, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "86.868", + "raw_stderr": "1.932", + "latency": "132.761", + "cost_per_test": "0.142988", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.932, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.5/91fd8a77-77d5-47be-ba05-c5f4f87deb78.json b/data/vals_ai/openai/gpt-5.5/91fd8a77-77d5-47be-ba05-c5f4f87deb78.json new file mode 100644 index 0000000000000000000000000000000000000000..6316a22162b1aa8be8a3ae6fa3142fe366e35fba --- /dev/null +++ b/data/vals_ai/openai/gpt-5.5/91fd8a77-77d5-47be-ba05-c5f4f87deb78.json @@ -0,0 +1,364 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/openai_gpt-5.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.5", + "id": "openai/gpt-5.5", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.5", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 39.683, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "39.683", + "raw_stderr": "6.148", + "latency": "91.895", + "cost_per_test": "0.142524", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.148, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "sage:automata:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 40.476, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "40.476", + "raw_stderr": "8.871", + "latency": "91.557", + "cost_per_test": "0.151694", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 8.871, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "sage:calculus:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 72.222, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "72.222", + "raw_stderr": "5.571", + "latency": "54.876", + "cost_per_test": "0.112212", + "temperature": "1", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.571, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 74.436, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "74.436", + "raw_stderr": "4.923", + "latency": "44.698", + "cost_per_test": "0.114649", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.923, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "sage:overall:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 51.532, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "51.532", + "raw_stderr": "3.954", + "latency": "76.05", + "cost_per_test": "0.136289", + "temperature": "1", + "max_output_tokens": "30000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.954, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.5/92afdfba-b428-4464-a7a6-0a3873987c72.json b/data/vals_ai/openai/gpt-5.5/92afdfba-b428-4464-a7a6-0a3873987c72.json new file mode 100644 index 0000000000000000000000000000000000000000..9b4a8e523c420111244c1324885ed8afd1ed2a64 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.5/92afdfba-b428-4464-a7a6-0a3873987c72.json @@ -0,0 +1,438 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/openai_gpt-5.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.5", + "id": "openai/gpt-5.5", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.5", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 89.644, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.644", + "raw_stderr": "0.697", + "latency": "11.168", + "cost_per_test": "0.00784", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.697, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.836, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.836", + "raw_stderr": "0.194", + "latency": "15.562", + "cost_per_test": "0.014325", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.194, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 91.193, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "91.193", + "raw_stderr": "0.547", + "latency": "6.799", + "cost_per_test": "0.00728", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.547, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 86.515, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.515", + "raw_stderr": "0.41", + "latency": "18.139", + "cost_per_test": "0.015403", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.41, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.448, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.448", + "raw_stderr": "1.076", + "latency": "7.139", + "cost_per_test": "0.006809", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.076, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 85.453, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.453", + "raw_stderr": "1.488", + "latency": "50.029", + "cost_per_test": "0.040762", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.488, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.5/9b77eb18-c96f-42f6-a26b-8f30996a5863.json b/data/vals_ai/openai/gpt-5.5/9b77eb18-c96f-42f6-a26b-8f30996a5863.json new file mode 100644 index 0000000000000000000000000000000000000000..9f9dd1188a02bd63acf75c68ebaf1c30410964c1 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.5/9b77eb18-c96f-42f6-a26b-8f30996a5863.json @@ -0,0 +1,536 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_multimodal_index/openai_gpt-5.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Multimodal Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.5", + "id": "openai/gpt-5.5", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.5", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_multimodal_index:corp_fin_v2:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Multimodal Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (CorpFin v2).", + "metric_id": "vals_ai.vals_multimodal_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 69.58, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "69.58", + "raw_stderr": "1.571", + "latency": "26.764", + "cost_per_test": "0.43166", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.571, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:finance_agent:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.finance_agent", + "source_data": { + "dataset_name": "Vals Multimodal Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_multimodal_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 51.834, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "51.834", + "raw_stderr": "0.461", + "latency": "616.56", + "cost_per_test": "3.784562", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.461, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:mortgage_tax:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.mortgage_tax", + "source_data": { + "dataset_name": "Vals Multimodal Index - Mortgage Tax", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "mortgage_tax", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Mortgage Tax).", + "metric_id": "vals_ai.vals_multimodal_index.mortgage_tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 68.76, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "mortgage_tax", + "task_name": "Mortgage Tax", + "dataset_type": "private", + "industry": "index", + "raw_score": "68.76", + "raw_stderr": "0.912", + "latency": "28.243", + "cost_per_test": "0.071407", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.912, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:overall:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.overall", + "source_data": { + "dataset_name": "Vals Multimodal Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Overall).", + "metric_id": "vals_ai.vals_multimodal_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 68.071, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "68.071", + "raw_stderr": "1.419", + "latency": "501.86", + "cost_per_test": "3.314182", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.419, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:sage:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.sage", + "source_data": { + "dataset_name": "Vals Multimodal Index - SAGE", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "sage", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SAGE).", + "metric_id": "vals_ai.vals_multimodal_index.sage.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 51.532, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "sage", + "task_name": "SAGE", + "dataset_type": "private", + "industry": "index", + "raw_score": "51.532", + "raw_stderr": "3.954", + "latency": "76.05", + "cost_per_test": "0.130637", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.954, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:swebench:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.swebench", + "source_data": { + "dataset_name": "Vals Multimodal Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SWE-bench).", + "metric_id": "vals_ai.vals_multimodal_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 82.6, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "82.6", + "raw_stderr": "1.695", + "latency": "426.425", + "cost_per_test": "1.361765" + }, + "uncertainty": { + "standard_error": { + "value": 1.695, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:terminal_bench_2_1:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Multimodal Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_multimodal_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 76.404, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "76.404", + "raw_stderr": "0.649", + "latency": "427.412", + "cost_per_test": "0.741883", + "max_output_tokens": "128000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.649, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:vibe_code_bench:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Multimodal Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_multimodal_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 77.099, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "77.099", + "raw_stderr": "7.028", + "latency": "1911.567", + "cost_per_test": "16.661911", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.028, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.5/c812337c-058a-40c0-9a78-faa62c3323a2.json b/data/vals_ai/openai/gpt-5.5/c812337c-058a-40c0-9a78-faa62c3323a2.json new file mode 100644 index 0000000000000000000000000000000000000000..46d494a97ca92e05386cb0c47a05e48edd2420eb --- /dev/null +++ b/data/vals_ai/openai/gpt-5.5/c812337c-058a-40c0-9a78-faa62c3323a2.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/code-migration/openai_gpt-5.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Code Migration", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.5", + "id": "openai/gpt-5.5", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.5", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "code-migration:cli:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.code-migration.cli", + "source_data": { + "dataset_name": "Code Migration - CLI", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cli", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (CLI).", + "metric_id": "vals_ai.code-migration.cli.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 36.882, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cli", + "task_name": "CLI", + "dataset_type": "private", + "industry": "coding", + "raw_score": "36.882", + "raw_stderr": "2.212", + "latency": "1746.371", + "cost_per_test": "7.539299", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.212, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "code-migration:cobol:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.code-migration.cobol", + "source_data": { + "dataset_name": "Code Migration - COBOL", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cobol", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (COBOL).", + "metric_id": "vals_ai.code-migration.cobol.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 70.0, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cobol", + "task_name": "COBOL", + "dataset_type": "private", + "industry": "coding", + "raw_score": "70", + "raw_stderr": "15.275", + "latency": "2340.697", + "cost_per_test": "3.151116", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 15.275, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "code-migration:code_quality:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.code-migration.code_quality", + "source_data": { + "dataset_name": "Code Migration - Code Quality", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "code_quality", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Code Quality).", + "metric_id": "vals_ai.code-migration.code_quality.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 3.498, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "code_quality", + "task_name": "Code Quality", + "dataset_type": "private", + "industry": "coding", + "raw_score": "3.498", + "raw_stderr": "0.048", + "latency": "1894.953", + "cost_per_test": "6.442253", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.048, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "code-migration:overall:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.code-migration.overall", + "source_data": { + "dataset_name": "Code Migration - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Overall).", + "metric_id": "vals_ai.code-migration.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 45.161, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "45.161", + "raw_stderr": "4.164", + "latency": "1894.953", + "cost_per_test": "6.442253", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.164, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.5/ca644dba-b008-4126-84b4-1393ea15d50a.json b/data/vals_ai/openai/gpt-5.5/ca644dba-b008-4126-84b4-1393ea15d50a.json new file mode 100644 index 0000000000000000000000000000000000000000..06380514a99ea79543ac2f27662753536357b3bd --- /dev/null +++ b/data/vals_ai/openai/gpt-5.5/ca644dba-b008-4126-84b4-1393ea15d50a.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/openai_gpt-5.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.5", + "id": "openai/gpt-5.5", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.5", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 93.434, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.434", + "raw_stderr": "1.822", + "latency": "118.4", + "cost_per_test": "0.092359", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.822, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 92.929, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.929", + "raw_stderr": "1.822", + "latency": "106.252", + "cost_per_test": "0.073175", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.822, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 93.182, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.182", + "raw_stderr": "1.288", + "latency": "112.326", + "cost_per_test": "0.082767", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.288, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.5/df9cb991-e0a2-4aef-bbdb-3ad14544e899.json b/data/vals_ai/openai/gpt-5.5/df9cb991-e0a2-4aef-bbdb-3ad14544e899.json new file mode 100644 index 0000000000000000000000000000000000000000..612fbbeb3440bb2ea65568408bf6367e198d1ef3 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.5/df9cb991-e0a2-4aef-bbdb-3ad14544e899.json @@ -0,0 +1,1188 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_research/openai_gpt-5.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Legal Research Bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.5", + "id": "openai/gpt-5.5", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.5", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_research:administrative_regulatory:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.legal_research.administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 53.03, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "53.03", + "raw_stderr": "6.19", + "latency": "2093.383", + "cost_per_test": "7.953883", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.19, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:business_commercial:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.legal_research.business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 37.209, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "37.209", + "raw_stderr": "7.458", + "latency": "2266.219", + "cost_per_test": "6.684872", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.458, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:civil_litigation:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.legal_research.civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 44.444, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "44.444", + "raw_stderr": "9.745", + "latency": "2408.394", + "cost_per_test": "7.00366", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 9.745, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:constitutional_civil_rights:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.legal_research.constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 28.947, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "28.947", + "raw_stderr": "7.456", + "latency": "2156.642", + "cost_per_test": "5.731692", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.456, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:criminal:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.legal_research.criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 36.842, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "36.842", + "raw_stderr": "6.446", + "latency": "2310.474", + "cost_per_test": "7.322657", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.446, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:family:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.legal_research.family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 13.636, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "13.636", + "raw_stderr": "7.489", + "latency": "2680.585", + "cost_per_test": "7.622096", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.489, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:health:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.legal_research.health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "66.667", + "raw_stderr": "8.754", + "latency": "2062.802", + "cost_per_test": "9.360291", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 8.754, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:immigration:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.legal_research.immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 36.364, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "36.364", + "raw_stderr": "15.212", + "latency": "2512.615", + "cost_per_test": "8.597148", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 15.212, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:overall:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.legal_research.overall", + "source_data": { + "dataset_name": "Legal Research Bench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Overall).", + "metric_id": "vals_ai.legal_research.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 40.385, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "legal", + "raw_score": "40.385", + "raw_stderr": "3.41", + "latency": "2265.105", + "cost_per_test": "7.402471", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.41, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass", + "source_data": { + "dataset_name": "Legal Research Bench - Weighted Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Weighted Pass).", + "metric_id": "vals_ai.legal_research.weighted_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 83.964, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass", + "task_name": "Weighted Pass", + "dataset_type": "private", + "industry": "legal", + "raw_score": "83.964", + "raw_stderr": "1.474", + "latency": "2265.105", + "cost_per_test": "7.402471", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.474, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__administrative_regulatory:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.weighted_pass__administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 86.669, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "86.669", + "raw_stderr": "2.717", + "latency": "2093.383", + "cost_per_test": "7.953883", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.717, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__business_commercial:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.weighted_pass__business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 86.764, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "86.764", + "raw_stderr": "2.947", + "latency": "2266.219", + "cost_per_test": "6.684872", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.947, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__civil_litigation:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.weighted_pass__civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 87.616, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "87.616", + "raw_stderr": "3.561", + "latency": "2408.394", + "cost_per_test": "7.00366", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.561, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__constitutional_civil_rights:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 83.745, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "83.745", + "raw_stderr": "2.425", + "latency": "2156.642", + "cost_per_test": "5.731692", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.425, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__criminal:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.weighted_pass__criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 81.837, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "81.837", + "raw_stderr": "2.851", + "latency": "2310.474", + "cost_per_test": "7.322657", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.851, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__family:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.weighted_pass__family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 70.836, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "70.836", + "raw_stderr": "4.872", + "latency": "2680.585", + "cost_per_test": "7.622096", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.872, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__health:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.weighted_pass__health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 93.59, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "93.59", + "raw_stderr": "2.623", + "latency": "2062.802", + "cost_per_test": "9.360291", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.623, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__immigration:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.weighted_pass__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 82.114, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "82.114", + "raw_stderr": "6.489", + "latency": "2512.615", + "cost_per_test": "8.597148", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.489, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.5/e20593b4-3916-4992-b8c0-5b3784aef13e.json b/data/vals_ai/openai/gpt-5.5/e20593b4-3916-4992-b8c0-5b3784aef13e.json new file mode 100644 index 0000000000000000000000000000000000000000..9268039971d788f8aab6bcde3f0bc88c572be2e5 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.5/e20593b4-3916-4992-b8c0-5b3784aef13e.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/openai_gpt-5.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.5", + "id": "openai/gpt-5.5", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.5", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 76.153, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "76.153", + "raw_stderr": "1.201", + "latency": "39.334", + "cost_per_test": "0.079986", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.201, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 68.76, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.76", + "raw_stderr": "0.912", + "latency": "28.243", + "cost_per_test": "0.071407", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.912, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 61.367, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.367", + "raw_stderr": "1.373", + "latency": "17.152", + "cost_per_test": "0.062828", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.373, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.5/fe1bfa4c-48be-4c35-bd01-6ec463c79a3e.json b/data/vals_ai/openai/gpt-5.5/fe1bfa4c-48be-4c35-bd01-6ec463c79a3e.json new file mode 100644 index 0000000000000000000000000000000000000000..080ad5a1f4257ad9dc63f2435bf4fde86d934a3b --- /dev/null +++ b/data/vals_ai/openai/gpt-5.5/fe1bfa4c-48be-4c35-bd01-6ec463c79a3e.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/openai_gpt-5.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.5", + "id": "openai/gpt-5.5", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.5", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 65.495, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.495", + "raw_stderr": "1.359", + "latency": "95.83", + "cost_per_test": "0.06042", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.359, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 74.98, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "74.98", + "raw_stderr": "0.854", + "latency": "95.83", + "cost_per_test": "0.06042", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.854, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:openai/gpt-5.5:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 84.464, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "84.464", + "raw_stderr": "1.036", + "latency": "95.83", + "cost_per_test": "0.06042", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.036, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-luna/1c5950f7-d679-4c36-b75f-d14bdfd34107.json b/data/vals_ai/openai/gpt-5.6-luna/1c5950f7-d679-4c36-b75f-d14bdfd34107.json new file mode 100644 index 0000000000000000000000000000000000000000..c305a22de310a0e92d125882ab21abfe39736bac --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-luna/1c5950f7-d679-4c36-b75f-d14bdfd34107.json @@ -0,0 +1,356 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/openai_gpt-5.6-luna/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-luna", + "id": "openai/gpt-5.6-luna", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-luna", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:openai/gpt-5.6-luna:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 38.095, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "38.095", + "raw_stderr": "6.103", + "latency": "52.28", + "cost_per_test": "0.053583", + "max_output_tokens": "30000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.103, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "sage:automata:openai/gpt-5.6-luna:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 31.35, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "31.35", + "raw_stderr": "5.772", + "latency": "70.071", + "cost_per_test": "0.054355", + "max_output_tokens": "30000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.772, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "sage:calculus:openai/gpt-5.6-luna:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 64.506, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "64.506", + "raw_stderr": "6.102", + "latency": "47.895", + "cost_per_test": "0.092103", + "max_output_tokens": "30000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.102, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:openai/gpt-5.6-luna:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 63.216, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "63.216", + "raw_stderr": "5.413", + "latency": "33.283", + "cost_per_test": "0.049738", + "max_output_tokens": "30000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.413, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "sage:overall:openai/gpt-5.6-luna:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 44.22, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "44.22", + "raw_stderr": "3.331", + "latency": "51.878", + "cost_per_test": "0.052559", + "max_output_tokens": "30000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.331, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-luna/4be54e8b-d3bd-465f-a6ca-32beba11774a.json b/data/vals_ai/openai/gpt-5.6-luna/4be54e8b-d3bd-465f-a6ca-32beba11774a.json new file mode 100644 index 0000000000000000000000000000000000000000..0f8275533382869530b7966c03af585ef2b74d50 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-luna/4be54e8b-d3bd-465f-a6ca-32beba11774a.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medcode/openai_gpt-5.6-luna/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedCode", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medcode", + "benchmark_name": "MedCode", + "benchmark_updated": "2026-07-08", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-luna", + "id": "openai/gpt-5.6-luna", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-luna", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medcode:overall:openai/gpt-5.6-luna:score", + "evaluation_name": "vals_ai.medcode.overall", + "source_data": { + "dataset_name": "MedCode - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medcode", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedCode (Overall).", + "metric_id": "vals_ai.medcode.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode" + } + }, + "score_details": { + "score": 42.391, + "details": { + "benchmark_slug": "medcode", + "benchmark_name": "MedCode", + "benchmark_updated": "2026-07-08", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "42.391", + "raw_stderr": "2.27", + "latency": "81.283", + "cost_per_test": "0.07985", + "max_output_tokens": "30000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.27, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-luna/5505b0c2-0fc3-4927-ad24-027c83f55b49.json b/data/vals_ai/openai/gpt-5.6-luna/5505b0c2-0fc3-4927-ad24-027c83f55b49.json new file mode 100644 index 0000000000000000000000000000000000000000..7c25d6a8530ec7c9faec87e737536df692d667c8 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-luna/5505b0c2-0fc3-4927-ad24-027c83f55b49.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/openai_gpt-5.6-luna/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-luna", + "id": "openai/gpt-5.6-luna", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-luna", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:openai/gpt-5.6-luna:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 77.055, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "77.055", + "raw_stderr": "3.069", + "latency": "1551.831", + "cost_per_test": "3.63213", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.069, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-luna/62a20058-bf0c-45cd-924e-e31928e162bc.json b/data/vals_ai/openai/gpt-5.6-luna/62a20058-bf0c-45cd-924e-e31928e162bc.json new file mode 100644 index 0000000000000000000000000000000000000000..ba7f964cd2355876949d24d554a37aa5375ff647 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-luna/62a20058-bf0c-45cd-924e-e31928e162bc.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/openai_gpt-5.6-luna/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-luna", + "id": "openai/gpt-5.6-luna", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-luna", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:openai/gpt-5.6-luna:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 63.52, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.52", + "raw_stderr": "1.643", + "latency": "10.874", + "cost_per_test": "0.009937", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.643, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:openai/gpt-5.6-luna:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 64.103, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.103", + "raw_stderr": "1.638", + "latency": "19.928", + "cost_per_test": "0.194349", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.638, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:openai/gpt-5.6-luna:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 64.219, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.219", + "raw_stderr": "0.945", + "latency": "16.968", + "cost_per_test": "0.095195", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.945, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:openai/gpt-5.6-luna:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 65.035, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.035", + "raw_stderr": "1.628", + "latency": "20.103", + "cost_per_test": "0.0813", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.628, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-luna/87c3572e-b2f9-49e3-ac20-8d661d6c5d36.json b/data/vals_ai/openai/gpt-5.6-luna/87c3572e-b2f9-49e3-ac20-8d661d6c5d36.json new file mode 100644 index 0000000000000000000000000000000000000000..f989150b22a5430c30c387b3cbc279bd61e9e8c0 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-luna/87c3572e-b2f9-49e3-ac20-8d661d6c5d36.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/openai_gpt-5.6-luna/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-luna", + "id": "openai/gpt-5.6-luna", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-luna", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:openai/gpt-5.6-luna:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 83.333, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "83.333", + "raw_stderr": "16.667", + "latency": "7915.469", + "cost_per_test": "5.981552", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 16.667, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:openai/gpt-5.6-luna:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 70.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "70", + "raw_stderr": "1.925", + "latency": "1055.396", + "cost_per_test": "0.79754", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.925, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:openai/gpt-5.6-luna:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 83.636, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "83.636", + "raw_stderr": "1.05", + "latency": "575.67", + "cost_per_test": "0.435022", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.05, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:openai/gpt-5.6-luna:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 79.026, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "79.026", + "raw_stderr": "0.991", + "latency": "355.751", + "cost_per_test": "0.268834", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.991, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-luna/be0c634e-70b8-40c2-bc6b-9680ca79c1d4.json b/data/vals_ai/openai/gpt-5.6-luna/be0c634e-70b8-40c2-bc6b-9680ca79c1d4.json new file mode 100644 index 0000000000000000000000000000000000000000..2d959ad8da3b829bbefcfb3d44d6d18750977d4d --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-luna/be0c634e-70b8-40c2-bc6b-9680ca79c1d4.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/openai_gpt-5.6-luna/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-luna", + "id": "openai/gpt-5.6-luna", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-luna", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:openai/gpt-5.6-luna:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 78.0, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "78", + "raw_stderr": "13.904", + "latency": "1730.527", + "cost_per_test": "1.303994", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 13.904, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:openai/gpt-5.6-luna:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 67.833, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "67.833", + "raw_stderr": "15.511", + "latency": "1610.589", + "cost_per_test": "1.32468", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 15.511, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "ioi:overall:openai/gpt-5.6-luna:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 72.917, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "72.917", + "raw_stderr": "5.083", + "latency": "1670.558", + "cost_per_test": "1.314337", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.083, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-luna/d2a2451f-5f89-4661-8c27-068691033d8e.json b/data/vals_ai/openai/gpt-5.6-luna/d2a2451f-5f89-4661-8c27-068691033d8e.json new file mode 100644 index 0000000000000000000000000000000000000000..c5924537489a406760d6813fc15bd8c98655a857 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-luna/d2a2451f-5f89-4661-8c27-068691033d8e.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/openai_gpt-5.6-luna/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-luna", + "id": "openai/gpt-5.6-luna", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-luna", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:openai/gpt-5.6-luna:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 54.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "54", + "raw_stderr": "5.009", + "latency": "387.84", + "cost_per_test": "0.449325", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.009, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-luna/f2724c97-472f-4322-852a-71bea23f8c06.json b/data/vals_ai/openai/gpt-5.6-luna/f2724c97-472f-4322-852a-71bea23f8c06.json new file mode 100644 index 0000000000000000000000000000000000000000..90945f133b245b3b008bba8d91de797f30bf3de7 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-luna/f2724c97-472f-4322-852a-71bea23f8c06.json @@ -0,0 +1,103 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/openai_gpt-5.6-luna/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-luna", + "id": "openai/gpt-5.6-luna", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-luna", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:openai/gpt-5.6-luna:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 85.029, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.029", + "raw_stderr": "0.858", + "latency": "35.102", + "cost_per_test": "0.032093", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.858, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-sol/03823d71-f6d7-4758-95ee-550399079537.json b/data/vals_ai/openai/gpt-5.6-sol/03823d71-f6d7-4758-95ee-550399079537.json new file mode 100644 index 0000000000000000000000000000000000000000..8d5f050256f8635204297d97d0541394921bdcea --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-sol/03823d71-f6d7-4758-95ee-550399079537.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/openai_gpt-5.6-sol/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-sol", + "id": "openai/gpt-5.6-sol", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-sol", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 94.949, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.949", + "raw_stderr": "1.556", + "latency": "72.403", + "cost_per_test": "0.135633", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.556, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 95.455, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "95.455", + "raw_stderr": "1.48", + "latency": "44.206", + "cost_per_test": "0.084288", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.48, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 95.202, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "95.202", + "raw_stderr": "1.074", + "latency": "58.304", + "cost_per_test": "0.10996", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.074, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-sol/166289c0-a153-46e8-b979-bb5701f04988.json b/data/vals_ai/openai/gpt-5.6-sol/166289c0-a153-46e8-b979-bb5701f04988.json new file mode 100644 index 0000000000000000000000000000000000000000..b1ceb09bc6bc3b61ab08d26505c8a89ebcca7ebf --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-sol/166289c0-a153-46e8-b979-bb5701f04988.json @@ -0,0 +1,356 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/openai_gpt-5.6-sol/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-sol", + "id": "openai/gpt-5.6-sol", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-sol", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 47.619, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "47.619", + "raw_stderr": "6.27", + "latency": "67.069", + "cost_per_test": "0.125985", + "max_output_tokens": "30000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.27, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "sage:automata:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 39.682, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "39.682", + "raw_stderr": "6.262", + "latency": "115.524", + "cost_per_test": "0.205378", + "max_output_tokens": "30000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.262, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "sage:calculus:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 67.284, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "67.284", + "raw_stderr": "5.681", + "latency": "46.739", + "cost_per_test": "0.157303", + "max_output_tokens": "30000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.681, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 70.388, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "70.388", + "raw_stderr": "5.146", + "latency": "43.593", + "cost_per_test": "0.112568", + "max_output_tokens": "30000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.146, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "sage:overall:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 52.563, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "52.563", + "raw_stderr": "3.416", + "latency": "75.395", + "cost_per_test": "0.147977", + "max_output_tokens": "30000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.416, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-sol/1d6f1e1d-6727-4e93-84a4-809ee1838394.json b/data/vals_ai/openai/gpt-5.6-sol/1d6f1e1d-6727-4e93-84a4-809ee1838394.json new file mode 100644 index 0000000000000000000000000000000000000000..9f111713807ed62e2c1cbf9556dd89fe86509bce --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-sol/1d6f1e1d-6727-4e93-84a4-809ee1838394.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/openai_gpt-5.6-sol/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-sol", + "id": "openai/gpt-5.6-sol", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-sol", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.538, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.538", + "raw_stderr": "1.661", + "latency": "8.564", + "cost_per_test": "0.026875", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.661, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 65.152, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.152", + "raw_stderr": "1.627", + "latency": "18.943", + "cost_per_test": "1.060797", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.627, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 64.375, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.375", + "raw_stderr": "0.943", + "latency": "15.941", + "cost_per_test": "0.484982", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.943, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 66.434, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.434", + "raw_stderr": "1.612", + "latency": "20.315", + "cost_per_test": "0.367275", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.612, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-sol/2203ee15-4e91-4ab3-89b4-cef4c8088f37.json b/data/vals_ai/openai/gpt-5.6-sol/2203ee15-4e91-4ab3-89b4-cef4c8088f37.json new file mode 100644 index 0000000000000000000000000000000000000000..27d8f2d88f5db69e5b1fda8a121795506b56dfc9 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-sol/2203ee15-4e91-4ab3-89b4-cef4c8088f37.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/public-benefits-bench/openai_gpt-5.6-sol/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Public Benefits Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "social mobility", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-sol", + "id": "openai/gpt-5.6-sol", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-sol", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "public-benefits-bench:overall:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.public-benefits-bench.overall", + "source_data": { + "dataset_name": "Public Benefits Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Public Benefits Bench v1.1 (Overall).", + "metric_id": "vals_ai.public-benefits-bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "score_details": { + "score": 66.509, + "details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "social mobility", + "raw_score": "66.509", + "raw_stderr": "1.228", + "latency": "3796.706", + "cost_per_test": "8.847882", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.228, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-sol/2f64779f-5157-4eb1-af95-c273dc20cb1a.json b/data/vals_ai/openai/gpt-5.6-sol/2f64779f-5157-4eb1-af95-c273dc20cb1a.json new file mode 100644 index 0000000000000000000000000000000000000000..b0111ce7dc3fbf3ca170f6759be3077ca73ea28f --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-sol/2f64779f-5157-4eb1-af95-c273dc20cb1a.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/openai_gpt-5.6-sol/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-sol", + "id": "openai/gpt-5.6-sol", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-sol", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 85.233, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "85.233", + "raw_stderr": "1.973", + "latency": "94.404", + "cost_per_test": "0.276691", + "max_output_tokens": "30000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.973, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-sol/36e1809c-6f3e-4198-aa9a-c8975b825842.json b/data/vals_ai/openai/gpt-5.6-sol/36e1809c-6f3e-4198-aa9a-c8975b825842.json new file mode 100644 index 0000000000000000000000000000000000000000..4fccb6b50cd177f78198a2a5e6540d45a5f00e72 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-sol/36e1809c-6f3e-4198-aa9a-c8975b825842.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/openai_gpt-5.6-sol/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-sol", + "id": "openai/gpt-5.6-sol", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-sol", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 77.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "77", + "raw_stderr": "4.208", + "latency": "516.531", + "cost_per_test": "1.736601", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.208, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-sol/38b04ffa-423e-4aa5-aafa-b70e8637c9ad.json b/data/vals_ai/openai/gpt-5.6-sol/38b04ffa-423e-4aa5-aafa-b70e8637c9ad.json new file mode 100644 index 0000000000000000000000000000000000000000..de6aad13ed7d43f70ec2217565750642a52afc7e --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-sol/38b04ffa-423e-4aa5-aafa-b70e8637c9ad.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/openai_gpt-5.6-sol/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-sol", + "id": "openai/gpt-5.6-sol", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-sol", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 65.004, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.004", + "raw_stderr": "1.364", + "latency": "62.034", + "cost_per_test": "0.138418", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.364, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 74.775, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "74.775", + "raw_stderr": "0.856", + "latency": "62.034", + "cost_per_test": "0.138418", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.856, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 84.546, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "84.546", + "raw_stderr": "1.034", + "latency": "62.034", + "cost_per_test": "0.138418", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.034, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-sol/3d35d0c5-cdc9-499a-b80c-8ca5168aa022.json b/data/vals_ai/openai/gpt-5.6-sol/3d35d0c5-cdc9-499a-b80c-8ca5168aa022.json new file mode 100644 index 0000000000000000000000000000000000000000..9b0bc1b71c73892fe3e5db588b19a3c0ec36a90f --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-sol/3d35d0c5-cdc9-499a-b80c-8ca5168aa022.json @@ -0,0 +1,371 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/openai_gpt-5.6-sol/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-sol", + "id": "openai/gpt-5.6-sol", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-sol", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 97.619, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "97.619", + "raw_stderr": "2.381", + "latency": "270.353", + "cost_per_test": "1.750776", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.381, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 95.019, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "95.019", + "raw_stderr": "1.349", + "latency": "186.034", + "cost_per_test": "1.213534", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.349, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 97.423, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "97.423", + "raw_stderr": "1.141", + "latency": "142.548", + "cost_per_test": "0.926284", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.141, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 100.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "100", + "raw_stderr": "0", + "latency": "1206.142", + "cost_per_test": "1.895371", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:overall:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 96.2, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "96.2", + "raw_stderr": "0.856", + "latency": "182.365", + "cost_per_test": "1.1513", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.856, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-sol/4640c7cc-ba92-44cc-b799-7f1d5c8d1fb8.json b/data/vals_ai/openai/gpt-5.6-sol/4640c7cc-ba92-44cc-b799-7f1d5c8d1fb8.json new file mode 100644 index 0000000000000000000000000000000000000000..adc7ecf19d7508f5ab696c7cfc1db92f5394a5b1 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-sol/4640c7cc-ba92-44cc-b799-7f1d5c8d1fb8.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/openai_gpt-5.6-sol/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-sol", + "id": "openai/gpt-5.6-sol", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-sol", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.561, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.561", + "raw_stderr": "0.847", + "latency": "7.724", + "cost_per_test": "0.018769", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.847, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.283, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.283", + "raw_stderr": "0.891", + "latency": "11.262", + "cost_per_test": "0.0265", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.891, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.961, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.961", + "raw_stderr": "0.808", + "latency": "14.047", + "cost_per_test": "0.038842", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.808, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.732, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.732", + "raw_stderr": "1.432", + "latency": "20.3", + "cost_per_test": "0.030318", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.432, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.64, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.64", + "raw_stderr": "1.003", + "latency": "6.99", + "cost_per_test": "0.018513", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.003, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.822, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.822", + "raw_stderr": "1.051", + "latency": "46.83", + "cost_per_test": "0.096053", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.051, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.108, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.108", + "raw_stderr": "1.278", + "latency": "11.833", + "cost_per_test": "0.025299", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.278, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.89, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.89", + "raw_stderr": "1.973", + "latency": "9.19", + "cost_per_test": "0.021669", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.973, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.11, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.11", + "raw_stderr": "1.225", + "latency": "25.942", + "cost_per_test": "0.053742", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.225, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 95.263, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "95.263", + "raw_stderr": "0.578", + "latency": "9.448", + "cost_per_test": "0.021345", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.578, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.662, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.662", + "raw_stderr": "1.082", + "latency": "10.365", + "cost_per_test": "0.023076", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.082, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.1, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.1", + "raw_stderr": "0.308", + "latency": "15.654", + "cost_per_test": "0.032087", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.308, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.377, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.377", + "raw_stderr": "1.435", + "latency": "24.805", + "cost_per_test": "0.025924", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.435, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.764, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.764", + "raw_stderr": "0.719", + "latency": "13.717", + "cost_per_test": "0.032534", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.719, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.223, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.223", + "raw_stderr": "1.098", + "latency": "6.702", + "cost_per_test": "0.016633", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.098, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-sol/64d5b246-3f52-4807-861a-af1287771262.json b/data/vals_ai/openai/gpt-5.6-sol/64d5b246-3f52-4807-861a-af1287771262.json new file mode 100644 index 0000000000000000000000000000000000000000..a28c4304087e15cde20374d5d18c8a1dbf271f25 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-sol/64d5b246-3f52-4807-861a-af1287771262.json @@ -0,0 +1,103 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/openai_gpt-5.6-sol/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-sol", + "id": "openai/gpt-5.6-sol", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-sol", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 88.844, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.844", + "raw_stderr": "0.757", + "latency": "26.198", + "cost_per_test": "0.055795", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.757, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-sol/68344f5c-d585-4105-b053-f345d0bef0d8.json b/data/vals_ai/openai/gpt-5.6-sol/68344f5c-d585-4105-b053-f345d0bef0d8.json new file mode 100644 index 0000000000000000000000000000000000000000..8c415ea5d3cc0b02451ac33560393d9270023bf3 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-sol/68344f5c-d585-4105-b053-f345d0bef0d8.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/openai_gpt-5.6-sol/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-sol", + "id": "openai/gpt-5.6-sol", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-sol", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.137, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.137", + "raw_stderr": "0.754", + "latency": "4.226", + "cost_per_test": "0.007722", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.754, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "lcb:hard:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 67.429, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "67.429", + "raw_stderr": "2.505", + "latency": "146.86", + "cost_per_test": "0.226187", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.505, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "lcb:medium:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 82.245, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "82.245", + "raw_stderr": "1.953", + "latency": "18.454", + "cost_per_test": "0.035584", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.953, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "lcb:overall:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 82.604, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "82.604", + "raw_stderr": "1.088", + "latency": "56.513", + "cost_per_test": "0.089831", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.088, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-sol/6f89d906-236f-461b-9257-eec58d3c087d.json b/data/vals_ai/openai/gpt-5.6-sol/6f89d906-236f-461b-9257-eec58d3c087d.json new file mode 100644 index 0000000000000000000000000000000000000000..6284ea4dc8068b5d184eaac8da704ecc6b0e272e --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-sol/6f89d906-236f-461b-9257-eec58d3c087d.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medcode/openai_gpt-5.6-sol/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedCode", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medcode", + "benchmark_name": "MedCode", + "benchmark_updated": "2026-07-08", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-sol", + "id": "openai/gpt-5.6-sol", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-sol", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medcode:overall:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.medcode.overall", + "source_data": { + "dataset_name": "MedCode - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medcode", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedCode (Overall).", + "metric_id": "vals_ai.medcode.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode" + } + }, + "score_details": { + "score": 43.974, + "details": { + "benchmark_slug": "medcode", + "benchmark_name": "MedCode", + "benchmark_updated": "2026-07-08", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "43.974", + "raw_stderr": "2.258", + "latency": "96.589", + "cost_per_test": "0.280517", + "max_output_tokens": "30000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.258, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-sol/754a0b58-6fa0-400f-94fb-3f37f98b1dbe.json b/data/vals_ai/openai/gpt-5.6-sol/754a0b58-6fa0-400f-94fb-3f37f98b1dbe.json new file mode 100644 index 0000000000000000000000000000000000000000..f2b97918339d9fd6632d48f0f11d5d3d9c4909b0 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-sol/754a0b58-6fa0-400f-94fb-3f37f98b1dbe.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/programbench/openai_gpt-5.6-sol/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProgramBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-sol", + "id": "openai/gpt-5.6-sol", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-sol", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "programbench:almost:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.programbench.almost", + "source_data": { + "dataset_name": "ProgramBench - Almost Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "almost", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Almost Resolved).", + "metric_id": "vals_ai.programbench.almost.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 23.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "almost", + "task_name": "Almost Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "23", + "raw_stderr": "2.983", + "latency": "1764.488", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.983, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "programbench:overall:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.programbench.overall", + "source_data": { + "dataset_name": "ProgramBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Overall).", + "metric_id": "vals_ai.programbench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 1.5, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "1.5", + "raw_stderr": "0.862", + "latency": "1764.488", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.862, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "programbench:partial:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.programbench.partial", + "source_data": { + "dataset_name": "ProgramBench - Raw Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "partial", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Raw Pass Rate).", + "metric_id": "vals_ai.programbench.partial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 77.644, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "partial", + "task_name": "Raw Pass Rate", + "dataset_type": "public", + "industry": "coding", + "raw_score": "77.644", + "raw_stderr": "1.689", + "latency": "1764.488", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.689, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "programbench:strict:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.programbench.strict", + "source_data": { + "dataset_name": "ProgramBench - Fully Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "strict", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Fully Resolved).", + "metric_id": "vals_ai.programbench.strict.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 1.5, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "strict", + "task_name": "Fully Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "1.5", + "raw_stderr": "0.862", + "latency": "1764.488", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.862, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-sol/7c6d07b1-03a3-48f4-88f5-44598748846d.json b/data/vals_ai/openai/gpt-5.6-sol/7c6d07b1-03a3-48f4-88f5-44598748846d.json new file mode 100644 index 0000000000000000000000000000000000000000..8d495374790add80af7a6892ae5104c7d372132d --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-sol/7c6d07b1-03a3-48f4-88f5-44598748846d.json @@ -0,0 +1,408 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/openai_gpt-5.6-sol/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-sol", + "id": "openai/gpt-5.6-sol", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-sol", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 66.434, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "66.434", + "raw_stderr": "1.612", + "latency": "20.315", + "cost_per_test": "0.367275", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.612, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 54.955, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "54.955", + "raw_stderr": "0.439", + "latency": "1104.076", + "cost_per_test": "1.180225", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.439, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_index:overall:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 72.63, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "72.63", + "raw_stderr": "1.071", + "latency": "748.418", + "cost_per_test": "7.425917", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.071, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 99.02, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "99.02", + "raw_stderr": "0.976", + "latency": "237.142", + "cost_per_test": "1.311279" + }, + "uncertainty": { + "standard_error": { + "value": 0.976, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 85.768, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "85.768", + "raw_stderr": "1.35", + "latency": "372.819", + "cost_per_test": "1.023319", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.35, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 86.968, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "86.968", + "raw_stderr": "4.544", + "latency": "2007.736", + "cost_per_test": "33.247485", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.544, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-sol/7e02eccf-5335-42cc-868b-9ea3329d3357.json b/data/vals_ai/openai/gpt-5.6-sol/7e02eccf-5335-42cc-868b-9ea3329d3357.json new file mode 100644 index 0000000000000000000000000000000000000000..ea9f3b80ed6b3987fcc0a6cc5ce1f65b23b550cf --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-sol/7e02eccf-5335-42cc-868b-9ea3329d3357.json @@ -0,0 +1,676 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/emb/openai_gpt-5.6-sol/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - EMB", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-sol", + "id": "openai/gpt-5.6-sol", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-sol", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "emb:comps:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.emb.comps", + "source_data": { + "dataset_name": "EMB - Comps", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "comps", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Comps).", + "metric_id": "vals_ai.emb.comps.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 71.309, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "comps", + "task_name": "Comps", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.309", + "raw_stderr": "6.203", + "latency": "814.317", + "cost_per_test": "9.611009", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.203, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:dataroom:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.emb.dataroom", + "source_data": { + "dataset_name": "EMB - Dataroom", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dataroom", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Dataroom).", + "metric_id": "vals_ai.emb.dataroom.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 85.243, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dataroom", + "task_name": "Dataroom", + "dataset_type": "private", + "industry": "finance", + "raw_score": "85.243", + "raw_stderr": "3.94", + "latency": "539.538", + "cost_per_test": "4.744315", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.94, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:dcf:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.emb.dcf", + "source_data": { + "dataset_name": "EMB - DCF", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dcf", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (DCF).", + "metric_id": "vals_ai.emb.dcf.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 62.764, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dcf", + "task_name": "DCF", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.764", + "raw_stderr": "7.225", + "latency": "510.322", + "cost_per_test": "3.850982", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.225, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:lbo:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.emb.lbo", + "source_data": { + "dataset_name": "EMB - LBO", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "lbo", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (LBO).", + "metric_id": "vals_ai.emb.lbo.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 67.242, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "lbo", + "task_name": "LBO", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.242", + "raw_stderr": "4.87", + "latency": "725.253", + "cost_per_test": "11.104033", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.87, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:ma:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.emb.ma", + "source_data": { + "dataset_name": "EMB - M&A", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "ma", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (M&A).", + "metric_id": "vals_ai.emb.ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 83.873, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "ma", + "task_name": "M&A", + "dataset_type": "private", + "industry": "finance", + "raw_score": "83.873", + "raw_stderr": "2.837", + "latency": "511.166", + "cost_per_test": "3.781208", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.837, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:misc:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.emb.misc", + "source_data": { + "dataset_name": "EMB - Misc", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "misc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Misc).", + "metric_id": "vals_ai.emb.misc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 68.681, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "misc", + "task_name": "Misc", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.681", + "raw_stderr": "3.132", + "latency": "545.114", + "cost_per_test": "3.282938", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.132, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:operating:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.emb.operating", + "source_data": { + "dataset_name": "EMB - Operating", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "operating", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Operating).", + "metric_id": "vals_ai.emb.operating.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 67.257, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "operating", + "task_name": "Operating", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.257", + "raw_stderr": "6.512", + "latency": "448.591", + "cost_per_test": "4.311989", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.512, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:overall:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.emb.overall", + "source_data": { + "dataset_name": "EMB - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Overall).", + "metric_id": "vals_ai.emb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 72.339, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "72.339", + "raw_stderr": "2.373", + "latency": "590.258", + "cost_per_test": "6.010773", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.373, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:scratch:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.emb.scratch", + "source_data": { + "dataset_name": "EMB - Scratch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "scratch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Scratch).", + "metric_id": "vals_ai.emb.scratch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 69.078, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "scratch", + "task_name": "Scratch", + "dataset_type": "private", + "industry": "finance", + "raw_score": "69.078", + "raw_stderr": "2.916", + "latency": "622.321", + "cost_per_test": "4.126546", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.916, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:template:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.emb.template", + "source_data": { + "dataset_name": "EMB - Template", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "template", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Template).", + "metric_id": "vals_ai.emb.template.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 75.303, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "template", + "task_name": "Template", + "dataset_type": "private", + "industry": "finance", + "raw_score": "75.303", + "raw_stderr": "3.068", + "latency": "558.195", + "cost_per_test": "7.895", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.068, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-sol/82c35c74-01c8-4dc5-a9eb-101ee731323d.json b/data/vals_ai/openai/gpt-5.6-sol/82c35c74-01c8-4dc5-a9eb-101ee731323d.json new file mode 100644 index 0000000000000000000000000000000000000000..0f63e5574204d80838ae1da2328f7ac8fd5a947c --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-sol/82c35c74-01c8-4dc5-a9eb-101ee731323d.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/openai_gpt-5.6-sol/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-sol", + "id": "openai/gpt-5.6-sol", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-sol", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 75.119, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "75.119", + "raw_stderr": "1.219", + "latency": "15.561", + "cost_per_test": "0.079634", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.219, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 67.289, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.289", + "raw_stderr": "0.922", + "latency": "12.618", + "cost_per_test": "0.071198", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.922, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 59.459, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.459", + "raw_stderr": "1.384", + "latency": "9.674", + "cost_per_test": "0.062761", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.384, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-sol/834ff301-7c94-4a3b-aa8a-745718231577.json b/data/vals_ai/openai/gpt-5.6-sol/834ff301-7c94-4a3b-aa8a-745718231577.json new file mode 100644 index 0000000000000000000000000000000000000000..469adc15a52896cd75ea9a1867f934d1605ab6e1 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-sol/834ff301-7c94-4a3b-aa8a-745718231577.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/openai_gpt-5.6-sol/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-sol", + "id": "openai/gpt-5.6-sol", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-sol", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 91.667, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "91.667", + "raw_stderr": "8.333", + "latency": "8295.214", + "cost_per_test": "22.768842", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 8.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 84.444, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "84.444", + "raw_stderr": "2.94", + "latency": "1106.029", + "cost_per_test": "3.035846", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.94, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 86.061, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "86.061", + "raw_stderr": "2.185", + "latency": "603.288", + "cost_per_test": "1.655916", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.185, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 85.768, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "85.768", + "raw_stderr": "1.35", + "latency": "372.819", + "cost_per_test": "1.023319", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.35, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-sol/8f531481-9231-4c9c-807b-8006b1af4380.json b/data/vals_ai/openai/gpt-5.6-sol/8f531481-9231-4c9c-807b-8006b1af4380.json new file mode 100644 index 0000000000000000000000000000000000000000..5b666c1fa16e8e12fff77e277669f55c14a99c71 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-sol/8f531481-9231-4c9c-807b-8006b1af4380.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/openai_gpt-5.6-sol/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-sol", + "id": "openai/gpt-5.6-sol", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-sol", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 85.0, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "85", + "raw_stderr": "13.693", + "latency": "1314.969", + "cost_per_test": "3.002041", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 13.693, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 88.333, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "88.333", + "raw_stderr": "10.65", + "latency": "1306.389", + "cost_per_test": "2.96886", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 10.65, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "ioi:overall:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 86.667, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "86.667", + "raw_stderr": "1.667", + "latency": "1310.679", + "cost_per_test": "2.98545", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.667, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-sol/a50013db-e503-4c01-8cf1-91ab4fe368c8.json b/data/vals_ai/openai/gpt-5.6-sol/a50013db-e503-4c01-8cf1-91ab4fe368c8.json new file mode 100644 index 0000000000000000000000000000000000000000..ce21ab3a3aab2825cfbbdbf0ee272a56216bf5af --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-sol/a50013db-e503-4c01-8cf1-91ab4fe368c8.json @@ -0,0 +1,195 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/cyber/openai_gpt-5.6-sol/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CyberBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "beta", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-sol", + "id": "openai/gpt-5.6-sol", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-sol", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "cyber:overall:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.cyber.overall", + "source_data": { + "dataset_name": "CyberBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (Overall).", + "metric_id": "vals_ai.cyber.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 88.136, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "beta", + "raw_score": "88.136", + "raw_stderr": "4.079", + "latency": "536.987", + "cost_per_test": "1.146805", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.079, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "cyber:patch:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.cyber.patch", + "source_data": { + "dataset_name": "CyberBench - Patch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "patch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (Patch).", + "metric_id": "vals_ai.cyber.patch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 83.051, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "patch", + "task_name": "Patch", + "dataset_type": "private", + "industry": "beta", + "raw_score": "83.051", + "raw_stderr": "4.884", + "latency": "875.103", + "cost_per_test": "1.281995", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.884, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "cyber:poc:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.cyber.poc", + "source_data": { + "dataset_name": "CyberBench - PoC", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "poc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (PoC).", + "metric_id": "vals_ai.cyber.poc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 93.22, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "poc", + "task_name": "PoC", + "dataset_type": "private", + "industry": "beta", + "raw_score": "93.22", + "raw_stderr": "3.273", + "latency": "198.872", + "cost_per_test": "1.011615", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.273, + "method": "vals_reported" + } + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-sol/b63e94e3-2e50-4dce-9157-860886072d28.json b/data/vals_ai/openai/gpt-5.6-sol/b63e94e3-2e50-4dce-9157-860886072d28.json new file mode 100644 index 0000000000000000000000000000000000000000..d3fab02eb4685b83392336db45ddac4b9537ebd8 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-sol/b63e94e3-2e50-4dce-9157-860886072d28.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/openai_gpt-5.6-sol/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-sol", + "id": "openai/gpt-5.6-sol", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-sol", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 80.495, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "80.495", + "raw_stderr": "3.716", + "latency": "2033.047", + "cost_per_test": "33.403634", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.716, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-sol/c0d4d236-c7a2-452b-a886-9512ad87bad9.json b/data/vals_ai/openai/gpt-5.6-sol/c0d4d236-c7a2-452b-a886-9512ad87bad9.json new file mode 100644 index 0000000000000000000000000000000000000000..c81d208cc9715c009e54178d66061d70e91d18e4 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-sol/c0d4d236-c7a2-452b-a886-9512ad87bad9.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/code-migration/openai_gpt-5.6-sol/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Code Migration", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-sol", + "id": "openai/gpt-5.6-sol", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-sol", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "code-migration:cli:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.code-migration.cli", + "source_data": { + "dataset_name": "Code Migration - CLI", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cli", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (CLI).", + "metric_id": "vals_ai.code-migration.cli.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 47.221, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cli", + "task_name": "CLI", + "dataset_type": "private", + "industry": "coding", + "raw_score": "47.221", + "raw_stderr": "2.786", + "latency": "4369.049", + "cost_per_test": "31.186416", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.786, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "code-migration:cobol:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.code-migration.cobol", + "source_data": { + "dataset_name": "Code Migration - COBOL", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cobol", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (COBOL).", + "metric_id": "vals_ai.code-migration.cobol.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 70.0, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cobol", + "task_name": "COBOL", + "dataset_type": "private", + "industry": "coding", + "raw_score": "70", + "raw_stderr": "15.275", + "latency": "1100.355", + "cost_per_test": "4.606544", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 15.275, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "code-migration:code_quality:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.code-migration.code_quality", + "source_data": { + "dataset_name": "Code Migration - Code Quality", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "code_quality", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Code Quality).", + "metric_id": "vals_ai.code-migration.code_quality.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 3.847, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "code_quality", + "task_name": "Code Quality", + "dataset_type": "private", + "industry": "coding", + "raw_score": "3.847", + "raw_stderr": "0.06", + "latency": "3551.876", + "cost_per_test": "24.541448", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.06, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "code-migration:overall:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.code-migration.overall", + "source_data": { + "dataset_name": "Code Migration - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Overall).", + "metric_id": "vals_ai.code-migration.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 52.916, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "52.916", + "raw_stderr": "4.353", + "latency": "3551.876", + "cost_per_test": "24.541448", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.353, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-sol/ccc05cf1-30d3-429a-a856-74ee247cb4ce.json b/data/vals_ai/openai/gpt-5.6-sol/ccc05cf1-30d3-429a-a856-74ee247cb4ce.json new file mode 100644 index 0000000000000000000000000000000000000000..b6b588d2a57394aa3f2ec528d87d8d0e68a2324a --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-sol/ccc05cf1-30d3-429a-a856-74ee247cb4ce.json @@ -0,0 +1,1188 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_research/openai_gpt-5.6-sol/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Legal Research Bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-sol", + "id": "openai/gpt-5.6-sol", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-sol", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_research:administrative_regulatory:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.legal_research.administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 60.606, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "60.606", + "raw_stderr": "6.061", + "latency": "4138.255", + "cost_per_test": "19.689902", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.061, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:business_commercial:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.legal_research.business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 53.488, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "53.488", + "raw_stderr": "7.696", + "latency": "4581.547", + "cost_per_test": "19.023356", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.696, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:civil_litigation:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.legal_research.civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 40.741, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "40.741", + "raw_stderr": "9.636", + "latency": "3816.903", + "cost_per_test": "18.560919", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 9.636, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:constitutional_civil_rights:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.legal_research.constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 34.211, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "34.211", + "raw_stderr": "7.799", + "latency": "3868.674", + "cost_per_test": "14.526185", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.799, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:criminal:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.legal_research.criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 43.86, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "43.86", + "raw_stderr": "6.631", + "latency": "4987.986", + "cost_per_test": "24.12378", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.631, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:family:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.legal_research.family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 13.636, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "13.636", + "raw_stderr": "7.489", + "latency": "6569.663", + "cost_per_test": "30.084962", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.489, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:health:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.legal_research.health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 80.0, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "80", + "raw_stderr": "7.428", + "latency": "4151.253", + "cost_per_test": "22.836198", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.428, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:immigration:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.legal_research.immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 45.455, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "45.455", + "raw_stderr": "15.746", + "latency": "5565.396", + "cost_per_test": "29.782421", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 15.746, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:overall:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.legal_research.overall", + "source_data": { + "dataset_name": "Legal Research Bench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Overall).", + "metric_id": "vals_ai.legal_research.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 48.077, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "legal", + "raw_score": "48.077", + "raw_stderr": "3.473", + "latency": "4631.383", + "cost_per_test": "21.605566", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.473, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass", + "source_data": { + "dataset_name": "Legal Research Bench - Weighted Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Weighted Pass).", + "metric_id": "vals_ai.legal_research.weighted_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 87.816, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass", + "task_name": "Weighted Pass", + "dataset_type": "private", + "industry": "legal", + "raw_score": "87.816", + "raw_stderr": "1.262", + "latency": "4631.383", + "cost_per_test": "21.605566", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.262, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__administrative_regulatory:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.weighted_pass__administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 90.69, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "90.69", + "raw_stderr": "2.118", + "latency": "4138.255", + "cost_per_test": "19.689902", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.118, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__business_commercial:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.weighted_pass__business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 89.003, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "89.003", + "raw_stderr": "2.864", + "latency": "4581.547", + "cost_per_test": "19.023356", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.864, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__civil_litigation:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.weighted_pass__civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 90.225, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "90.225", + "raw_stderr": "2.781", + "latency": "3816.903", + "cost_per_test": "18.560919", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.781, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__constitutional_civil_rights:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 88.31, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "88.31", + "raw_stderr": "1.827", + "latency": "3868.674", + "cost_per_test": "14.526185", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.827, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__criminal:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.weighted_pass__criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 86.381, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "86.381", + "raw_stderr": "2.307", + "latency": "4987.986", + "cost_per_test": "24.12378", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.307, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__family:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.weighted_pass__family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 76.736, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "76.736", + "raw_stderr": "4.515", + "latency": "6569.663", + "cost_per_test": "30.084962", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.515, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__health:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.weighted_pass__health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 97.511, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "97.511", + "raw_stderr": "1.194", + "latency": "4151.253", + "cost_per_test": "22.836198", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.194, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__immigration:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.weighted_pass__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 81.155, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "81.155", + "raw_stderr": "7.633", + "latency": "5565.396", + "cost_per_test": "29.782421", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.633, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-sol/dc3e870e-a191-49bb-bf7d-b0afcd84b248.json b/data/vals_ai/openai/gpt-5.6-sol/dc3e870e-a191-49bb-bf7d-b0afcd84b248.json new file mode 100644 index 0000000000000000000000000000000000000000..bbfe0a2c394afce036dc210a847dee0d49050c00 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-sol/dc3e870e-a191-49bb-bf7d-b0afcd84b248.json @@ -0,0 +1,3386 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/hlab/openai_gpt-5.6-sol/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Harvey's Legal Agent Benchmark", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-sol", + "id": "openai/gpt-5.6-sol", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-sol", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "hlab:antitrust_competition:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1172.092", + "cost_per_test": "8.940461", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:arbitration_international_dispute_resolution:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1371.899", + "cost_per_test": "17.046465", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:banking_finance:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1562.148", + "cost_per_test": "10.733229", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:bankruptcy_restructuring:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1473.251", + "cost_per_test": "13.525029", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:capital_markets:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1245.513", + "cost_per_test": "5.036098", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:corporate_governance:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1349.549", + "cost_per_test": "10.906051", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:corporate_ma:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1261.031", + "cost_per_test": "12.611383", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Criteria Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Criteria Pass Rate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 80.547, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate", + "task_name": "Criteria Pass Rate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.547", + "raw_stderr": "0.476", + "latency": "1370.186", + "cost_per_test": "10.374554", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.476, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__antitrust_competition:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 78.623, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.623", + "raw_stderr": "2.472", + "latency": "1172.092", + "cost_per_test": "8.940461", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.472, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__arbitration_international_dispute_resolution:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 82.079, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.079", + "raw_stderr": "2.3", + "latency": "1371.899", + "cost_per_test": "17.046465", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.3, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__banking_finance:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 81.413, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.413", + "raw_stderr": "2.376", + "latency": "1562.148", + "cost_per_test": "10.733229", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.376, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__bankruptcy_restructuring:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 72.664, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "72.664", + "raw_stderr": "2.626", + "latency": "1473.251", + "cost_per_test": "13.525029", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.626, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__capital_markets:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 71.372, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "71.372", + "raw_stderr": "2.836", + "latency": "1245.513", + "cost_per_test": "5.036098", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.836, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_governance:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 75.685, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "75.685", + "raw_stderr": "2.515", + "latency": "1349.549", + "cost_per_test": "10.906051", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.515, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_ma:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 85.577, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.577", + "raw_stderr": "1.992", + "latency": "1261.031", + "cost_per_test": "12.611383", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.992, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__data_privacy_cybersecurity:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 90.106, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.106", + "raw_stderr": "1.778", + "latency": "1177.994", + "cost_per_test": "9.289775", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.778, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__emerging_companies_venture_capital:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 85.324, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.324", + "raw_stderr": "2.071", + "latency": "1288.92", + "cost_per_test": "10.388772", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.071, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__employment_labor:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 67.041, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "67.041", + "raw_stderr": "2.882", + "latency": "1419.997", + "cost_per_test": "7.485275", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.882, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__energy_natural_resources:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 92.739, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "92.739", + "raw_stderr": "1.493", + "latency": "1722.813", + "cost_per_test": "17.774066", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.493, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__environmental_esg:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 80.645, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.645", + "raw_stderr": "2.37", + "latency": "1111.078", + "cost_per_test": "7.575723", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.37, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__funds_asset_management:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 74.927, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "74.927", + "raw_stderr": "2.344", + "latency": "3106.836", + "cost_per_test": "13.033612", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.344, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__healthcare_life_sciences:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 82.508, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.508", + "raw_stderr": "2.186", + "latency": "1594.308", + "cost_per_test": "11.601398", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.186, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__immigration:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 84.364, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.364", + "raw_stderr": "2.194", + "latency": "1017.421", + "cost_per_test": "5.823174", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.194, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__insurance:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 84.249, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.249", + "raw_stderr": "2.209", + "latency": "1239.26", + "cost_per_test": "8.5933", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.209, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__intellectual_property:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 85.512, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.512", + "raw_stderr": "2.096", + "latency": "1126.249", + "cost_per_test": "8.439215", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.096, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__international_trade_sanctions:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 80.556, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.556", + "raw_stderr": "2.498", + "latency": "1023.829", + "cost_per_test": "6.235021", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.498, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__litigation_dispute_resolution:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 83.895, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.895", + "raw_stderr": "2.254", + "latency": "907.362", + "cost_per_test": "7.56413", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.254, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__real_estate:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 79.009, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.009", + "raw_stderr": "2.202", + "latency": "1732.607", + "cost_per_test": "12.830219", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.202, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__structured_finance_securitization:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 79.365, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.365", + "raw_stderr": "2.284", + "latency": "1252.704", + "cost_per_test": "14.536468", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.284, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__tax:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 70.903, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "70.903", + "raw_stderr": "2.631", + "latency": "1159.828", + "cost_per_test": "10.025086", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.631, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__trusts_estates_private_client:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 86.26, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.26", + "raw_stderr": "2.131", + "latency": "1368.648", + "cost_per_test": "8.582153", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.131, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__white_collar_defense_investigations:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 78.451, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.451", + "raw_stderr": "2.39", + "latency": "1199.139", + "cost_per_test": "10.413192", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.39, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:data_privacy_cybersecurity:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1177.994", + "cost_per_test": "9.289775", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:emerging_companies_venture_capital:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1288.92", + "cost_per_test": "10.388772", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:employment_labor:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1419.997", + "cost_per_test": "7.485275", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:energy_natural_resources:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "1722.813", + "cost_per_test": "17.774066", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:environmental_esg:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1111.078", + "cost_per_test": "7.575723", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:funds_asset_management:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "3106.836", + "cost_per_test": "13.033612", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:healthcare_life_sciences:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1594.308", + "cost_per_test": "11.601398", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:immigration:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1017.421", + "cost_per_test": "5.823174", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:insurance:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1239.26", + "cost_per_test": "8.5933", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:intellectual_property:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1126.249", + "cost_per_test": "8.439215", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:international_trade_sanctions:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1023.829", + "cost_per_test": "6.235021", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:litigation_dispute_resolution:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "907.362", + "cost_per_test": "7.56413", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:overall:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.overall", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Overall).", + "metric_id": "vals_ai.hlab.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 2.5, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "2.5", + "raw_stderr": "0.833", + "latency": "1370.186", + "cost_per_test": "10.374554", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.833, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:real_estate:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1732.607", + "cost_per_test": "12.830219", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:structured_finance_securitization:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1252.704", + "cost_per_test": "14.536468", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:tax:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1159.828", + "cost_per_test": "10.025086", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:trusts_estates_private_client:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1368.648", + "cost_per_test": "8.582153", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:white_collar_defense_investigations:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.hlab.white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1199.139", + "cost_per_test": "10.413192", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-sol/ee060483-3142-477b-8a6a-5d57638c190c.json b/data/vals_ai/openai/gpt-5.6-sol/ee060483-3142-477b-8a6a-5d57638c190c.json new file mode 100644 index 0000000000000000000000000000000000000000..ae1e26223f5d4e0ae4fc0532a1ef9cb72bf07379 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-sol/ee060483-3142-477b-8a6a-5d57638c190c.json @@ -0,0 +1,740 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/openai_gpt-5.6-sol/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-sol", + "id": "openai/gpt-5.6-sol", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-sol", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 49.233, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "49.233", + "raw_stderr": "0.804", + "latency": "1135.714", + "cost_per_test": "0.941562", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.804, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 41.362, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "41.362", + "raw_stderr": "0.068", + "latency": "1165.258", + "cost_per_test": "1.253419", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.068, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 45.284, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "45.284", + "raw_stderr": "0.98", + "latency": "1405.294", + "cost_per_test": "1.258215", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.98, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 63.329, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.329", + "raw_stderr": "3.028", + "latency": "1405.72", + "cost_per_test": "1.27232", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.028, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 74.351, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "74.351", + "raw_stderr": "1.294", + "latency": "423.559", + "cost_per_test": "0.687883", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.294, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 20.706, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "20.706", + "raw_stderr": "0.501", + "latency": "968.534", + "cost_per_test": "1.476372", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.501, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 64.933, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.933", + "raw_stderr": "3.336", + "latency": "1605.836", + "cost_per_test": "1.147273", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.336, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 75.613, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "75.613", + "raw_stderr": "1.885", + "latency": "851.845", + "cost_per_test": "0.724133", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.885, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 64.163, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.163", + "raw_stderr": "1.612", + "latency": "875.032", + "cost_per_test": "1.828342", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.612, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:overall:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 53.756, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "53.756", + "raw_stderr": "0.846", + "latency": "1165.258", + "cost_per_test": "1.253419", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.846, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 26.192, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "26.192", + "raw_stderr": "0.847", + "latency": "1935.492", + "cost_per_test": "2.11729", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.847, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-sol/f17f8d6c-c8ae-41c9-ad6e-616a30606630.json b/data/vals_ai/openai/gpt-5.6-sol/f17f8d6c-c8ae-41c9-ad6e-616a30606630.json new file mode 100644 index 0000000000000000000000000000000000000000..11b69f278f189620f17f11e32b7ad1b7f26c8930 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-sol/f17f8d6c-c8ae-41c9-ad6e-616a30606630.json @@ -0,0 +1,438 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/openai_gpt-5.6-sol/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-sol", + "id": "openai/gpt-5.6-sol", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-sol", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 90.492, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.492", + "raw_stderr": "0.937", + "latency": "2.88", + "cost_per_test": "0.005618", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.937, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 84.398, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.398", + "raw_stderr": "0.203", + "latency": "7.666", + "cost_per_test": "0.01698", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.203, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 91.438, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "91.438", + "raw_stderr": "0.535", + "latency": "3.001", + "cost_per_test": "0.008536", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 86.965, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.965", + "raw_stderr": "0.412", + "latency": "6.198", + "cost_per_test": "0.012678", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.412, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.054, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.054", + "raw_stderr": "1.12", + "latency": "2.608", + "cost_per_test": "0.00629", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.12, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 86.443, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.443", + "raw_stderr": "1.335", + "latency": "14.835", + "cost_per_test": "0.025965", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.335, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-sol/f2fe2491-b092-4686-a77e-c0fc4f06dc92.json b/data/vals_ai/openai/gpt-5.6-sol/f2fe2491-b092-4686-a77e-c0fc4f06dc92.json new file mode 100644 index 0000000000000000000000000000000000000000..3eb1ad95213f6823f38d87d288acb24e57c9fe9c --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-sol/f2fe2491-b092-4686-a77e-c0fc4f06dc92.json @@ -0,0 +1,536 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_multimodal_index/openai_gpt-5.6-sol/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Multimodal Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-sol", + "id": "openai/gpt-5.6-sol", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-sol", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_multimodal_index:corp_fin_v2:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.vals_multimodal_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Multimodal Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (CorpFin v2).", + "metric_id": "vals_ai.vals_multimodal_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 66.434, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "66.434", + "raw_stderr": "1.612", + "latency": "20.315", + "cost_per_test": "0.367275", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.612, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:finance_agent:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.vals_multimodal_index.finance_agent", + "source_data": { + "dataset_name": "Vals Multimodal Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_multimodal_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 54.955, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "54.955", + "raw_stderr": "0.439", + "latency": "1104.076", + "cost_per_test": "1.180225", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.439, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:mortgage_tax:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.vals_multimodal_index.mortgage_tax", + "source_data": { + "dataset_name": "Vals Multimodal Index - Mortgage Tax", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "mortgage_tax", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Mortgage Tax).", + "metric_id": "vals_ai.vals_multimodal_index.mortgage_tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 67.289, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "mortgage_tax", + "task_name": "Mortgage Tax", + "dataset_type": "private", + "industry": "index", + "raw_score": "67.289", + "raw_stderr": "0.922", + "latency": "12.618", + "cost_per_test": "0.071198", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.922, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:overall:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.vals_multimodal_index.overall", + "source_data": { + "dataset_name": "Vals Multimodal Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Overall).", + "metric_id": "vals_ai.vals_multimodal_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 72.191, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "72.191", + "raw_stderr": "0.979", + "latency": "547.157", + "cost_per_test": "5.335537", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.979, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:sage:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.vals_multimodal_index.sage", + "source_data": { + "dataset_name": "Vals Multimodal Index - SAGE", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "sage", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SAGE).", + "metric_id": "vals_ai.vals_multimodal_index.sage.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 52.563, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "sage", + "task_name": "SAGE", + "dataset_type": "private", + "industry": "index", + "raw_score": "52.563", + "raw_stderr": "3.416", + "latency": "75.395", + "cost_per_test": "0.147977", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.416, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:swebench:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.vals_multimodal_index.swebench", + "source_data": { + "dataset_name": "Vals Multimodal Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SWE-bench).", + "metric_id": "vals_ai.vals_multimodal_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 99.02, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "99.02", + "raw_stderr": "0.976", + "latency": "237.142", + "cost_per_test": "1.311279" + }, + "uncertainty": { + "standard_error": { + "value": 0.976, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:terminal_bench_2_1:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.vals_multimodal_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Multimodal Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_multimodal_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 85.768, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "85.768", + "raw_stderr": "1.35", + "latency": "372.819", + "cost_per_test": "1.023319", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.35, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:vibe_code_bench:openai/gpt-5.6-sol:score", + "evaluation_name": "vals_ai.vals_multimodal_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Multimodal Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_multimodal_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 86.968, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "86.968", + "raw_stderr": "4.544", + "latency": "2007.736", + "cost_per_test": "33.247485", + "max_output_tokens": "128000", + "reasoning_effort": "max", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.544, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-terra/0003243d-a69c-4a5f-84ec-5802cb1e65bc.json b/data/vals_ai/openai/gpt-5.6-terra/0003243d-a69c-4a5f-84ec-5802cb1e65bc.json new file mode 100644 index 0000000000000000000000000000000000000000..185760a1894809b5f54b659d5bcd7c2891aded4e --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-terra/0003243d-a69c-4a5f-84ec-5802cb1e65bc.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/openai_gpt-5.6-terra/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-terra", + "id": "openai/gpt-5.6-terra", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-terra", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 82.869, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "82.869", + "raw_stderr": "1.948", + "latency": "35.526", + "cost_per_test": "0.078235", + "max_output_tokens": "30000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.948, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-terra/028a2282-70c5-4980-84d2-e721a8183c93.json b/data/vals_ai/openai/gpt-5.6-terra/028a2282-70c5-4980-84d2-e721a8183c93.json new file mode 100644 index 0000000000000000000000000000000000000000..78abff2d62ec04acccc6fd4bd43a5f5a94ee05cc --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-terra/028a2282-70c5-4980-84d2-e721a8183c93.json @@ -0,0 +1,1188 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_research/openai_gpt-5.6-terra/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Legal Research Bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-terra", + "id": "openai/gpt-5.6-terra", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-terra", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_research:administrative_regulatory:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.legal_research.administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 57.576, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "57.576", + "raw_stderr": "6.13", + "latency": "1182.969", + "cost_per_test": "7.823015", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.13, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:business_commercial:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.legal_research.business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 37.209, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "37.209", + "raw_stderr": "7.458", + "latency": "1332.727", + "cost_per_test": "7.297443", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.458, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:civil_litigation:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.legal_research.civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "33.333", + "raw_stderr": "9.245", + "latency": "1228.914", + "cost_per_test": "7.499439", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 9.245, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:constitutional_civil_rights:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.legal_research.constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 36.842, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "36.842", + "raw_stderr": "7.93", + "latency": "1076.949", + "cost_per_test": "6.26741", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.93, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:criminal:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.legal_research.criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 36.842, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "36.842", + "raw_stderr": "6.446", + "latency": "1356.442", + "cost_per_test": "7.827654", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.446, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:family:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.legal_research.family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 13.636, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "13.636", + "raw_stderr": "7.489", + "latency": "1423.695", + "cost_per_test": "8.612481", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.489, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:health:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.legal_research.health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 76.667, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "76.667", + "raw_stderr": "7.854", + "latency": "1166.454", + "cost_per_test": "9.203349", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.854, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:immigration:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.legal_research.immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 27.273, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "27.273", + "raw_stderr": "14.084", + "latency": "1616.189", + "cost_per_test": "9.017179", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 14.084, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:overall:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.legal_research.overall", + "source_data": { + "dataset_name": "Legal Research Bench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Overall).", + "metric_id": "vals_ai.legal_research.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 40.865, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "legal", + "raw_score": "40.865", + "raw_stderr": "3.417", + "latency": "1268.948", + "cost_per_test": "7.604427", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.417, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass", + "source_data": { + "dataset_name": "Legal Research Bench - Weighted Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Weighted Pass).", + "metric_id": "vals_ai.legal_research.weighted_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 84.033, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass", + "task_name": "Weighted Pass", + "dataset_type": "private", + "industry": "legal", + "raw_score": "84.033", + "raw_stderr": "1.408", + "latency": "1268.948", + "cost_per_test": "7.604427", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.408, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__administrative_regulatory:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.weighted_pass__administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 86.696, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "86.696", + "raw_stderr": "2.539", + "latency": "1182.969", + "cost_per_test": "7.823015", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.539, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__business_commercial:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.weighted_pass__business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 84.329, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "84.329", + "raw_stderr": "3.27", + "latency": "1332.727", + "cost_per_test": "7.297443", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.27, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__civil_litigation:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.weighted_pass__civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 88.495, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "88.495", + "raw_stderr": "2.917", + "latency": "1228.914", + "cost_per_test": "7.499439", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.917, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__constitutional_civil_rights:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 85.254, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "85.254", + "raw_stderr": "2.493", + "latency": "1076.949", + "cost_per_test": "6.26741", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.493, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__criminal:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.weighted_pass__criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 83.389, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "83.389", + "raw_stderr": "2.503", + "latency": "1356.442", + "cost_per_test": "7.827654", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.503, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__family:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.weighted_pass__family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 74.235, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "74.235", + "raw_stderr": "4.347", + "latency": "1423.695", + "cost_per_test": "8.612481", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.347, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__health:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.weighted_pass__health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 94.002, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "94.002", + "raw_stderr": "2.403", + "latency": "1166.454", + "cost_per_test": "9.203349", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.403, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__immigration:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.weighted_pass__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 80.122, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "80.122", + "raw_stderr": "7.615", + "latency": "1616.189", + "cost_per_test": "9.017179", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.615, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-terra/1048fca1-940e-4c5b-8062-b702a501eb58.json b/data/vals_ai/openai/gpt-5.6-terra/1048fca1-940e-4c5b-8062-b702a501eb58.json new file mode 100644 index 0000000000000000000000000000000000000000..57560906dc1c1ab88fd0edc496e66e549f9125c7 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-terra/1048fca1-940e-4c5b-8062-b702a501eb58.json @@ -0,0 +1,408 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/openai_gpt-5.6-terra/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-terra", + "id": "openai/gpt-5.6-terra", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-terra", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 66.434, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "66.434", + "raw_stderr": "1.612", + "latency": "9.294", + "cost_per_test": "0.345071", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.612, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 52.626, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "52.626", + "raw_stderr": "0.537", + "latency": "388.501", + "cost_per_test": "0.599116", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.537, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_index:overall:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 65.135, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "65.135", + "raw_stderr": "1.715", + "latency": "352.049", + "cost_per_test": "2.655165", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.715, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 73.529, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "73.529", + "raw_stderr": "4.368", + "latency": "142.056", + "cost_per_test": "0.803534" + }, + "uncertainty": { + "standard_error": { + "value": 4.368, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 73.408, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "73.408", + "raw_stderr": "2.085", + "latency": "260.9", + "cost_per_test": "0.712339", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.085, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 72.818, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "72.818", + "raw_stderr": "7.59", + "latency": "959.495", + "cost_per_test": "10.815764", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.59, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-terra/18f1f49c-5812-4338-bac0-4a1870e8a866.json b/data/vals_ai/openai/gpt-5.6-terra/18f1f49c-5812-4338-bac0-4a1870e8a866.json new file mode 100644 index 0000000000000000000000000000000000000000..9f8f84e2d113c26a59ac553041f48b40c8c4d774 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-terra/18f1f49c-5812-4338-bac0-4a1870e8a866.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/openai_gpt-5.6-terra/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-terra", + "id": "openai/gpt-5.6-terra", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-terra", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 67.842, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "67.842", + "raw_stderr": "5.103", + "latency": "959.495", + "cost_per_test": "10.815764", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.103, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-terra/28bd21b8-3007-410a-ad77-a4810f6cea2b.json b/data/vals_ai/openai/gpt-5.6-terra/28bd21b8-3007-410a-ad77-a4810f6cea2b.json new file mode 100644 index 0000000000000000000000000000000000000000..a102f7f762388a2304dec732496e471d91df3c63 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-terra/28bd21b8-3007-410a-ad77-a4810f6cea2b.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/openai_gpt-5.6-terra/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-terra", + "id": "openai/gpt-5.6-terra", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-terra", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 65.822, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.822", + "raw_stderr": "1.356", + "latency": "37.28", + "cost_per_test": "0.089773", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.356, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 76.166, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "76.166", + "raw_stderr": "0.836", + "latency": "37.28", + "cost_per_test": "0.089773", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.836, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 86.509, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "86.509", + "raw_stderr": "0.977", + "latency": "37.28", + "cost_per_test": "0.089773", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.977, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-terra/37824af8-6686-4774-a7f3-720f52fd9a93.json b/data/vals_ai/openai/gpt-5.6-terra/37824af8-6686-4774-a7f3-720f52fd9a93.json new file mode 100644 index 0000000000000000000000000000000000000000..b5e6a3060d6600bd0635a59321064a2249246e1c --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-terra/37824af8-6686-4774-a7f3-720f52fd9a93.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/programbench/openai_gpt-5.6-terra/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProgramBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-terra", + "id": "openai/gpt-5.6-terra", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-terra", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "programbench:almost:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.programbench.almost", + "source_data": { + "dataset_name": "ProgramBench - Almost Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "almost", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Almost Resolved).", + "metric_id": "vals_ai.programbench.almost.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 6.5, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "almost", + "task_name": "Almost Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "6.5", + "raw_stderr": "1.748", + "latency": "564.4", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.748, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "programbench:overall:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.programbench.overall", + "source_data": { + "dataset_name": "ProgramBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Overall).", + "metric_id": "vals_ai.programbench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.5, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0.5", + "raw_stderr": "0.5", + "latency": "564.4", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.5, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "programbench:partial:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.programbench.partial", + "source_data": { + "dataset_name": "ProgramBench - Raw Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "partial", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Raw Pass Rate).", + "metric_id": "vals_ai.programbench.partial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 66.313, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "partial", + "task_name": "Raw Pass Rate", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66.313", + "raw_stderr": "1.8", + "latency": "564.4", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.8, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "programbench:strict:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.programbench.strict", + "source_data": { + "dataset_name": "ProgramBench - Fully Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "strict", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Fully Resolved).", + "metric_id": "vals_ai.programbench.strict.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.5, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "strict", + "task_name": "Fully Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0.5", + "raw_stderr": "0.5", + "latency": "564.4", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.5, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-terra/39d5a6fe-ad89-4f41-95e1-141512c09dde.json b/data/vals_ai/openai/gpt-5.6-terra/39d5a6fe-ad89-4f41-95e1-141512c09dde.json new file mode 100644 index 0000000000000000000000000000000000000000..70e8917b7eba09ddaa7868d1d0955dd3e82d408b --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-terra/39d5a6fe-ad89-4f41-95e1-141512c09dde.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/code-migration/openai_gpt-5.6-terra/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Code Migration", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-terra", + "id": "openai/gpt-5.6-terra", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-terra", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "code-migration:cli:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.code-migration.cli", + "source_data": { + "dataset_name": "Code Migration - CLI", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cli", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (CLI).", + "metric_id": "vals_ai.code-migration.cli.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 26.645, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cli", + "task_name": "CLI", + "dataset_type": "private", + "industry": "coding", + "raw_score": "26.645", + "raw_stderr": "2.18", + "latency": "1281.86", + "cost_per_test": "2.70452", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.18, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "code-migration:cobol:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.code-migration.cobol", + "source_data": { + "dataset_name": "Code Migration - COBOL", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cobol", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (COBOL).", + "metric_id": "vals_ai.code-migration.cobol.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 65.844, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cobol", + "task_name": "COBOL", + "dataset_type": "private", + "industry": "coding", + "raw_score": "65.844", + "raw_stderr": "14.637", + "latency": "537.072", + "cost_per_test": "1.174847", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 14.637, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "code-migration:code_quality:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.code-migration.code_quality", + "source_data": { + "dataset_name": "Code Migration - Code Quality", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "code_quality", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Code Quality).", + "metric_id": "vals_ai.code-migration.code_quality.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 3.305, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "code_quality", + "task_name": "Code Quality", + "dataset_type": "private", + "industry": "coding", + "raw_score": "3.305", + "raw_stderr": "0.059", + "latency": "1095.663", + "cost_per_test": "2.322102", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.059, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "code-migration:overall:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.code-migration.overall", + "source_data": { + "dataset_name": "Code Migration - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Overall).", + "metric_id": "vals_ai.code-migration.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 36.445, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "36.445", + "raw_stderr": "4.008", + "latency": "1095.663", + "cost_per_test": "2.322102", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.008, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-terra/451ccc6c-7d4a-418c-aa42-4b77aa040982.json b/data/vals_ai/openai/gpt-5.6-terra/451ccc6c-7d4a-418c-aa42-4b77aa040982.json new file mode 100644 index 0000000000000000000000000000000000000000..296eccabd8be861c4b35c125a8d8f1f79066d3c7 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-terra/451ccc6c-7d4a-418c-aa42-4b77aa040982.json @@ -0,0 +1,536 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_multimodal_index/openai_gpt-5.6-terra/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Multimodal Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-terra", + "id": "openai/gpt-5.6-terra", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-terra", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_multimodal_index:corp_fin_v2:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.vals_multimodal_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Multimodal Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (CorpFin v2).", + "metric_id": "vals_ai.vals_multimodal_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 66.434, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "66.434", + "raw_stderr": "1.612", + "latency": "9.294", + "cost_per_test": "0.345071", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.612, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:finance_agent:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.vals_multimodal_index.finance_agent", + "source_data": { + "dataset_name": "Vals Multimodal Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_multimodal_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 52.626, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "52.626", + "raw_stderr": "0.537", + "latency": "388.501", + "cost_per_test": "0.599116", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.537, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:mortgage_tax:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.vals_multimodal_index.mortgage_tax", + "source_data": { + "dataset_name": "Vals Multimodal Index - Mortgage Tax", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "mortgage_tax", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Mortgage Tax).", + "metric_id": "vals_ai.vals_multimodal_index.mortgage_tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 67.329, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "mortgage_tax", + "task_name": "Mortgage Tax", + "dataset_type": "private", + "industry": "index", + "raw_score": "67.329", + "raw_stderr": "0.925", + "latency": "6.316", + "cost_per_test": "0.056764", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.925, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:overall:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.vals_multimodal_index.overall", + "source_data": { + "dataset_name": "Vals Multimodal Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Overall).", + "metric_id": "vals_ai.vals_multimodal_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 65.07, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "65.07", + "raw_stderr": "1.571", + "latency": "258.088", + "cost_per_test": "1.918153", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.571, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:sage:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.vals_multimodal_index.sage", + "source_data": { + "dataset_name": "Vals Multimodal Index - SAGE", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "sage", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SAGE).", + "metric_id": "vals_ai.vals_multimodal_index.sage.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 47.003, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "sage", + "task_name": "SAGE", + "dataset_type": "private", + "industry": "index", + "raw_score": "47.003", + "raw_stderr": "3.403", + "latency": "40.057", + "cost_per_test": "0.094483", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.403, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:swebench:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.vals_multimodal_index.swebench", + "source_data": { + "dataset_name": "Vals Multimodal Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SWE-bench).", + "metric_id": "vals_ai.vals_multimodal_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 73.529, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "73.529", + "raw_stderr": "4.368", + "latency": "142.056", + "cost_per_test": "0.803534" + }, + "uncertainty": { + "standard_error": { + "value": 4.368, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:terminal_bench_2_1:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.vals_multimodal_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Multimodal Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_multimodal_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 73.408, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "73.408", + "raw_stderr": "2.085", + "latency": "260.9", + "cost_per_test": "0.712339", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.085, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:vibe_code_bench:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.vals_multimodal_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Multimodal Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_multimodal_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 72.818, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "72.818", + "raw_stderr": "7.59", + "latency": "959.495", + "cost_per_test": "10.815764", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.59, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-terra/559fefac-74f4-4fe2-960b-f6c58f9453bc.json b/data/vals_ai/openai/gpt-5.6-terra/559fefac-74f4-4fe2-960b-f6c58f9453bc.json new file mode 100644 index 0000000000000000000000000000000000000000..688358b3bae819a915862280eda013b9bd127ea1 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-terra/559fefac-74f4-4fe2-960b-f6c58f9453bc.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medcode/openai_gpt-5.6-terra/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedCode", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medcode", + "benchmark_name": "MedCode", + "benchmark_updated": "2026-07-08", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-terra", + "id": "openai/gpt-5.6-terra", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-terra", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medcode:overall:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.medcode.overall", + "source_data": { + "dataset_name": "MedCode - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medcode", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedCode (Overall).", + "metric_id": "vals_ai.medcode.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode" + } + }, + "score_details": { + "score": 43.414, + "details": { + "benchmark_slug": "medcode", + "benchmark_name": "MedCode", + "benchmark_updated": "2026-07-08", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "43.414", + "raw_stderr": "2.173", + "latency": "18.411", + "cost_per_test": "0.058197", + "max_output_tokens": "30000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.173, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-terra/5b02ce76-156e-476d-9d69-97ef126eabba.json b/data/vals_ai/openai/gpt-5.6-terra/5b02ce76-156e-476d-9d69-97ef126eabba.json new file mode 100644 index 0000000000000000000000000000000000000000..c32fd95a0809cadff89b3ab87ebbdfcc3417c2c3 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-terra/5b02ce76-156e-476d-9d69-97ef126eabba.json @@ -0,0 +1,3386 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/hlab/openai_gpt-5.6-terra/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Harvey's Legal Agent Benchmark", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-terra", + "id": "openai/gpt-5.6-terra", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-terra", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "hlab:antitrust_competition:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "347.997", + "cost_per_test": "1.997694", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:arbitration_international_dispute_resolution:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "360.323", + "cost_per_test": "2.631968", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:banking_finance:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "369.987", + "cost_per_test": "2.118158", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:bankruptcy_restructuring:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "463.625", + "cost_per_test": "2.645725", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:capital_markets:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "285.363", + "cost_per_test": "1.437455", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:corporate_governance:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "364.882", + "cost_per_test": "2.12016", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:corporate_ma:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "354.82", + "cost_per_test": "1.940389", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Criteria Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Criteria Pass Rate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 78.477, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate", + "task_name": "Criteria Pass Rate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.477", + "raw_stderr": "0.494", + "latency": "368.79", + "cost_per_test": "1.980069", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.494, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__antitrust_competition:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 76.449, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "76.449", + "raw_stderr": "2.559", + "latency": "347.997", + "cost_per_test": "1.997694", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.559, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__arbitration_international_dispute_resolution:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 74.552, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "74.552", + "raw_stderr": "2.612", + "latency": "360.323", + "cost_per_test": "2.631968", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.612, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__banking_finance:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 83.643, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.643", + "raw_stderr": "2.259", + "latency": "369.987", + "cost_per_test": "2.118158", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.259, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__bankruptcy_restructuring:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 75.432, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "75.432", + "raw_stderr": "2.537", + "latency": "463.625", + "cost_per_test": "2.645725", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.537, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__capital_markets:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 85.882, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.882", + "raw_stderr": "2.185", + "latency": "285.363", + "cost_per_test": "1.437455", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.185, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_governance:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 75.342, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "75.342", + "raw_stderr": "2.527", + "latency": "364.882", + "cost_per_test": "2.12016", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.527, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_ma:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 83.013, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.013", + "raw_stderr": "2.129", + "latency": "354.82", + "cost_per_test": "1.940389", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.129, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__data_privacy_cybersecurity:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 86.219, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.219", + "raw_stderr": "2.053", + "latency": "365.331", + "cost_per_test": "2.276851", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.053, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__emerging_companies_venture_capital:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 81.911, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.911", + "raw_stderr": "2.253", + "latency": "399.866", + "cost_per_test": "1.784431", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.253, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__employment_labor:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 73.408, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "73.408", + "raw_stderr": "2.709", + "latency": "364.527", + "cost_per_test": "1.566207", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.709, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__energy_natural_resources:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 87.789, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.789", + "raw_stderr": "1.884", + "latency": "418.089", + "cost_per_test": "3.130221", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.884, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__environmental_esg:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 73.477, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "73.477", + "raw_stderr": "2.648", + "latency": "337.746", + "cost_per_test": "1.424326", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.648, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__funds_asset_management:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 72.886, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "72.886", + "raw_stderr": "2.404", + "latency": "472.316", + "cost_per_test": "2.513431", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.404, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__healthcare_life_sciences:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 73.597, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "73.597", + "raw_stderr": "2.537", + "latency": "484.054", + "cost_per_test": "2.268041", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.537, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__immigration:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 73.818, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "73.818", + "raw_stderr": "2.656", + "latency": "337.706", + "cost_per_test": "1.450882", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.656, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__insurance:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 78.022, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.022", + "raw_stderr": "2.511", + "latency": "333.322", + "cost_per_test": "1.401004", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.511, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__intellectual_property:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 86.219, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.219", + "raw_stderr": "2.053", + "latency": "330.588", + "cost_per_test": "1.46568", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.053, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__international_trade_sanctions:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 76.19, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "76.19", + "raw_stderr": "2.688", + "latency": "318.425", + "cost_per_test": "1.162607", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.688, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__litigation_dispute_resolution:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 77.154, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.154", + "raw_stderr": "2.574", + "latency": "403.154", + "cost_per_test": "2.523421", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.574, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__real_estate:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 88.63, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.63", + "raw_stderr": "1.717", + "latency": "375.414", + "cost_per_test": "2.32119", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.717, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__structured_finance_securitization:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 76.825, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "76.825", + "raw_stderr": "2.381", + "latency": "306.878", + "cost_per_test": "1.47357", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.381, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__tax:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 69.9, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "69.9", + "raw_stderr": "2.657", + "latency": "407.619", + "cost_per_test": "2.663856", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.657, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__trusts_estates_private_client:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 82.824, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.824", + "raw_stderr": "2.335", + "latency": "352.622", + "cost_per_test": "1.816456", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.335, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__white_collar_defense_investigations:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 70.034, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "70.034", + "raw_stderr": "2.663", + "latency": "296.298", + "cost_per_test": "1.387928", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.663, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:data_privacy_cybersecurity:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "365.331", + "cost_per_test": "2.276851", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:emerging_companies_venture_capital:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "399.866", + "cost_per_test": "1.784431", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:employment_labor:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "364.527", + "cost_per_test": "1.566207", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:energy_natural_resources:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "418.089", + "cost_per_test": "3.130221", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:environmental_esg:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "337.746", + "cost_per_test": "1.424326", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:funds_asset_management:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "472.316", + "cost_per_test": "2.513431", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:healthcare_life_sciences:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "484.054", + "cost_per_test": "2.268041", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:immigration:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "337.706", + "cost_per_test": "1.450882", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:insurance:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "333.322", + "cost_per_test": "1.401004", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:intellectual_property:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "330.588", + "cost_per_test": "1.46568", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:international_trade_sanctions:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "318.425", + "cost_per_test": "1.162607", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:litigation_dispute_resolution:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "403.154", + "cost_per_test": "2.523421", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:overall:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.overall", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Overall).", + "metric_id": "vals_ai.hlab.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.417, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0.417", + "raw_stderr": "0", + "latency": "368.79", + "cost_per_test": "1.980069", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:real_estate:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "375.414", + "cost_per_test": "2.32119", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:structured_finance_securitization:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "306.878", + "cost_per_test": "1.47357", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:tax:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "407.619", + "cost_per_test": "2.663856", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:trusts_estates_private_client:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "352.622", + "cost_per_test": "1.816456", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "hlab:white_collar_defense_investigations:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.hlab.white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "296.298", + "cost_per_test": "1.387928", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-terra/6255f8a0-dcc6-4e26-b345-d586a3ac8610.json b/data/vals_ai/openai/gpt-5.6-terra/6255f8a0-dcc6-4e26-b345-d586a3ac8610.json new file mode 100644 index 0000000000000000000000000000000000000000..cbbfa1c2207ad1ef705ba5994c7d2802662db504 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-terra/6255f8a0-dcc6-4e26-b345-d586a3ac8610.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/openai_gpt-5.6-terra/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-terra", + "id": "openai/gpt-5.6-terra", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-terra", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 74.324, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "74.324", + "raw_stderr": "1.232", + "latency": "7.021", + "cost_per_test": "0.058489", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.232, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 67.329, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.329", + "raw_stderr": "0.925", + "latency": "6.316", + "cost_per_test": "0.056764", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.925, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 60.334, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.334", + "raw_stderr": "1.379", + "latency": "5.612", + "cost_per_test": "0.05504", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.379, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-terra/6e49d05f-f619-4666-a518-328e060f2f8c.json b/data/vals_ai/openai/gpt-5.6-terra/6e49d05f-f619-4666-a518-328e060f2f8c.json new file mode 100644 index 0000000000000000000000000000000000000000..d592d2c6cc0ec9552e10b4768ae4606362a08517 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-terra/6e49d05f-f619-4666-a518-328e060f2f8c.json @@ -0,0 +1,438 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/openai_gpt-5.6-terra/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-terra", + "id": "openai/gpt-5.6-terra", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-terra", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 90.327, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.327", + "raw_stderr": "0.916", + "latency": "2.373", + "cost_per_test": "0.004621", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.916, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.178, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.178", + "raw_stderr": "0.186", + "latency": "2.352", + "cost_per_test": "0.006047", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.186, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 89.392, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.392", + "raw_stderr": "0.567", + "latency": "2.159", + "cost_per_test": "0.00709", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.567, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 85.105, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.105", + "raw_stderr": "0.45", + "latency": "2.782", + "cost_per_test": "0.006608", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.45, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.098, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.098", + "raw_stderr": "1.094", + "latency": "2.004", + "cost_per_test": "0.005242", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.094, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.528, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.528", + "raw_stderr": "1.635", + "latency": "5.02", + "cost_per_test": "0.010042", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.635, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-terra/81aa7fa4-ebe8-4029-a83b-5057536cdf1b.json b/data/vals_ai/openai/gpt-5.6-terra/81aa7fa4-ebe8-4029-a83b-5057536cdf1b.json new file mode 100644 index 0000000000000000000000000000000000000000..2d5ee3025ad17a4221d20403f322f832835c8506 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-terra/81aa7fa4-ebe8-4029-a83b-5057536cdf1b.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/openai_gpt-5.6-terra/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-terra", + "id": "openai/gpt-5.6-terra", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-terra", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 66.167, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66.167", + "raw_stderr": "14.997", + "latency": "678.342", + "cost_per_test": "1.023201", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 14.997, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 64.333, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "64.333", + "raw_stderr": "12.199", + "latency": "1983.157", + "cost_per_test": "1.412998", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 12.199, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "ioi:overall:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 65.25, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "65.25", + "raw_stderr": "0.917", + "latency": "1330.749", + "cost_per_test": "1.2181", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.917, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-terra/8ccda8c4-4d7a-466c-9fe0-368d5ec31bc8.json b/data/vals_ai/openai/gpt-5.6-terra/8ccda8c4-4d7a-466c-9fe0-368d5ec31bc8.json new file mode 100644 index 0000000000000000000000000000000000000000..4c7c6dc0d19340ef32c7b8b224dc87acccbd5116 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-terra/8ccda8c4-4d7a-466c-9fe0-368d5ec31bc8.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/openai_gpt-5.6-terra/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-terra", + "id": "openai/gpt-5.6-terra", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-terra", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.026, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.026", + "raw_stderr": "0.951", + "latency": "9.097", + "cost_per_test": "0.013874", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.951, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.875, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.875", + "raw_stderr": "1.025", + "latency": "6.029", + "cost_per_test": "0.017204", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.025, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.194, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.194", + "raw_stderr": "0.884", + "latency": "7.923", + "cost_per_test": "0.02491", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.884, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.976, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.976", + "raw_stderr": "1.415", + "latency": "6.133", + "cost_per_test": "0.020138", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.415, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.981, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.981", + "raw_stderr": "1.078", + "latency": "4.591", + "cost_per_test": "0.013355", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.078, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.869, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.869", + "raw_stderr": "1.21", + "latency": "26.937", + "cost_per_test": "0.056925", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.21, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.763, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.763", + "raw_stderr": "1.321", + "latency": "5.144", + "cost_per_test": "0.015212", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.321, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.528, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.528", + "raw_stderr": "2.067", + "latency": "10.208", + "cost_per_test": "0.015083", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.067, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.387, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.387", + "raw_stderr": "1.317", + "latency": "9.223", + "cost_per_test": "0.028453", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.317, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.152, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.152", + "raw_stderr": "0.638", + "latency": "5.296", + "cost_per_test": "0.013023", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.638, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.091, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.091", + "raw_stderr": "1.203", + "latency": "5.556", + "cost_per_test": "0.015167", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.203, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.659, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.659", + "raw_stderr": "0.333", + "latency": "8.173", + "cost_per_test": "0.020098", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.968, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.968", + "raw_stderr": "1.642", + "latency": "6.126", + "cost_per_test": "0.016369", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.642, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.07, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.07", + "raw_stderr": "0.791", + "latency": "7.846", + "cost_per_test": "0.020128", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.791, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.341, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.341", + "raw_stderr": "1.216", + "latency": "4.315", + "cost_per_test": "0.011528", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.216, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-terra/8f22132a-3817-432e-ba0e-ea8f6b4ed05b.json b/data/vals_ai/openai/gpt-5.6-terra/8f22132a-3817-432e-ba0e-ea8f6b4ed05b.json new file mode 100644 index 0000000000000000000000000000000000000000..38d1df25d3a0582792e9d60b9a4e8fd46c1b45a8 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-terra/8f22132a-3817-432e-ba0e-ea8f6b4ed05b.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/openai_gpt-5.6-terra/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-terra", + "id": "openai/gpt-5.6-terra", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-terra", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 63.17, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.17", + "raw_stderr": "1.647", + "latency": "4.35", + "cost_per_test": "0.017807", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.647, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 66.317, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.317", + "raw_stderr": "1.614", + "latency": "9.507", + "cost_per_test": "1.039485", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.614, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 65.307, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.307", + "raw_stderr": "0.938", + "latency": "7.717", + "cost_per_test": "0.467454", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.938, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 66.434, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.434", + "raw_stderr": "1.612", + "latency": "9.294", + "cost_per_test": "0.345071", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.612, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-terra/abb96536-330e-4f32-89b0-f5e3ffd3e46c.json b/data/vals_ai/openai/gpt-5.6-terra/abb96536-330e-4f32-89b0-f5e3ffd3e46c.json new file mode 100644 index 0000000000000000000000000000000000000000..874cfd8a6f2cda1bd2c55da6edaa28747b8a5bf6 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-terra/abb96536-330e-4f32-89b0-f5e3ffd3e46c.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/openai_gpt-5.6-terra/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-terra", + "id": "openai/gpt-5.6-terra", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-terra", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 83.333, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "83.333", + "raw_stderr": "16.667", + "latency": "5805.032", + "cost_per_test": "15.849543", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 16.667, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 63.333, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "63.333", + "raw_stderr": "3.849", + "latency": "774.004", + "cost_per_test": "2.113272", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.849, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 78.182, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "78.182", + "raw_stderr": "2.777", + "latency": "422.184", + "cost_per_test": "1.152694", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.777, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 73.408, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "73.408", + "raw_stderr": "2.085", + "latency": "260.9", + "cost_per_test": "0.712339", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.085, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-terra/af53f889-e17b-45f8-a01e-16bf73faf0f4.json b/data/vals_ai/openai/gpt-5.6-terra/af53f889-e17b-45f8-a01e-16bf73faf0f4.json new file mode 100644 index 0000000000000000000000000000000000000000..b04f607782cf47980d2707d0b72fe45c2ee66dc2 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-terra/af53f889-e17b-45f8-a01e-16bf73faf0f4.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/public-benefits-bench/openai_gpt-5.6-terra/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Public Benefits Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "social mobility", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-terra", + "id": "openai/gpt-5.6-terra", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-terra", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "public-benefits-bench:overall:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.public-benefits-bench.overall", + "source_data": { + "dataset_name": "Public Benefits Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Public Benefits Bench v1.1 (Overall).", + "metric_id": "vals_ai.public-benefits-bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "score_details": { + "score": 62.382, + "details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "social mobility", + "raw_score": "62.382", + "raw_stderr": "1.26", + "latency": "1314.247", + "cost_per_test": "3.061002", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.26, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-terra/b58e9736-862a-4538-85c2-574f6a7875a9.json b/data/vals_ai/openai/gpt-5.6-terra/b58e9736-862a-4538-85c2-574f6a7875a9.json new file mode 100644 index 0000000000000000000000000000000000000000..02cd821fec8cd8a9c85b0f54e9c2f91ff5397a21 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-terra/b58e9736-862a-4538-85c2-574f6a7875a9.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/openai_gpt-5.6-terra/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-terra", + "id": "openai/gpt-5.6-terra", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-terra", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 97.205, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "97.205", + "raw_stderr": "0.919", + "latency": "4.104", + "cost_per_test": "0.006125", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.919, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "lcb:hard:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 72.857, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "72.857", + "raw_stderr": "2.377", + "latency": "113.134", + "cost_per_test": "0.198055", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.377, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "lcb:medium:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 87.728, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "87.728", + "raw_stderr": "1.677", + "latency": "15.755", + "cost_per_test": "0.02909", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.677, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "lcb:overall:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 85.93, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "85.93", + "raw_stderr": "1.017", + "latency": "44.331", + "cost_per_test": "0.077757", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.017, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-terra/b89c4db1-a4d2-44ed-86d0-9e2d38321803.json b/data/vals_ai/openai/gpt-5.6-terra/b89c4db1-a4d2-44ed-86d0-9e2d38321803.json new file mode 100644 index 0000000000000000000000000000000000000000..6d36fe006114cfa4581a787e77cc052a7cb7f513 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-terra/b89c4db1-a4d2-44ed-86d0-9e2d38321803.json @@ -0,0 +1,676 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/emb/openai_gpt-5.6-terra/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - EMB", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-terra", + "id": "openai/gpt-5.6-terra", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-terra", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "emb:comps:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.emb.comps", + "source_data": { + "dataset_name": "EMB - Comps", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "comps", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Comps).", + "metric_id": "vals_ai.emb.comps.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 72.539, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "comps", + "task_name": "Comps", + "dataset_type": "private", + "industry": "finance", + "raw_score": "72.539", + "raw_stderr": "3.884", + "latency": "652.479", + "cost_per_test": "7.52738", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.884, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:dataroom:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.emb.dataroom", + "source_data": { + "dataset_name": "EMB - Dataroom", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dataroom", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Dataroom).", + "metric_id": "vals_ai.emb.dataroom.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 81.591, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dataroom", + "task_name": "Dataroom", + "dataset_type": "private", + "industry": "finance", + "raw_score": "81.591", + "raw_stderr": "5.802", + "latency": "376.538", + "cost_per_test": "2.796274", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.802, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:dcf:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.emb.dcf", + "source_data": { + "dataset_name": "EMB - DCF", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dcf", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (DCF).", + "metric_id": "vals_ai.emb.dcf.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 63.137, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dcf", + "task_name": "DCF", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.137", + "raw_stderr": "7.603", + "latency": "440.003", + "cost_per_test": "3.055422", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.603, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:lbo:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.emb.lbo", + "source_data": { + "dataset_name": "EMB - LBO", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "lbo", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (LBO).", + "metric_id": "vals_ai.emb.lbo.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 57.396, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "lbo", + "task_name": "LBO", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.396", + "raw_stderr": "4.191", + "latency": "539.225", + "cost_per_test": "6.266524", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.191, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:ma:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.emb.ma", + "source_data": { + "dataset_name": "EMB - M&A", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "ma", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (M&A).", + "metric_id": "vals_ai.emb.ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 76.375, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "ma", + "task_name": "M&A", + "dataset_type": "private", + "industry": "finance", + "raw_score": "76.375", + "raw_stderr": "6.973", + "latency": "420.088", + "cost_per_test": "2.827391", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.973, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:misc:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.emb.misc", + "source_data": { + "dataset_name": "EMB - Misc", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "misc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Misc).", + "metric_id": "vals_ai.emb.misc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 61.827, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "misc", + "task_name": "Misc", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.827", + "raw_stderr": "6.659", + "latency": "436.308", + "cost_per_test": "2.943363", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.659, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:operating:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.emb.operating", + "source_data": { + "dataset_name": "EMB - Operating", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "operating", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Operating).", + "metric_id": "vals_ai.emb.operating.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 60.433, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "operating", + "task_name": "Operating", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.433", + "raw_stderr": "7.308", + "latency": "410.635", + "cost_per_test": "2.58596", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.308, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:overall:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.emb.overall", + "source_data": { + "dataset_name": "EMB - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Overall).", + "metric_id": "vals_ai.emb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 67.614, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.614", + "raw_stderr": "3.035", + "latency": "472.411", + "cost_per_test": "4.106916", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.035, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:scratch:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.emb.scratch", + "source_data": { + "dataset_name": "EMB - Scratch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "scratch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Scratch).", + "metric_id": "vals_ai.emb.scratch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 64.708, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "scratch", + "task_name": "Scratch", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.708", + "raw_stderr": "2.901", + "latency": "532.233", + "cost_per_test": "4.410967", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.901, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "emb:template:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.emb.template", + "source_data": { + "dataset_name": "EMB - Template", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "template", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Template).", + "metric_id": "vals_ai.emb.template.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 70.257, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "template", + "task_name": "Template", + "dataset_type": "private", + "industry": "finance", + "raw_score": "70.257", + "raw_stderr": "3.816", + "latency": "412.589", + "cost_per_test": "3.802864", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.816, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-terra/c45f7dfb-ae45-44d0-9a1e-a0c81935c859.json b/data/vals_ai/openai/gpt-5.6-terra/c45f7dfb-ae45-44d0-9a1e-a0c81935c859.json new file mode 100644 index 0000000000000000000000000000000000000000..4dcea728ea25f0e7f570538b1d7ab5adf8ec6e3a --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-terra/c45f7dfb-ae45-44d0-9a1e-a0c81935c859.json @@ -0,0 +1,103 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/openai_gpt-5.6-terra/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-terra", + "id": "openai/gpt-5.6-terra", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-terra", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 86.474, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.474", + "raw_stderr": "0.822", + "latency": "13.033", + "cost_per_test": "0.040937", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.822, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-terra/d82b0ca3-27b5-44cd-a594-aa2bd4b558ac.json b/data/vals_ai/openai/gpt-5.6-terra/d82b0ca3-27b5-44cd-a594-aa2bd4b558ac.json new file mode 100644 index 0000000000000000000000000000000000000000..8e3e6dde3e9c40a0f0936706b8a5cd565c4ae9fa --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-terra/d82b0ca3-27b5-44cd-a594-aa2bd4b558ac.json @@ -0,0 +1,740 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/openai_gpt-5.6-terra/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-terra", + "id": "openai/gpt-5.6-terra", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-terra", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 50.442, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "50.442", + "raw_stderr": "2.997", + "latency": "420.636", + "cost_per_test": "0.59218", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.997, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 40.448, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "40.448", + "raw_stderr": "0.458", + "latency": "425.115", + "cost_per_test": "0.654944", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.458, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 45.218, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "45.218", + "raw_stderr": "0.909", + "latency": "592.363", + "cost_per_test": "0.85302", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.909, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 61.33, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.33", + "raw_stderr": "0.856", + "latency": "442.847", + "cost_per_test": "0.543184", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.856, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 73.664, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "73.664", + "raw_stderr": "1.613", + "latency": "197.036", + "cost_per_test": "0.457029", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.613, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 21.398, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "21.398", + "raw_stderr": "1.368", + "latency": "367.572", + "cost_per_test": "0.788792", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.368, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 61.601, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.601", + "raw_stderr": "2.945", + "latency": "520.158", + "cost_per_test": "0.482387", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.945, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 71.921, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.921", + "raw_stderr": "2.009", + "latency": "363.769", + "cost_per_test": "0.484043", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.009, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 60.119, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.119", + "raw_stderr": "0.921", + "latency": "297.739", + "cost_per_test": "0.853652", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.921, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:overall:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 52.42, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.42", + "raw_stderr": "0.711", + "latency": "425.115", + "cost_per_test": "0.654944", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.711, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 26.085, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "26.085", + "raw_stderr": "2.875", + "latency": "706.415", + "cost_per_test": "0.904686", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.875, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-terra/e23153c1-045f-4a37-a113-9810361952c5.json b/data/vals_ai/openai/gpt-5.6-terra/e23153c1-045f-4a37-a113-9810361952c5.json new file mode 100644 index 0000000000000000000000000000000000000000..733ab89df93a76a421403f477a2fd1b703cdd4d0 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-terra/e23153c1-045f-4a37-a113-9810361952c5.json @@ -0,0 +1,371 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/openai_gpt-5.6-terra/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-terra", + "id": "openai/gpt-5.6-terra", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-terra", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 64.286, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "64.286", + "raw_stderr": "7.483", + "latency": "177.434", + "cost_per_test": "1.07697", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.483, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 75.479, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "75.479", + "raw_stderr": "2.668", + "latency": "125.913", + "cost_per_test": "0.75089", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.668, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 77.835, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "77.835", + "raw_stderr": "2.99", + "latency": "101.053", + "cost_per_test": "0.58", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.99, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "338.9", + "cost_per_test": "1.446064", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "swebench:overall:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 75.2, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "75.2", + "raw_stderr": "1.933", + "latency": "121.857", + "cost_per_test": "0.716007", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.933, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-terra/e6813661-cb7a-4ec2-bc78-4d40fafe00a7.json b/data/vals_ai/openai/gpt-5.6-terra/e6813661-cb7a-4ec2-bc78-4d40fafe00a7.json new file mode 100644 index 0000000000000000000000000000000000000000..801b90654c956e06db48c1963262f02fb21516e6 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-terra/e6813661-cb7a-4ec2-bc78-4d40fafe00a7.json @@ -0,0 +1,356 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/openai_gpt-5.6-terra/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-terra", + "id": "openai/gpt-5.6-terra", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-terra", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 38.095, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "38.095", + "raw_stderr": "6.103", + "latency": "44.223", + "cost_per_test": "0.106104", + "max_output_tokens": "30000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.103, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "sage:automata:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 38.889, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "38.889", + "raw_stderr": "6.166", + "latency": "50.623", + "cost_per_test": "0.092952", + "max_output_tokens": "30000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.166, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "sage:calculus:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 59.259, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "59.259", + "raw_stderr": "6.315", + "latency": "31.529", + "cost_per_test": "0.138603", + "max_output_tokens": "30000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.315, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 64.025, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "64.025", + "raw_stderr": "5.379", + "latency": "25.326", + "cost_per_test": "0.084394", + "max_output_tokens": "30000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.379, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "sage:overall:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 47.003, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "47.003", + "raw_stderr": "3.403", + "latency": "40.057", + "cost_per_test": "0.094483", + "max_output_tokens": "30000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.403, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-terra/f44c4a59-a2a5-4d33-be46-3194c615259e.json b/data/vals_ai/openai/gpt-5.6-terra/f44c4a59-a2a5-4d33-be46-3194c615259e.json new file mode 100644 index 0000000000000000000000000000000000000000..eae580d3462fe3151de590cdf7f508a12b695fae --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-terra/f44c4a59-a2a5-4d33-be46-3194c615259e.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/openai_gpt-5.6-terra/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-terra", + "id": "openai/gpt-5.6-terra", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-terra", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 89.899, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.899", + "raw_stderr": "2.711", + "latency": "43.438", + "cost_per_test": "0.083172", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.711, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 91.919, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.919", + "raw_stderr": "2.68", + "latency": "30.09", + "cost_per_test": "0.086815", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.68, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 90.909, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.909", + "raw_stderr": "1.906", + "latency": "36.764", + "cost_per_test": "0.084994", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.906, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/gpt-5.6-terra/f6219b9b-c9e8-4dda-80a3-d56bf0c48fa0.json b/data/vals_ai/openai/gpt-5.6-terra/f6219b9b-c9e8-4dda-80a3-d56bf0c48fa0.json new file mode 100644 index 0000000000000000000000000000000000000000..57f6a2934690d7625a96e78694cbb72da0be4f94 --- /dev/null +++ b/data/vals_ai/openai/gpt-5.6-terra/f6219b9b-c9e8-4dda-80a3-d56bf0c48fa0.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/openai_gpt-5.6-terra/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "gpt-5.6-terra", + "id": "openai/gpt-5.6-terra", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/gpt-5.6-terra", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:openai/gpt-5.6-terra:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 71.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "71", + "raw_stderr": "4.56", + "latency": "421.678", + "cost_per_test": "2.918608", + "max_output_tokens": "128000", + "reasoning_effort": "xhigh", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.56, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 128000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "xhigh" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o1-2024-12-17/24acaf3c-f89a-43b2-92b4-0c9ffd736108.json b/data/vals_ai/openai/o1-2024-12-17/24acaf3c-f89a-43b2-92b4-0c9ffd736108.json new file mode 100644 index 0000000000000000000000000000000000000000..5c889270287aa01b903c76c49a2cf74b73f64a27 --- /dev/null +++ b/data/vals_ai/openai/o1-2024-12-17/24acaf3c-f89a-43b2-92b4-0c9ffd736108.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/openai_o1-2024-12-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o1-2024-12-17", + "id": "openai/o1-2024-12-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o1-2024-12-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 73.913, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "73.913", + "raw_stderr": "2.447", + "latency": "15.647", + "cost_per_test": "0.065374", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.447, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:hard:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 22.571, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "22.571", + "raw_stderr": "2.235", + "latency": "172.829", + "cost_per_test": "0.622193", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.235, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:medium:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 54.308, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "54.308", + "raw_stderr": "2.545", + "latency": "87.77", + "cost_per_test": "0.296634", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.545, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:overall:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 50.264, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "50.264", + "raw_stderr": "1.393", + "latency": "92.082", + "cost_per_test": "0.328067", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.393, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o1-2024-12-17/4998af91-1444-4dec-ad36-951f27a9ef94.json b/data/vals_ai/openai/o1-2024-12-17/4998af91-1444-4dec-ad36-951f27a9ef94.json new file mode 100644 index 0000000000000000000000000000000000000000..a8c42fdbfcd6537423a71488094ea610cb99fe66 --- /dev/null +++ b/data/vals_ai/openai/o1-2024-12-17/4998af91-1444-4dec-ad36-951f27a9ef94.json @@ -0,0 +1,103 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/openai_o1-2024-12-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o1-2024-12-17", + "id": "openai/o1-2024-12-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o1-2024-12-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 77.412, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.412", + "raw_stderr": "1.005", + "latency": "26.415", + "cost_per_test": "0.14586", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.005, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o1-2024-12-17/4b7cfbb0-902e-43f2-8fb8-cfda9bcc3da1.json b/data/vals_ai/openai/o1-2024-12-17/4b7cfbb0-902e-43f2-8fb8-cfda9bcc3da1.json new file mode 100644 index 0000000000000000000000000000000000000000..e118a7e1c722af943457a226dcdc94a0d450f267 --- /dev/null +++ b/data/vals_ai/openai/o1-2024-12-17/4b7cfbb0-902e-43f2-8fb8-cfda9bcc3da1.json @@ -0,0 +1,432 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/openai_o1-2024-12-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o1-2024-12-17", + "id": "openai/o1-2024-12-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o1-2024-12-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 86.892, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.892", + "raw_stderr": "0.71", + "latency": "7.02", + "max_output_tokens": "2048", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.71, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 2048, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 75.313, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "75.313", + "raw_stderr": "0.213", + "latency": "8.514", + "max_output_tokens": "2048", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.213, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 2048, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.004, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.004", + "raw_stderr": "0.656", + "latency": "9.267", + "max_output_tokens": "2048", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.656, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 2048, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.393, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.393", + "raw_stderr": "0.578", + "latency": "8.338", + "max_output_tokens": "2048", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.578, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 2048, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.891, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.891", + "raw_stderr": "1.248", + "latency": "8.145", + "max_output_tokens": "2048", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.248, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 2048, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.864, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.864", + "raw_stderr": "2.414", + "latency": "8.742", + "max_output_tokens": "2048", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.414, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 2048, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o1-2024-12-17/7216ba1f-6cd1-4ab0-8b60-4f12c098dc12.json b/data/vals_ai/openai/o1-2024-12-17/7216ba1f-6cd1-4ab0-8b60-4f12c098dc12.json new file mode 100644 index 0000000000000000000000000000000000000000..6681afd089bacad7188bbc6d5500c0eaf7d39986 --- /dev/null +++ b/data/vals_ai/openai/o1-2024-12-17/7216ba1f-6cd1-4ab0-8b60-4f12c098dc12.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/openai_o1-2024-12-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o1-2024-12-17", + "id": "openai/o1-2024-12-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o1-2024-12-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.911, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.911", + "raw_stderr": "1.018", + "latency": "20.622", + "cost_per_test": "0.093755", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.018, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.1, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.1", + "raw_stderr": "1.109", + "latency": "23.018", + "cost_per_test": "0.100986", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.109, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.661, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.661", + "raw_stderr": "1.013", + "latency": "34.973", + "cost_per_test": "0.152426", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.013, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.854, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.854", + "raw_stderr": "1.721", + "latency": "28.946", + "cost_per_test": "0.119714", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.721, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.204, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.204", + "raw_stderr": "1.15", + "latency": "20.018", + "cost_per_test": "0.088583", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.15, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.2, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.2", + "raw_stderr": "1.406", + "latency": "62.165", + "cost_per_test": "0.232772", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.406, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.951, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.951", + "raw_stderr": "1.4", + "latency": "23.197", + "cost_per_test": "0.0985", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.4, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.066, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.066", + "raw_stderr": "2.216", + "latency": "23.108", + "cost_per_test": "0.103124", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.216, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.571, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.571", + "raw_stderr": "1.359", + "latency": "29.349", + "cost_per_test": "0.137459", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.359, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.932, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.932", + "raw_stderr": "0.744", + "latency": "24.494", + "cost_per_test": "0.12593", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.744, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.684, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.684", + "raw_stderr": "1.245", + "latency": "16.101", + "cost_per_test": "0.081239", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.245, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.488, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.488", + "raw_stderr": "0.364", + "latency": "26.867", + "cost_per_test": "0.116223", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.364, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.962, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.962", + "raw_stderr": "1.758", + "latency": "19.121", + "cost_per_test": "0.086531", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.758, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.529, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.529", + "raw_stderr": "0.917", + "latency": "33.158", + "cost_per_test": "0.131873", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.917, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.211, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.211", + "raw_stderr": "1.291", + "latency": "17.866", + "cost_per_test": "0.074227", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.291, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o1-2024-12-17/d16fd0ac-c2d6-47d3-9c77-5162dbd0511c.json b/data/vals_ai/openai/o1-2024-12-17/d16fd0ac-c2d6-47d3-9c77-5162dbd0511c.json new file mode 100644 index 0000000000000000000000000000000000000000..8d02b948273fa3df9c5a802aa612a87ef816604a --- /dev/null +++ b/data/vals_ai/openai/o1-2024-12-17/d16fd0ac-c2d6-47d3-9c77-5162dbd0511c.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/openai_o1-2024-12-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o1-2024-12-17", + "id": "openai/o1-2024-12-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o1-2024-12-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 64.432, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.432", + "raw_stderr": "1.369", + "latency": "19.477", + "cost_per_test": "0.12173", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.369, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 74.284, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "74.284", + "raw_stderr": "0.861", + "latency": "19.477", + "cost_per_test": "0.12173", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.861, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 84.137, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "84.137", + "raw_stderr": "1.045", + "latency": "19.477", + "cost_per_test": "0.12173", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.045, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o1-2024-12-17/d1aebcbc-39d7-4f60-9c89-4200d28b074c.json b/data/vals_ai/openai/o1-2024-12-17/d1aebcbc-39d7-4f60-9c89-4200d28b074c.json new file mode 100644 index 0000000000000000000000000000000000000000..0c139e23e140d32dc88c5f37ab753d7989c94e27 --- /dev/null +++ b/data/vals_ai/openai/o1-2024-12-17/d1aebcbc-39d7-4f60-9c89-4200d28b074c.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/openai_o1-2024-12-17/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o1-2024-12-17", + "id": "openai/o1-2024-12-17", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o1-2024-12-17", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 72.727, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.727", + "raw_stderr": "3.165", + "latency": "35.372", + "cost_per_test": "0.227864", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.165, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 73.737, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.737", + "raw_stderr": "3.146", + "latency": "44.979", + "cost_per_test": "0.277401", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.146, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:openai/o1-2024-12-17:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 73.232, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.232", + "raw_stderr": "2.231", + "latency": "40.176", + "cost_per_test": "0.252632", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.231, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o3-2025-04-16/10aabcda-08ad-4c37-9250-5fb9bb1873b3.json b/data/vals_ai/openai/o3-2025-04-16/10aabcda-08ad-4c37-9250-5fb9bb1873b3.json new file mode 100644 index 0000000000000000000000000000000000000000..bc481f32380782850e52e3dd005d2e6762c5b3c1 --- /dev/null +++ b/data/vals_ai/openai/o3-2025-04-16/10aabcda-08ad-4c37-9250-5fb9bb1873b3.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/openai_o3-2025-04-16/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o3-2025-04-16", + "id": "openai/o3-2025-04-16", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o3-2025-04-16", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 73.847, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "73.847", + "raw_stderr": "1.239", + "latency": "25.416", + "cost_per_test": "0.011546", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.239, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 65.7, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.7", + "raw_stderr": "0.933", + "latency": "21.216", + "cost_per_test": "0.009972", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.933, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 57.552, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.552", + "raw_stderr": "1.394", + "latency": "17.015", + "cost_per_test": "0.008397", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.394, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o3-2025-04-16/393d091a-5a07-4a91-a9ab-b29b15f2186d.json b/data/vals_ai/openai/o3-2025-04-16/393d091a-5a07-4a91-a9ab-b29b15f2186d.json new file mode 100644 index 0000000000000000000000000000000000000000..7f124d35889c65e4c67154ef13f72d062fd00eb6 --- /dev/null +++ b/data/vals_ai/openai/o3-2025-04-16/393d091a-5a07-4a91-a9ab-b29b15f2186d.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/openai_o3-2025-04-16/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o3-2025-04-16", + "id": "openai/o3-2025-04-16", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o3-2025-04-16", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.447, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.447", + "raw_stderr": "0.689", + "latency": "10.019", + "cost_per_test": "0.005918", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.689, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:hard:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 64.0, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "64", + "raw_stderr": "2.566", + "latency": "100.796", + "cost_per_test": "0.068741", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.566, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:medium:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 89.295, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "89.295", + "raw_stderr": "1.58", + "latency": "81.028", + "cost_per_test": "0.024229", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.58, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:overall:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 83.914, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "83.914", + "raw_stderr": "1.03", + "latency": "63.948", + "cost_per_test": "0.032963", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.03, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o3-2025-04-16/76413d17-d281-4528-8d0f-dfe497cbfe25.json b/data/vals_ai/openai/o3-2025-04-16/76413d17-d281-4528-8d0f-dfe497cbfe25.json new file mode 100644 index 0000000000000000000000000000000000000000..e7feb0a452daa1b626c41b190a4785a90e217813 --- /dev/null +++ b/data/vals_ai/openai/o3-2025-04-16/76413d17-d281-4528-8d0f-dfe497cbfe25.json @@ -0,0 +1,432 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/openai_o3-2025-04-16/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o3-2025-04-16", + "id": "openai/o3-2025-04-16", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o3-2025-04-16", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 88.283, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.283", + "raw_stderr": "0.797", + "latency": "5.751", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.797, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.236, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.236", + "raw_stderr": "0.202", + "latency": "5.062", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.202, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.587, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.587", + "raw_stderr": "0.574", + "latency": "4.6", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.574, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.761, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.761", + "raw_stderr": "0.424", + "latency": "5.979", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.424, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.512, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.512", + "raw_stderr": "0.866", + "latency": "3.629", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.866, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.187, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.187", + "raw_stderr": "1.654", + "latency": "10.852", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.654, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o3-2025-04-16/83f016b7-2d5f-4c7a-9037-086ed30baab8.json b/data/vals_ai/openai/o3-2025-04-16/83f016b7-2d5f-4c7a-9037-086ed30baab8.json new file mode 100644 index 0000000000000000000000000000000000000000..e3e4f03cecf513ca7e705bc114574cdb3bc66df7 --- /dev/null +++ b/data/vals_ai/openai/o3-2025-04-16/83f016b7-2d5f-4c7a-9037-086ed30baab8.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/openai_o3-2025-04-16/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o3-2025-04-16", + "id": "openai/o3-2025-04-16", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o3-2025-04-16", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 64.023, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.023", + "raw_stderr": "1.372", + "latency": "24.024", + "cost_per_test": "0.012997", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.372, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 74.571, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "74.571", + "raw_stderr": "0.854", + "latency": "24.024", + "cost_per_test": "0.012997", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.854, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 85.119, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "85.119", + "raw_stderr": "1.018", + "latency": "24.024", + "cost_per_test": "0.012997", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.018, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o3-2025-04-16/8fc3638c-e6b1-48bf-8b5d-0dd5a9879d99.json b/data/vals_ai/openai/o3-2025-04-16/8fc3638c-e6b1-48bf-8b5d-0dd5a9879d99.json new file mode 100644 index 0000000000000000000000000000000000000000..26bee30ccccd696e1d138c617e8fa67cdc6bae51 --- /dev/null +++ b/data/vals_ai/openai/o3-2025-04-16/8fc3638c-e6b1-48bf-8b5d-0dd5a9879d99.json @@ -0,0 +1,103 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/openai_o3-2025-04-16/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o3-2025-04-16", + "id": "openai/o3-2025-04-16", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o3-2025-04-16", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 80.416, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.416", + "raw_stderr": "0.954", + "latency": "39.275", + "cost_per_test": "0.015444", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.954, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o3-2025-04-16/a9de2a10-6cbc-4dad-a3dd-011f6430e437.json b/data/vals_ai/openai/o3-2025-04-16/a9de2a10-6cbc-4dad-a3dd-011f6430e437.json new file mode 100644 index 0000000000000000000000000000000000000000..de857c3d213b19896d2371361b2fd26a498b37a3 --- /dev/null +++ b/data/vals_ai/openai/o3-2025-04-16/a9de2a10-6cbc-4dad-a3dd-011f6430e437.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/openai_o3-2025-04-16/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o3-2025-04-16", + "id": "openai/o3-2025-04-16", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o3-2025-04-16", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 63.52, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.52", + "raw_stderr": "1.643", + "latency": "10.487", + "cost_per_test": "0.008451", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.643, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 56.061, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.061", + "raw_stderr": "1.694", + "latency": "28.918", + "cost_per_test": "0.260856", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.694, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 59.713, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.713", + "raw_stderr": "0.965", + "latency": "19.153", + "cost_per_test": "0.133487", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.965, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 59.557, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.557", + "raw_stderr": "1.675", + "latency": "18.054", + "cost_per_test": "0.131154", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.675, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o3-2025-04-16/aa1f2102-ec90-4ed8-83fd-4d9d20096ccd.json b/data/vals_ai/openai/o3-2025-04-16/aa1f2102-ec90-4ed8-83fd-4d9d20096ccd.json new file mode 100644 index 0000000000000000000000000000000000000000..da9d0a6cafbdcfc02b649a14de5bda249842ca6e --- /dev/null +++ b/data/vals_ai/openai/o3-2025-04-16/aa1f2102-ec90-4ed8-83fd-4d9d20096ccd.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/openai_o3-2025-04-16/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o3-2025-04-16", + "id": "openai/o3-2025-04-16", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o3-2025-04-16", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 76.654, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "76.654", + "raw_stderr": "1.871", + "latency": "46.392", + "cost_per_test": "0.040334", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.871, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o3-2025-04-16/b171fcda-593c-4d43-84a6-b7475d1db70a.json b/data/vals_ai/openai/o3-2025-04-16/b171fcda-593c-4d43-84a6-b7475d1db70a.json new file mode 100644 index 0000000000000000000000000000000000000000..01bf95669e9aae18acb10222acb1988b804a4163 --- /dev/null +++ b/data/vals_ai/openai/o3-2025-04-16/b171fcda-593c-4d43-84a6-b7475d1db70a.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/openai_o3-2025-04-16/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o3-2025-04-16", + "id": "openai/o3-2025-04-16", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o3-2025-04-16", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.608, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.608", + "raw_stderr": "0.977", + "latency": "10.958", + "cost_per_test": "0.008551", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.977, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.987, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.987", + "raw_stderr": "1.069", + "latency": "14.834", + "cost_per_test": "0.011072", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.069, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.134, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.134", + "raw_stderr": "0.925", + "latency": "22.617", + "cost_per_test": "0.016523", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.925, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.073, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.073", + "raw_stderr": "1.657", + "latency": "15.171", + "cost_per_test": "0.011691", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.657, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.981, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.981", + "raw_stderr": "1.078", + "latency": "9.432", + "cost_per_test": "0.007898", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.078, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.638, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.638", + "raw_stderr": "1.317", + "latency": "47.699", + "cost_per_test": "0.029288", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.317, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.685, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.685", + "raw_stderr": "1.38", + "latency": "9.436", + "cost_per_test": "0.007691", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.38, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.953, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.953", + "raw_stderr": "2.124", + "latency": "9.482", + "cost_per_test": "0.009669", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.124, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.297, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.297", + "raw_stderr": "1.333", + "latency": "20.864", + "cost_per_test": "0.012843", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.004, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.004", + "raw_stderr": "0.646", + "latency": "20.683", + "cost_per_test": "0.01422", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.646, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.416, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.416", + "raw_stderr": "1.193", + "latency": "9.539", + "cost_per_test": "0.007168", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.193, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.595, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.595", + "raw_stderr": "0.345", + "latency": "16.735", + "cost_per_test": "0.011827", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.345, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.17, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.17", + "raw_stderr": "1.591", + "latency": "14.017", + "cost_per_test": "0.007963", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.591, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.915, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.915", + "raw_stderr": "0.835", + "latency": "17.865", + "cost_per_test": "0.013916", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.835, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.466, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.466", + "raw_stderr": "1.211", + "latency": "11.696", + "cost_per_test": "0.007081", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.211, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o3-2025-04-16/bce4753e-fd6e-439c-a8a2-f97e8f9954a3.json b/data/vals_ai/openai/o3-2025-04-16/bce4753e-fd6e-439c-a8a2-f97e8f9954a3.json new file mode 100644 index 0000000000000000000000000000000000000000..00e9eba169891766693a4db1831c699a7115b9d0 --- /dev/null +++ b/data/vals_ai/openai/o3-2025-04-16/bce4753e-fd6e-439c-a8a2-f97e8f9954a3.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/openai_o3-2025-04-16/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o3-2025-04-16", + "id": "openai/o3-2025-04-16", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o3-2025-04-16", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 82.828, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.828", + "raw_stderr": "2.68", + "latency": "74.323", + "cost_per_test": "0.037891", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.68, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 85.354, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.354", + "raw_stderr": "2.583", + "latency": "57.231", + "cost_per_test": "0.033212", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.583, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 84.091, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.091", + "raw_stderr": "1.861", + "latency": "65.777", + "cost_per_test": "0.035552", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.861, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o3-2025-04-16/dc63d530-4aca-4894-8bb5-794a904652ce.json b/data/vals_ai/openai/o3-2025-04-16/dc63d530-4aca-4894-8bb5-794a904652ce.json new file mode 100644 index 0000000000000000000000000000000000000000..ba0d266e8d8ed93858d75a136a119c679b3fbe23 --- /dev/null +++ b/data/vals_ai/openai/o3-2025-04-16/dc63d530-4aca-4894-8bb5-794a904652ce.json @@ -0,0 +1,356 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/openai_o3-2025-04-16/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o3-2025-04-16", + "id": "openai/o3-2025-04-16", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o3-2025-04-16", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 26.984, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "26.984", + "raw_stderr": "5.57", + "latency": "56.811", + "cost_per_test": "0.02951", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.57, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:automata:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 36.905, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "36.905", + "raw_stderr": "6.162", + "latency": "38.076", + "cost_per_test": "0.020412", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.162, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:calculus:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 55.556, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "55.556", + "raw_stderr": "6.306", + "latency": "145.43", + "cost_per_test": "0.06546", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.306, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 61.423, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "61.423", + "raw_stderr": "5.468", + "latency": "49.184", + "cost_per_test": "0.033352", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.468, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:overall:openai/o3-2025-04-16:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 41.771, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "41.771", + "raw_stderr": "3.315", + "latency": "48.024", + "cost_per_test": "0.027758", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.315, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o3-mini-2025-01-31/345f45ae-d09c-4bd6-aee8-ab3b8dfb8556.json b/data/vals_ai/openai/o3-mini-2025-01-31/345f45ae-d09c-4bd6-aee8-ab3b8dfb8556.json new file mode 100644 index 0000000000000000000000000000000000000000..477b5782418d467382782aab22d5ddd11df1e17f --- /dev/null +++ b/data/vals_ai/openai/o3-mini-2025-01-31/345f45ae-d09c-4bd6-aee8-ab3b8dfb8556.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/openai_o3-mini-2025-01-31/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o3-mini-2025-01-31", + "id": "openai/o3-mini-2025-01-31", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o3-mini-2025-01-31", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 56.41, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.41", + "raw_stderr": "1.693", + "latency": "12.114", + "cost_per_test": "0.004987", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.693, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 33.566, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "33.566", + "raw_stderr": "1.612", + "latency": "57.794", + "cost_per_test": "0.155768", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.612, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 45.299, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "45.299", + "raw_stderr": "0.964", + "latency": "31.047", + "cost_per_test": "0.078121", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.964, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 45.921, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "45.921", + "raw_stderr": "1.701", + "latency": "23.233", + "cost_per_test": "0.073609", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.701, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o3-mini-2025-01-31/59724d67-1198-4298-9579-0f6da4bc72ea.json b/data/vals_ai/openai/o3-mini-2025-01-31/59724d67-1198-4298-9579-0f6da4bc72ea.json new file mode 100644 index 0000000000000000000000000000000000000000..ac9888e1b9c71dd3cbe87b3ca26af70bd99eef04 --- /dev/null +++ b/data/vals_ai/openai/o3-mini-2025-01-31/59724d67-1198-4298-9579-0f6da4bc72ea.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/openai_o3-mini-2025-01-31/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o3-mini-2025-01-31", + "id": "openai/o3-mini-2025-01-31", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o3-mini-2025-01-31", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.758, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.758", + "raw_stderr": "0.617", + "latency": "17.082", + "cost_per_test": "0.008187", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.617, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:hard:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 39.714, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "39.714", + "raw_stderr": "2.615", + "latency": "90.45", + "cost_per_test": "0.057867", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.615, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:medium:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 75.979, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "75.979", + "raw_stderr": "2.183", + "latency": "53.868", + "cost_per_test": "0.03331", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.183, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:overall:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 71.484, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "71.484", + "raw_stderr": "1.154", + "latency": "53.8", + "cost_per_test": "0.033121", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.154, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o3-mini-2025-01-31/7cd54ff2-6df4-41c0-86bd-051c3714a574.json b/data/vals_ai/openai/o3-mini-2025-01-31/7cd54ff2-6df4-41c0-86bd-051c3714a574.json new file mode 100644 index 0000000000000000000000000000000000000000..5e14dda2ec5b126e44a6ae358ce2b8fbe4547aa7 --- /dev/null +++ b/data/vals_ai/openai/o3-mini-2025-01-31/7cd54ff2-6df4-41c0-86bd-051c3714a574.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/openai_o3-mini-2025-01-31/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o3-mini-2025-01-31", + "id": "openai/o3-mini-2025-01-31", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o3-mini-2025-01-31", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 59.362, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.362", + "raw_stderr": "1.404", + "latency": "74.312", + "cost_per_test": "0.018142", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.404, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 69.42, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "69.42", + "raw_stderr": "0.909", + "latency": "74.312", + "cost_per_test": "0.018142", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.909, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 79.477, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "79.477", + "raw_stderr": "1.155", + "latency": "74.312", + "cost_per_test": "0.018142", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.155, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o3-mini-2025-01-31/b66fce5b-a93d-46c6-9f37-48ff1307468d.json b/data/vals_ai/openai/o3-mini-2025-01-31/b66fce5b-a93d-46c6-9f37-48ff1307468d.json new file mode 100644 index 0000000000000000000000000000000000000000..0bbf729ddb090b11193db8d711f1b6f6c249dbe5 --- /dev/null +++ b/data/vals_ai/openai/o3-mini-2025-01-31/b66fce5b-a93d-46c6-9f37-48ff1307468d.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/openai_o3-mini-2025-01-31/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o3-mini-2025-01-31", + "id": "openai/o3-mini-2025-01-31", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o3-mini-2025-01-31", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 74.242, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.242", + "raw_stderr": "3.108", + "latency": "91.417", + "cost_per_test": "0.036567", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.108, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 76.768, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.768", + "raw_stderr": "3.046", + "latency": "108.55", + "cost_per_test": "0.035147", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.046, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 75.505, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.505", + "raw_stderr": "2.176", + "latency": "99.984", + "cost_per_test": "0.035857", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.176, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o3-mini-2025-01-31/c81a12d2-ef78-4395-a1e5-c141e41e43d2.json b/data/vals_ai/openai/o3-mini-2025-01-31/c81a12d2-ef78-4395-a1e5-c141e41e43d2.json new file mode 100644 index 0000000000000000000000000000000000000000..e1f845472a25ec589ba9a50be46eac2155103628 --- /dev/null +++ b/data/vals_ai/openai/o3-mini-2025-01-31/c81a12d2-ef78-4395-a1e5-c141e41e43d2.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/openai_o3-mini-2025-01-31/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o3-mini-2025-01-31", + "id": "openai/o3-mini-2025-01-31", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o3-mini-2025-01-31", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.237, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.237", + "raw_stderr": "1.108", + "latency": "15.13", + "cost_per_test": "0.009205", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.108, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.692, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.692", + "raw_stderr": "1.209", + "latency": "24.632", + "cost_per_test": "0.012235", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.209, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.749, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.749", + "raw_stderr": "1.008", + "latency": "24.487", + "cost_per_test": "0.015137", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.008, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.634, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.634", + "raw_stderr": "1.781", + "latency": "18.417", + "cost_per_test": "0.01223", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.781, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.071, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.071", + "raw_stderr": "1.227", + "latency": "15.538", + "cost_per_test": "0.009259", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.227, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 68.627, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "68.627", + "raw_stderr": "1.491", + "latency": "50.797", + "cost_per_test": "0.029435", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.491, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.751, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.751", + "raw_stderr": "1.454", + "latency": "18.661", + "cost_per_test": "0.009501", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.454, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 65.879, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "65.879", + "raw_stderr": "2.429", + "latency": "20.203", + "cost_per_test": "0.012645", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.429, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 50.772, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "50.772", + "raw_stderr": "1.507", + "latency": "31.103", + "cost_per_test": "0.018975", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.507, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.264, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.264", + "raw_stderr": "0.682", + "latency": "20.432", + "cost_per_test": "0.011498", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.682, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.242, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.242", + "raw_stderr": "1.439", + "latency": "17.041", + "cost_per_test": "0.010356", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.439, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.689, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.689", + "raw_stderr": "0.394", + "latency": "22.375", + "cost_per_test": "0.012991", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.394, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 69.138, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.138", + "raw_stderr": "2.068", + "latency": "20.18", + "cost_per_test": "0.009987", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.068, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.76, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.76", + "raw_stderr": "0.909", + "latency": "22.351", + "cost_per_test": "0.013089", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.909, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.827, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.827", + "raw_stderr": "1.394", + "latency": "14.281", + "cost_per_test": "0.008321", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.394, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o3-mini-2025-01-31/e77eee86-eeef-4119-8482-7e3857fccde6.json b/data/vals_ai/openai/o3-mini-2025-01-31/e77eee86-eeef-4119-8482-7e3857fccde6.json new file mode 100644 index 0000000000000000000000000000000000000000..879f786021c64f0a46fddc9a4b79ad0bf81b27d8 --- /dev/null +++ b/data/vals_ai/openai/o3-mini-2025-01-31/e77eee86-eeef-4119-8482-7e3857fccde6.json @@ -0,0 +1,432 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/openai_o3-mini-2025-01-31/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o3-mini-2025-01-31", + "id": "openai/o3-mini-2025-01-31", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o3-mini-2025-01-31", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 84.584, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.584", + "raw_stderr": "0.811", + "latency": "9.492", + "max_output_tokens": "4096", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.811, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 70.04, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "70.04", + "raw_stderr": "0.237", + "latency": "8.804", + "max_output_tokens": "4096", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.237, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 65.64, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "65.64", + "raw_stderr": "0.812", + "latency": "14.422", + "max_output_tokens": "4096", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.812, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 71.539, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "71.539", + "raw_stderr": "0.72", + "latency": "12.011", + "max_output_tokens": "4096", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.72, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.53, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.53", + "raw_stderr": "1.143", + "latency": "11.706", + "max_output_tokens": "4096", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.143, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:openai/o3-mini-2025-01-31:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 59.9, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "59.9", + "raw_stderr": "3.205", + "latency": "15.631", + "max_output_tokens": "4096", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.205, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o4-mini-2025-04-16/0f209c75-7cde-4306-9698-9ffc52b9a7d0.json b/data/vals_ai/openai/o4-mini-2025-04-16/0f209c75-7cde-4306-9698-9ffc52b9a7d0.json new file mode 100644 index 0000000000000000000000000000000000000000..eed3c9676d860abbe9ade341905a6cf4e12a677d --- /dev/null +++ b/data/vals_ai/openai/o4-mini-2025-04-16/0f209c75-7cde-4306-9698-9ffc52b9a7d0.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/openai_o4-mini-2025-04-16/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o4-mini-2025-04-16", + "id": "openai/o4-mini-2025-04-16", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o4-mini-2025-04-16", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.4, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.4", + "raw_stderr": "1.15", + "latency": "8.895", + "cost_per_test": "0.004693", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.15, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.072, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.072", + "raw_stderr": "1.194", + "latency": "10.359", + "cost_per_test": "0.005291", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.194, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.661, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.661", + "raw_stderr": "1.011", + "latency": "10.034", + "cost_per_test": "0.006322", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.011, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.878, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.878", + "raw_stderr": "1.769", + "latency": "11.322", + "cost_per_test": "0.005717", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.769, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.308, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.308", + "raw_stderr": "1.219", + "latency": "6.938", + "cost_per_test": "0.004229", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.219, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 72.136, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.136", + "raw_stderr": "1.442", + "latency": "24.109", + "cost_per_test": "0.010718", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.442, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.406, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.406", + "raw_stderr": "1.485", + "latency": "8.799", + "cost_per_test": "0.004377", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.485, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 70.866, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.866", + "raw_stderr": "2.328", + "latency": "7.315", + "cost_per_test": "0.005467", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.328, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 61.399, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "61.399", + "raw_stderr": "1.467", + "latency": "12.946", + "cost_per_test": "0.006345", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.467, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.302, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.302", + "raw_stderr": "0.725", + "latency": "9.219", + "cost_per_test": "0.0055", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.725, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.68, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.68", + "raw_stderr": "1.347", + "latency": "7.297", + "cost_per_test": "0.004122", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.347, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.561, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.561", + "raw_stderr": "0.386", + "latency": "10.443", + "cost_per_test": "0.005524", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.386, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.547, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.547", + "raw_stderr": "1.975", + "latency": "11.216", + "cost_per_test": "0.004966", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.975, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.375, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.375", + "raw_stderr": "0.922", + "latency": "10.661", + "cost_per_test": "0.005527", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.922, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.83, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.83", + "raw_stderr": "1.365", + "latency": "7.088", + "cost_per_test": "0.004056", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.365, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o4-mini-2025-04-16/152dda38-b914-4046-b40a-923f82185be0.json b/data/vals_ai/openai/o4-mini-2025-04-16/152dda38-b914-4046-b40a-923f82185be0.json new file mode 100644 index 0000000000000000000000000000000000000000..94746ec410e59fc276fcff257e66de725e52c367 --- /dev/null +++ b/data/vals_ai/openai/o4-mini-2025-04-16/152dda38-b914-4046-b40a-923f82185be0.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/openai_o4-mini-2025-04-16/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o4-mini-2025-04-16", + "id": "openai/o4-mini-2025-04-16", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o4-mini-2025-04-16", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 64.432, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.432", + "raw_stderr": "1.369", + "latency": "15.447", + "cost_per_test": "0.007057", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.369, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 74.776, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "74.776", + "raw_stderr": "0.853", + "latency": "15.447", + "cost_per_test": "0.007057", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.853, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 85.119, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "85.119", + "raw_stderr": "1.018", + "latency": "15.447", + "cost_per_test": "0.007057", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.018, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o4-mini-2025-04-16/179df2af-6bec-4e2c-a6c4-fea9b93992ba.json b/data/vals_ai/openai/o4-mini-2025-04-16/179df2af-6bec-4e2c-a6c4-fea9b93992ba.json new file mode 100644 index 0000000000000000000000000000000000000000..7b1232345025b510daa7523fde0c5f9bc3e041b6 --- /dev/null +++ b/data/vals_ai/openai/o4-mini-2025-04-16/179df2af-6bec-4e2c-a6c4-fea9b93992ba.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/openai_o4-mini-2025-04-16/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o4-mini-2025-04-16", + "id": "openai/o4-mini-2025-04-16", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o4-mini-2025-04-16", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.758, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.758", + "raw_stderr": "0.617", + "latency": "7.1", + "cost_per_test": "0.003345", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.617, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:hard:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 58.571, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "58.571", + "raw_stderr": "2.633", + "latency": "67.813", + "cost_per_test": "0.02575", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.633, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:medium:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 89.295, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "89.295", + "raw_stderr": "1.58", + "latency": "23.607", + "cost_per_test": "0.011845", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.58, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:overall:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 82.208, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "82.208", + "raw_stderr": "1.044", + "latency": "32.84", + "cost_per_test": "0.013647", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.044, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o4-mini-2025-04-16/1ab8aad4-898a-428c-89af-56725b5f0bf6.json b/data/vals_ai/openai/o4-mini-2025-04-16/1ab8aad4-898a-428c-89af-56725b5f0bf6.json new file mode 100644 index 0000000000000000000000000000000000000000..674adb44c34fade8ce1d3eb338a13fcf275daf19 --- /dev/null +++ b/data/vals_ai/openai/o4-mini-2025-04-16/1ab8aad4-898a-428c-89af-56725b5f0bf6.json @@ -0,0 +1,432 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/openai_o4-mini-2025-04-16/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o4-mini-2025-04-16", + "id": "openai/o4-mini-2025-04-16", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o4-mini-2025-04-16", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 86.286, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.286", + "raw_stderr": "0.799", + "latency": "3.623", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.799, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.212, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.212", + "raw_stderr": "0.208", + "latency": "3.557", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.208, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.542, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.542", + "raw_stderr": "0.7", + "latency": "3.993", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.7, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.185, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.185", + "raw_stderr": "0.498", + "latency": "4.03", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.498, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 70.599, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "70.599", + "raw_stderr": "1.493", + "latency": "3.466", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.493, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.285, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.285", + "raw_stderr": "1.676", + "latency": "5.51", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.676, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o4-mini-2025-04-16/35c6a018-44ee-4986-8456-e825d984693c.json b/data/vals_ai/openai/o4-mini-2025-04-16/35c6a018-44ee-4986-8456-e825d984693c.json new file mode 100644 index 0000000000000000000000000000000000000000..32f7197419aa8276808bfdaedf7c0ea6a1a2c754 --- /dev/null +++ b/data/vals_ai/openai/o4-mini-2025-04-16/35c6a018-44ee-4986-8456-e825d984693c.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/openai_o4-mini-2025-04-16/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o4-mini-2025-04-16", + "id": "openai/o4-mini-2025-04-16", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o4-mini-2025-04-16", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 69.139, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "69.139", + "raw_stderr": "1.957", + "latency": "81.959", + "cost_per_test": "0.040605", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.957, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o4-mini-2025-04-16/37fa3d90-e9cc-4501-a284-6c53c0e3a591.json b/data/vals_ai/openai/o4-mini-2025-04-16/37fa3d90-e9cc-4501-a284-6c53c0e3a591.json new file mode 100644 index 0000000000000000000000000000000000000000..bd8e990f63ca08297c5d2b9b8b963cf7a73c4822 --- /dev/null +++ b/data/vals_ai/openai/o4-mini-2025-04-16/37fa3d90-e9cc-4501-a284-6c53c0e3a591.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/openai_o4-mini-2025-04-16/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o4-mini-2025-04-16", + "id": "openai/o4-mini-2025-04-16", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o4-mini-2025-04-16", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 73.737, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.737", + "raw_stderr": "3.127", + "latency": "28.254", + "cost_per_test": "0.012662", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.127, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 75.253, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.253", + "raw_stderr": "3.067", + "latency": "31.754", + "cost_per_test": "0.01326", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.067, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 74.495, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.495", + "raw_stderr": "2.19", + "latency": "30.004", + "cost_per_test": "0.012961", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.19, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o4-mini-2025-04-16/3a09ee00-5b67-4415-bc65-57b22109a563.json b/data/vals_ai/openai/o4-mini-2025-04-16/3a09ee00-5b67-4415-bc65-57b22109a563.json new file mode 100644 index 0000000000000000000000000000000000000000..587d77233a211749eb085d576a3c287881385a22 --- /dev/null +++ b/data/vals_ai/openai/o4-mini-2025-04-16/3a09ee00-5b67-4415-bc65-57b22109a563.json @@ -0,0 +1,356 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/openai_o4-mini-2025-04-16/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o4-mini-2025-04-16", + "id": "openai/o4-mini-2025-04-16", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o4-mini-2025-04-16", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 31.746, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "31.746", + "raw_stderr": "5.841", + "latency": "26.294", + "cost_per_test": "0.023123", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.841, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:automata:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 37.302, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "37.302", + "raw_stderr": "6.187", + "latency": "26.42", + "cost_per_test": "0.025048", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.187, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:calculus:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "33.333", + "raw_stderr": "6.225", + "latency": "72.463", + "cost_per_test": "0.10162", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.225, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 54.135, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "54.135", + "raw_stderr": "5.597", + "latency": "28.93", + "cost_per_test": "0.040058", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.597, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:overall:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 41.061, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "41.061", + "raw_stderr": "3.395", + "latency": "27.215", + "cost_per_test": "0.02941", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.395, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o4-mini-2025-04-16/610a4b2d-7b53-4faa-8876-c17a79c4f5e2.json b/data/vals_ai/openai/o4-mini-2025-04-16/610a4b2d-7b53-4faa-8876-c17a79c4f5e2.json new file mode 100644 index 0000000000000000000000000000000000000000..fb5d8b4b2e82c84091f7d815a08dced0b196c826 --- /dev/null +++ b/data/vals_ai/openai/o4-mini-2025-04-16/610a4b2d-7b53-4faa-8876-c17a79c4f5e2.json @@ -0,0 +1,103 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/openai_o4-mini-2025-04-16/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o4-mini-2025-04-16", + "id": "openai/o4-mini-2025-04-16", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o4-mini-2025-04-16", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 79.665, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.665", + "raw_stderr": "0.967", + "latency": "19.718", + "cost_per_test": "0.010516", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.967, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o4-mini-2025-04-16/84f72438-8bad-4ae5-b97d-64077660c650.json b/data/vals_ai/openai/o4-mini-2025-04-16/84f72438-8bad-4ae5-b97d-64077660c650.json new file mode 100644 index 0000000000000000000000000000000000000000..af54e860a2c6b43429fdafd135936105dbfd252a --- /dev/null +++ b/data/vals_ai/openai/o4-mini-2025-04-16/84f72438-8bad-4ae5-b97d-64077660c650.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/openai_o4-mini-2025-04-16/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o4-mini-2025-04-16", + "id": "openai/o4-mini-2025-04-16", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o4-mini-2025-04-16", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 70.986, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "70.986", + "raw_stderr": "1.28", + "latency": "15.169", + "cost_per_test": "0.012721", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.28, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 64.826, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.826", + "raw_stderr": "0.944", + "latency": "12.996", + "cost_per_test": "0.011654", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.944, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 58.665, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.665", + "raw_stderr": "1.388", + "latency": "10.824", + "cost_per_test": "0.010586", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.388, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o4-mini-2025-04-16/b678008f-4748-41ac-beea-f06598a6e56c.json b/data/vals_ai/openai/o4-mini-2025-04-16/b678008f-4748-41ac-beea-f06598a6e56c.json new file mode 100644 index 0000000000000000000000000000000000000000..343f3f2e38f514372bc2babb258654188c7fb6d4 --- /dev/null +++ b/data/vals_ai/openai/o4-mini-2025-04-16/b678008f-4748-41ac-beea-f06598a6e56c.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/openai_o4-mini-2025-04-16/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o4-mini-2025-04-16", + "id": "openai/o4-mini-2025-04-16", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o4-mini-2025-04-16", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 7.667, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "7.667", + "raw_stderr": "4.647", + "latency": "376.345", + "cost_per_test": "0.114606", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.647, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 2.0, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "2", + "raw_stderr": "0.77", + "latency": "584.039", + "cost_per_test": "0.197548", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.77, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "ioi:overall:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 4.834, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "4.834", + "raw_stderr": "2.355", + "latency": "480.192", + "cost_per_test": "0.156077", + "max_output_tokens": "100000", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.355, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 100000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/openai/o4-mini-2025-04-16/d96aa592-0f73-48a2-8225-f6586798ac2e.json b/data/vals_ai/openai/o4-mini-2025-04-16/d96aa592-0f73-48a2-8225-f6586798ac2e.json new file mode 100644 index 0000000000000000000000000000000000000000..921a074d1c97ccd8a76cfb42175890f8232b5bc9 --- /dev/null +++ b/data/vals_ai/openai/o4-mini-2025-04-16/d96aa592-0f73-48a2-8225-f6586798ac2e.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/openai_o4-mini-2025-04-16/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "o4-mini-2025-04-16", + "id": "openai/o4-mini-2025-04-16", + "developer": "openai", + "additional_details": { + "vals_model_id": "openai/o4-mini-2025-04-16", + "vals_provider": "OpenAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 63.986, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.986", + "raw_stderr": "1.639", + "latency": "8.361", + "cost_per_test": "0.00449", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.639, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 55.012, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "55.012", + "raw_stderr": "1.698", + "latency": "27.127", + "cost_per_test": "0.144774", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.698, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 58.974, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.974", + "raw_stderr": "0.967", + "latency": "17.528", + "cost_per_test": "0.074104", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.967, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:openai/o4-mini-2025-04-16:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 57.925, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.925", + "raw_stderr": "1.685", + "latency": "17.095", + "cost_per_test": "0.073049", + "max_output_tokens": "16384", + "reasoning_effort": "high", + "provider": "OpenAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.685, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 16384, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/poolside/laguna-m.1/0ec5327e-d4ca-4e57-9041-984e7eef3650.json b/data/vals_ai/poolside/laguna-m.1/0ec5327e-d4ca-4e57-9041-984e7eef3650.json new file mode 100644 index 0000000000000000000000000000000000000000..dd91341e133af49d4224246f91f31c204cbb02a3 --- /dev/null +++ b/data/vals_ai/poolside/laguna-m.1/0ec5327e-d4ca-4e57-9041-984e7eef3650.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/poolside_laguna-m.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "laguna-m.1", + "id": "poolside/laguna-m.1", + "developer": "poolside", + "additional_details": { + "vals_model_id": "poolside/laguna-m.1", + "vals_provider": "Poolside" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 85.459, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.459", + "raw_stderr": "0.941", + "latency": "45.298", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.941, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 75.595, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "75.595", + "raw_stderr": "0.229", + "latency": "37.165", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.229, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.459, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.459", + "raw_stderr": "0.791", + "latency": "29.012", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.791, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 75.144, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "75.144", + "raw_stderr": "0.591", + "latency": "42.363", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.591, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 73.756, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "73.756", + "raw_stderr": "1.419", + "latency": "30.517", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.419, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 62.451, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "62.451", + "raw_stderr": "2.269", + "latency": "69.822", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 2.269, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/poolside/laguna-m.1/13144fde-8460-4a01-b1ab-6103222e13ca.json b/data/vals_ai/poolside/laguna-m.1/13144fde-8460-4a01-b1ab-6103222e13ca.json new file mode 100644 index 0000000000000000000000000000000000000000..e34225c6cd3e573233bcad77271e440d073fb00d --- /dev/null +++ b/data/vals_ai/poolside/laguna-m.1/13144fde-8460-4a01-b1ab-6103222e13ca.json @@ -0,0 +1,361 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/poolside_laguna-m.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "laguna-m.1", + "id": "poolside/laguna-m.1", + "developer": "poolside", + "additional_details": { + "vals_model_id": "poolside/laguna-m.1", + "vals_provider": "Poolside" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 23.81, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "23.81", + "raw_stderr": "6.652", + "latency": "1749.41", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 6.652, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 60.536, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "60.536", + "raw_stderr": "3.031", + "latency": "1551.546", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 3.031, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 79.897, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "79.897", + "raw_stderr": "2.885", + "latency": "1172.052", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 2.885, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "2161.717", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 64.6, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "64.6", + "raw_stderr": "2.141", + "latency": "1424.584", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 2.141, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/poolside/laguna-m.1/133751b0-731e-419d-9c8e-87801c05a675.json b/data/vals_ai/poolside/laguna-m.1/133751b0-731e-419d-9c8e-87801c05a675.json new file mode 100644 index 0000000000000000000000000000000000000000..876663797a5e4354e8432a210399abea560d6c6f --- /dev/null +++ b/data/vals_ai/poolside/laguna-m.1/133751b0-731e-419d-9c8e-87801c05a675.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/programbench/poolside_laguna-m.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProgramBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "laguna-m.1", + "id": "poolside/laguna-m.1", + "developer": "poolside", + "additional_details": { + "vals_model_id": "poolside/laguna-m.1", + "vals_provider": "Poolside" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "programbench:almost:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.programbench.almost", + "source_data": { + "dataset_name": "ProgramBench - Almost Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "almost", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Almost Resolved).", + "metric_id": "vals_ai.programbench.almost.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "almost", + "task_name": "Almost Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "4921.898", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "programbench:overall:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.programbench.overall", + "source_data": { + "dataset_name": "ProgramBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Overall).", + "metric_id": "vals_ai.programbench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "4921.898", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "programbench:partial:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.programbench.partial", + "source_data": { + "dataset_name": "ProgramBench - Raw Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "partial", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Raw Pass Rate).", + "metric_id": "vals_ai.programbench.partial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 15.833, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "partial", + "task_name": "Raw Pass Rate", + "dataset_type": "public", + "industry": "coding", + "raw_score": "15.833", + "raw_stderr": "1.474", + "latency": "4921.898", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.474, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "programbench:strict:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.programbench.strict", + "source_data": { + "dataset_name": "ProgramBench - Fully Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "strict", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Fully Resolved).", + "metric_id": "vals_ai.programbench.strict.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "strict", + "task_name": "Fully Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "4921.898", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/poolside/laguna-m.1/17599ec1-4d28-472f-8a5c-5e42bc6885f1.json b/data/vals_ai/poolside/laguna-m.1/17599ec1-4d28-472f-8a5c-5e42bc6885f1.json new file mode 100644 index 0000000000000000000000000000000000000000..49c1e410f460af761c4cdba06355e3b36f8dfdf5 --- /dev/null +++ b/data/vals_ai/poolside/laguna-m.1/17599ec1-4d28-472f-8a5c-5e42bc6885f1.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/poolside_laguna-m.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "laguna-m.1", + "id": "poolside/laguna-m.1", + "developer": "poolside", + "additional_details": { + "vals_model_id": "poolside/laguna-m.1", + "vals_provider": "Poolside" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 10.937, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "10.937", + "raw_stderr": "3.062", + "latency": "3206.87", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 3.062, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/poolside/laguna-m.1/1a47456a-7038-4033-b64a-9a49ce1e62a8.json b/data/vals_ai/poolside/laguna-m.1/1a47456a-7038-4033-b64a-9a49ce1e62a8.json new file mode 100644 index 0000000000000000000000000000000000000000..6d8d11bb4e2f53a4da0557e85dd2f83f8cf13921 --- /dev/null +++ b/data/vals_ai/poolside/laguna-m.1/1a47456a-7038-4033-b64a-9a49ce1e62a8.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/poolside_laguna-m.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "laguna-m.1", + "id": "poolside/laguna-m.1", + "developer": "poolside", + "additional_details": { + "vals_model_id": "poolside/laguna-m.1", + "vals_provider": "Poolside" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 4.545, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "4.545", + "raw_stderr": "1.48", + "latency": "233.988", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.48, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 49.495, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "49.495", + "raw_stderr": "3.547", + "latency": "421.926", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 3.547, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 27.02, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "27.02", + "raw_stderr": "1.922", + "latency": "327.957", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.922, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/poolside/laguna-m.1/1c699847-a1ca-42e1-8216-ed7cd5c5817a.json b/data/vals_ai/poolside/laguna-m.1/1c699847-a1ca-42e1-8216-ed7cd5c5817a.json new file mode 100644 index 0000000000000000000000000000000000000000..f77fd233ee6f73033dad8118530228668e6fea09 --- /dev/null +++ b/data/vals_ai/poolside/laguna-m.1/1c699847-a1ca-42e1-8216-ed7cd5c5817a.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/poolside_laguna-m.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "laguna-m.1", + "id": "poolside/laguna-m.1", + "developer": "poolside", + "additional_details": { + "vals_model_id": "poolside/laguna-m.1", + "vals_provider": "Poolside" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 91.615, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "91.615", + "raw_stderr": "1.545", + "latency": "221.272", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.545, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 27.429, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "27.429", + "raw_stderr": "2.385", + "latency": "524.522", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 2.385, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 65.274, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "65.274", + "raw_stderr": "2.433", + "latency": "496.972", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 2.433, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 61.439, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "61.439", + "raw_stderr": "1.247", + "latency": "414.255", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.247, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/poolside/laguna-m.1/25e99c4e-4e44-4e04-b47d-7b620d26728b.json b/data/vals_ai/poolside/laguna-m.1/25e99c4e-4e44-4e04-b47d-7b620d26728b.json new file mode 100644 index 0000000000000000000000000000000000000000..8463bf8e9778fef98a4979fbdcce25c0a464b9cb --- /dev/null +++ b/data/vals_ai/poolside/laguna-m.1/25e99c4e-4e44-4e04-b47d-7b620d26728b.json @@ -0,0 +1,280 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/poolside_laguna-m.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "laguna-m.1", + "id": "poolside/laguna-m.1", + "developer": "poolside", + "additional_details": { + "vals_model_id": "poolside/laguna-m.1", + "vals_provider": "Poolside" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.956, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.956", + "raw_stderr": "1.665", + "latency": "190.749", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.665, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 56.993, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.993", + "raw_stderr": "1.69", + "latency": "298.129", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.69, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 58.159, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.159", + "raw_stderr": "0.971", + "latency": "211.713", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.971, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 56.527, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.527", + "raw_stderr": "1.692", + "latency": "146.262", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.692, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/poolside/laguna-m.1/6da60ab0-5c0a-472f-8674-71461b303480.json b/data/vals_ai/poolside/laguna-m.1/6da60ab0-5c0a-472f-8674-71461b303480.json new file mode 100644 index 0000000000000000000000000000000000000000..4be38cf5283d5b5e53fe5fa18a40289f323f0613 --- /dev/null +++ b/data/vals_ai/poolside/laguna-m.1/6da60ab0-5c0a-472f-8674-71461b303480.json @@ -0,0 +1,219 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/poolside_laguna-m.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "laguna-m.1", + "id": "poolside/laguna-m.1", + "developer": "poolside", + "additional_details": { + "vals_model_id": "poolside/laguna-m.1", + "vals_provider": "Poolside" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 1.472, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "1.472", + "raw_stderr": "0.344", + "latency": "212.643", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.344, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 1.636, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "1.636", + "raw_stderr": "0.256", + "latency": "212.643", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.256, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 1.799, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "1.799", + "raw_stderr": "0.38", + "latency": "212.643", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.38, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/poolside/laguna-m.1/71572f03-097a-4270-9127-cab9f8898a2a.json b/data/vals_ai/poolside/laguna-m.1/71572f03-097a-4270-9127-cab9f8898a2a.json new file mode 100644 index 0000000000000000000000000000000000000000..24b03c1860dce897edd2cea2f96e12eb745b8110 --- /dev/null +++ b/data/vals_ai/poolside/laguna-m.1/71572f03-097a-4270-9127-cab9f8898a2a.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/poolside_laguna-m.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "laguna-m.1", + "id": "poolside/laguna-m.1", + "developer": "poolside", + "additional_details": { + "vals_model_id": "poolside/laguna-m.1", + "vals_provider": "Poolside" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 75.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "75", + "raw_stderr": "0", + "latency": "0", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 12.222, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "12.222", + "raw_stderr": "2.222", + "latency": "0", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 2.222, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 39.394, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "39.394", + "raw_stderr": "1.212", + "latency": "0", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.212, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 31.835, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "31.835", + "raw_stderr": "1.498", + "latency": "0", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.498, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/poolside/laguna-m.1/93099594-f15d-4667-9c60-5dee7d9c7c45.json b/data/vals_ai/poolside/laguna-m.1/93099594-f15d-4667-9c60-5dee7d9c7c45.json new file mode 100644 index 0000000000000000000000000000000000000000..13fc9be96335889579e9f5d82004d5a6fbff46db --- /dev/null +++ b/data/vals_ai/poolside/laguna-m.1/93099594-f15d-4667-9c60-5dee7d9c7c45.json @@ -0,0 +1,97 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/poolside_laguna-m.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "laguna-m.1", + "id": "poolside/laguna-m.1", + "developer": "poolside", + "additional_details": { + "vals_model_id": "poolside/laguna-m.1", + "vals_provider": "Poolside" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "0", + "raw_stderr": "0", + "latency": "749.317", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/poolside/laguna-m.1/ccf8f38f-97d8-42ef-9380-b62d77d1162f.json b/data/vals_ai/poolside/laguna-m.1/ccf8f38f-97d8-42ef-9380-b62d77d1162f.json new file mode 100644 index 0000000000000000000000000000000000000000..2faa425021256ebf7e9340df5cf27350ef747812 --- /dev/null +++ b/data/vals_ai/poolside/laguna-m.1/ccf8f38f-97d8-42ef-9380-b62d77d1162f.json @@ -0,0 +1,392 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/poolside_laguna-m.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "laguna-m.1", + "id": "poolside/laguna-m.1", + "developer": "poolside", + "additional_details": { + "vals_model_id": "poolside/laguna-m.1", + "vals_provider": "Poolside" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 56.527, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "56.527", + "raw_stderr": "1.692", + "latency": "146.262", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.692, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 26.352, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "26.352", + "raw_stderr": "1.242", + "latency": "1760.996", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.242, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:overall:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 35.272, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "35.272", + "raw_stderr": "1.046", + "latency": "1313.598", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.046, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 60.784, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "60.784", + "raw_stderr": "4.834", + "latency": "1453.86" + }, + "uncertainty": { + "standard_error": { + "value": 4.834, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 31.835, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "31.835", + "raw_stderr": "1.498", + "latency": "0", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.498, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 6.614, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "6.614", + "raw_stderr": "3.225", + "latency": "3206.87", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 3.225, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/poolside/laguna-m.1/f08bb89c-12c3-4e68-b75b-cda38b845fcc.json b/data/vals_ai/poolside/laguna-m.1/f08bb89c-12c3-4e68-b75b-cda38b845fcc.json new file mode 100644 index 0000000000000000000000000000000000000000..ab3262314c778879f72ca2a670766834b1b2f669 --- /dev/null +++ b/data/vals_ai/poolside/laguna-m.1/f08bb89c-12c3-4e68-b75b-cda38b845fcc.json @@ -0,0 +1,996 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/poolside_laguna-m.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "laguna-m.1", + "id": "poolside/laguna-m.1", + "developer": "poolside", + "additional_details": { + "vals_model_id": "poolside/laguna-m.1", + "vals_provider": "Poolside" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 14.226, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "14.226", + "raw_stderr": "1.283", + "latency": "112.859", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.283, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.468, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.468", + "raw_stderr": "1.5", + "latency": "202.121", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.5, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 62.191, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "62.191", + "raw_stderr": "1.456", + "latency": "313.08", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.456, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 70.732, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.732", + "raw_stderr": "2.289", + "latency": "176.788", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 2.289, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.517, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.517", + "raw_stderr": "1.377", + "latency": "150.67", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.377, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 48.194, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "48.194", + "raw_stderr": "1.602", + "latency": "357.603", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.602, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 70.905, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.905", + "raw_stderr": "1.62", + "latency": "137.801", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.62, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 64.304, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "64.304", + "raw_stderr": "2.462", + "latency": "102.222", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 2.462, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 51.499, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "51.499", + "raw_stderr": "1.507", + "latency": "184.274", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.507, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 57.957, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "57.957", + "raw_stderr": "1.347", + "latency": "209.844", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.347, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 72.078, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.078", + "raw_stderr": "1.496", + "latency": "128.291", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.496, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 63.475, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "63.475", + "raw_stderr": "0.447", + "latency": "181.438", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.447, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 67.134, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "67.134", + "raw_stderr": "2.116", + "latency": "111.533", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 2.116, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 72.132, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.132", + "raw_stderr": "1.258", + "latency": "258.418", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.258, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:poolside/laguna-m.1:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.316, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.316", + "raw_stderr": "1.534", + "latency": "94.624", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.534, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/poolside/laguna-xs.2/189e718b-6822-467c-9dd4-f75dd5ccdd3b.json b/data/vals_ai/poolside/laguna-xs.2/189e718b-6822-467c-9dd4-f75dd5ccdd3b.json new file mode 100644 index 0000000000000000000000000000000000000000..2f4186b847323f9416e616ccfeafa549b2fc35a8 --- /dev/null +++ b/data/vals_ai/poolside/laguna-xs.2/189e718b-6822-467c-9dd4-f75dd5ccdd3b.json @@ -0,0 +1,361 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/poolside_laguna-xs.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "laguna-xs.2", + "id": "poolside/laguna-xs.2", + "developer": "poolside", + "additional_details": { + "vals_model_id": "poolside/laguna-xs.2", + "vals_provider": "Poolside" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 16.667, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "16.667", + "raw_stderr": "5.82", + "latency": "1115.08", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 5.82, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 60.92, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "60.92", + "raw_stderr": "3.026", + "latency": "858.75", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 3.026, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 79.381, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "79.381", + "raw_stderr": "2.912", + "latency": "538.706", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 2.912, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1403.244", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 64.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "64", + "raw_stderr": "2.149", + "latency": "759.372", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 2.149, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/poolside/laguna-xs.2/27f7b39b-b203-4668-bdd5-638ed99e859e.json b/data/vals_ai/poolside/laguna-xs.2/27f7b39b-b203-4668-bdd5-638ed99e859e.json new file mode 100644 index 0000000000000000000000000000000000000000..845306a17f7a65ae75d2865a9e26acdd5ac39ccf --- /dev/null +++ b/data/vals_ai/poolside/laguna-xs.2/27f7b39b-b203-4668-bdd5-638ed99e859e.json @@ -0,0 +1,996 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/poolside_laguna-xs.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "laguna-xs.2", + "id": "poolside/laguna-xs.2", + "developer": "poolside", + "additional_details": { + "vals_model_id": "poolside/laguna-xs.2", + "vals_provider": "Poolside" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.821, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.821", + "raw_stderr": "1.811", + "latency": "47.792", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.811, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.115, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.115", + "raw_stderr": "1.683", + "latency": "58.719", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.683, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 68.905, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "68.905", + "raw_stderr": "1.469", + "latency": "120.919", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.469, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.659, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.659", + "raw_stderr": "2.307", + "latency": "50.037", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 2.307, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.109, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.109", + "raw_stderr": "1.673", + "latency": "38.412", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.673, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 53.354, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "53.354", + "raw_stderr": "1.594", + "latency": "192.074", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.594, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.027, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.027", + "raw_stderr": "1.733", + "latency": "13.872", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.733, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 60.892, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "60.892", + "raw_stderr": "2.561", + "latency": "30.61", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 2.561, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 51.862, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "51.862", + "raw_stderr": "1.461", + "latency": "20.132", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.461, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 66.469, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.469", + "raw_stderr": "1.32", + "latency": "55.48", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.32, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 68.939, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "68.939", + "raw_stderr": "1.609", + "latency": "12.964", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.609, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 69.409, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.409", + "raw_stderr": "0.478", + "latency": "56.076", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.478, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 60.321, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "60.321", + "raw_stderr": "2.238", + "latency": "32.327", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 2.238, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.441, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.441", + "raw_stderr": "1.358", + "latency": "96.905", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.358, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.815, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.815", + "raw_stderr": "1.698", + "latency": "14.828", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.698, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/poolside/laguna-xs.2/39040a8a-7344-4dad-9846-db38bed865cb.json b/data/vals_ai/poolside/laguna-xs.2/39040a8a-7344-4dad-9846-db38bed865cb.json new file mode 100644 index 0000000000000000000000000000000000000000..9045ba70845bc18ca89a0817402f554b58c3527a --- /dev/null +++ b/data/vals_ai/poolside/laguna-xs.2/39040a8a-7344-4dad-9846-db38bed865cb.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/poolside_laguna-xs.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "laguna-xs.2", + "id": "poolside/laguna-xs.2", + "developer": "poolside", + "additional_details": { + "vals_model_id": "poolside/laguna-xs.2", + "vals_provider": "Poolside" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 3.841, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "3.841", + "raw_stderr": "1.713", + "latency": "2050.795", + "cost_per_test": "0", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.713, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/poolside/laguna-xs.2/50e7ab3a-dca3-41ba-a9a4-3edb4dc5b1a3.json b/data/vals_ai/poolside/laguna-xs.2/50e7ab3a-dca3-41ba-a9a4-3edb4dc5b1a3.json new file mode 100644 index 0000000000000000000000000000000000000000..0f9e30cab153ccd035cc56ca5ef0711b7f95f9a0 --- /dev/null +++ b/data/vals_ai/poolside/laguna-xs.2/50e7ab3a-dca3-41ba-a9a4-3edb4dc5b1a3.json @@ -0,0 +1,707 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/poolside_laguna-xs.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "laguna-xs.2", + "id": "poolside/laguna-xs.2", + "developer": "poolside", + "additional_details": { + "vals_model_id": "poolside/laguna-xs.2", + "vals_provider": "Poolside" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 9.577, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "9.577", + "raw_stderr": "0.889", + "latency": "663.412", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.889, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 6.794, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "6.794", + "raw_stderr": "0.353", + "latency": "526.624", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.353, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 16.424, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "16.424", + "raw_stderr": "1.515", + "latency": "505.067", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.515, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 18.734, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "18.734", + "raw_stderr": "1.143", + "latency": "481.877", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.143, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 15.297, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "15.297", + "raw_stderr": "2.628", + "latency": "558.171", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 2.628, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 2.591, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "2.591", + "raw_stderr": "0.591", + "latency": "819.584", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.591, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 28.879, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "28.879", + "raw_stderr": "1.489", + "latency": "306.58", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.489, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 22.68, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "22.68", + "raw_stderr": "2.817", + "latency": "527.313", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 2.817, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 20.433, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "20.433", + "raw_stderr": "2.368", + "latency": "330.888", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 2.368, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:overall:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 15.601, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "15.601", + "raw_stderr": "0.326", + "latency": "526.624", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.326, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 5.79, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "5.79", + "raw_stderr": "0.903", + "latency": "540.253", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.903, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/poolside/laguna-xs.2/52785f61-6ee1-4371-970c-701a93085fc3.json b/data/vals_ai/poolside/laguna-xs.2/52785f61-6ee1-4371-970c-701a93085fc3.json new file mode 100644 index 0000000000000000000000000000000000000000..24bf2a82d9e1e4a40f329470761482c71a68d3e2 --- /dev/null +++ b/data/vals_ai/poolside/laguna-xs.2/52785f61-6ee1-4371-970c-701a93085fc3.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/poolside_laguna-xs.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "laguna-xs.2", + "id": "poolside/laguna-xs.2", + "developer": "poolside", + "additional_details": { + "vals_model_id": "poolside/laguna-xs.2", + "vals_provider": "Poolside" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 97.826, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "97.826", + "raw_stderr": "0.813", + "latency": "99.594", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.813, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 29.143, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "29.143", + "raw_stderr": "2.429", + "latency": "522.103", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 2.429, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 72.846, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "72.846", + "raw_stderr": "2.273", + "latency": "291.878", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 2.273, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 66.605, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66.605", + "raw_stderr": "1.142", + "latency": "304.525", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.142, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/poolside/laguna-xs.2/57ba1f58-56d7-44bf-838f-eecea02731d7.json b/data/vals_ai/poolside/laguna-xs.2/57ba1f58-56d7-44bf-838f-eecea02731d7.json new file mode 100644 index 0000000000000000000000000000000000000000..39afc4c95e8e360facba736b66aa5a064888ecf2 --- /dev/null +++ b/data/vals_ai/poolside/laguna-xs.2/57ba1f58-56d7-44bf-838f-eecea02731d7.json @@ -0,0 +1,420 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/poolside_laguna-xs.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "laguna-xs.2", + "id": "poolside/laguna-xs.2", + "developer": "poolside", + "additional_details": { + "vals_model_id": "poolside/laguna-xs.2", + "vals_provider": "Poolside" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.103, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.103", + "raw_stderr": "0.915", + "latency": "12.554", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.915, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 69.947, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "69.947", + "raw_stderr": "0.267", + "latency": "12.122", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.267, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 76.811, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "76.811", + "raw_stderr": "0.771", + "latency": "9.388", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.771, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 71.026, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "71.026", + "raw_stderr": "0.563", + "latency": "11.924", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.563, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 67.98, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "67.98", + "raw_stderr": "1.354", + "latency": "8.765", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.354, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 63.288, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "63.288", + "raw_stderr": "2.145", + "latency": "16.793", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 2.145, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/poolside/laguna-xs.2/725b626d-b7bb-4c22-9854-60f6144ccbe3.json b/data/vals_ai/poolside/laguna-xs.2/725b626d-b7bb-4c22-9854-60f6144ccbe3.json new file mode 100644 index 0000000000000000000000000000000000000000..71b60b23f62e40013f47bfd5a763d6a2d38f753f --- /dev/null +++ b/data/vals_ai/poolside/laguna-xs.2/725b626d-b7bb-4c22-9854-60f6144ccbe3.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/programbench/poolside_laguna-xs.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProgramBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "laguna-xs.2", + "id": "poolside/laguna-xs.2", + "developer": "poolside", + "additional_details": { + "vals_model_id": "poolside/laguna-xs.2", + "vals_provider": "Poolside" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "programbench:almost:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.programbench.almost", + "source_data": { + "dataset_name": "ProgramBench - Almost Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "almost", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Almost Resolved).", + "metric_id": "vals_ai.programbench.almost.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "almost", + "task_name": "Almost Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1541.675", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "programbench:overall:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.programbench.overall", + "source_data": { + "dataset_name": "ProgramBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Overall).", + "metric_id": "vals_ai.programbench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1541.675", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "programbench:partial:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.programbench.partial", + "source_data": { + "dataset_name": "ProgramBench - Raw Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "partial", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Raw Pass Rate).", + "metric_id": "vals_ai.programbench.partial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 17.15, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "partial", + "task_name": "Raw Pass Rate", + "dataset_type": "public", + "industry": "coding", + "raw_score": "17.15", + "raw_stderr": "1.333", + "latency": "1541.675", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "programbench:strict:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.programbench.strict", + "source_data": { + "dataset_name": "ProgramBench - Fully Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "strict", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Fully Resolved).", + "metric_id": "vals_ai.programbench.strict.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "strict", + "task_name": "Fully Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1541.675", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/poolside/laguna-xs.2/7b6ce6d8-12d7-483f-9602-85c546ae08bb.json b/data/vals_ai/poolside/laguna-xs.2/7b6ce6d8-12d7-483f-9602-85c546ae08bb.json new file mode 100644 index 0000000000000000000000000000000000000000..dcbd072106ca37a8d901e557f61612218cf0019c --- /dev/null +++ b/data/vals_ai/poolside/laguna-xs.2/7b6ce6d8-12d7-483f-9602-85c546ae08bb.json @@ -0,0 +1,392 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/poolside_laguna-xs.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "laguna-xs.2", + "id": "poolside/laguna-xs.2", + "developer": "poolside", + "additional_details": { + "vals_model_id": "poolside/laguna-xs.2", + "vals_provider": "Poolside" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 55.012, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "55.012", + "raw_stderr": "1.698", + "latency": "139.86", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.698, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 14.42, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "14.42", + "raw_stderr": "2.018", + "latency": "512.053", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 2.018, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:overall:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 29.148, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "29.148", + "raw_stderr": "0.979", + "latency": "705.3", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.979, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 54.902, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "54.902", + "raw_stderr": "4.927", + "latency": "823.794" + }, + "uncertainty": { + "standard_error": { + "value": 4.927, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 26.966, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "26.966", + "raw_stderr": "2.595", + "latency": "0", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 2.595, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 1.452, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "1.452", + "raw_stderr": "0.81", + "latency": "2050.795", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.81, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/poolside/laguna-xs.2/8cd5cff7-1fb6-4990-b00a-9b580ffbb96b.json b/data/vals_ai/poolside/laguna-xs.2/8cd5cff7-1fb6-4990-b00a-9b580ffbb96b.json new file mode 100644 index 0000000000000000000000000000000000000000..05a3a4a1365588dfe32baafb5f6da3671f679ada --- /dev/null +++ b/data/vals_ai/poolside/laguna-xs.2/8cd5cff7-1fb6-4990-b00a-9b580ffbb96b.json @@ -0,0 +1,97 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/poolside_laguna-xs.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "laguna-xs.2", + "id": "poolside/laguna-xs.2", + "developer": "poolside", + "additional_details": { + "vals_model_id": "poolside/laguna-xs.2", + "vals_provider": "Poolside" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 1.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "1", + "raw_stderr": "1", + "latency": "312.29", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/poolside/laguna-xs.2/a7babbf7-3dc1-4e73-9a00-1f371b6adbda.json b/data/vals_ai/poolside/laguna-xs.2/a7babbf7-3dc1-4e73-9a00-1f371b6adbda.json new file mode 100644 index 0000000000000000000000000000000000000000..e9c5507dc75c4403b2cf6f247e1989e2d60719b2 --- /dev/null +++ b/data/vals_ai/poolside/laguna-xs.2/a7babbf7-3dc1-4e73-9a00-1f371b6adbda.json @@ -0,0 +1,280 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/poolside_laguna-xs.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "laguna-xs.2", + "id": "poolside/laguna-xs.2", + "developer": "poolside", + "additional_details": { + "vals_model_id": "poolside/laguna-xs.2", + "vals_provider": "Poolside" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 59.557, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.557", + "raw_stderr": "1.675", + "latency": "146.965", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.675, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 54.429, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "54.429", + "raw_stderr": "1.7", + "latency": "277.907", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.7, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 56.333, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.333", + "raw_stderr": "0.976", + "latency": "188.244", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.976, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 55.012, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "55.012", + "raw_stderr": "1.698", + "latency": "139.86", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.698, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/poolside/laguna-xs.2/c479e56d-db12-47c6-b15c-40b9447c30f6.json b/data/vals_ai/poolside/laguna-xs.2/c479e56d-db12-47c6-b15c-40b9447c30f6.json new file mode 100644 index 0000000000000000000000000000000000000000..a221ab0cedd89a81a879b63554f8fb06f1df8587 --- /dev/null +++ b/data/vals_ai/poolside/laguna-xs.2/c479e56d-db12-47c6-b15c-40b9447c30f6.json @@ -0,0 +1,219 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/poolside_laguna-xs.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "laguna-xs.2", + "id": "poolside/laguna-xs.2", + "developer": "poolside", + "additional_details": { + "vals_model_id": "poolside/laguna-xs.2", + "vals_provider": "Poolside" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 47.097, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "47.097", + "raw_stderr": "1.427", + "latency": "15.74", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.427, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 59.976, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.976", + "raw_stderr": "0.956", + "latency": "15.74", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 0.956, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 72.854, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "72.854", + "raw_stderr": "1.272", + "latency": "15.74", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.272, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/poolside/laguna-xs.2/efd95187-fcab-4452-b784-ccbf4b88025e.json b/data/vals_ai/poolside/laguna-xs.2/efd95187-fcab-4452-b784-ccbf4b88025e.json new file mode 100644 index 0000000000000000000000000000000000000000..5c8b22857b4964c1f408baf350b91eb581aa8234 --- /dev/null +++ b/data/vals_ai/poolside/laguna-xs.2/efd95187-fcab-4452-b784-ccbf4b88025e.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/poolside_laguna-xs.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "laguna-xs.2", + "id": "poolside/laguna-xs.2", + "developer": "poolside", + "additional_details": { + "vals_model_id": "poolside/laguna-xs.2", + "vals_provider": "Poolside" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 83.333, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "83.333", + "raw_stderr": "8.333", + "latency": "0", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 8.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 11.111, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "11.111", + "raw_stderr": "1.111", + "latency": "0", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 1.111, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 31.515, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "31.515", + "raw_stderr": "3.207", + "latency": "0", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 3.207, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 26.966, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "26.966", + "raw_stderr": "2.595", + "latency": "0", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 2.595, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/poolside/laguna-xs.2/f23afd08-9114-4a1c-866c-4ac656c75dc7.json b/data/vals_ai/poolside/laguna-xs.2/f23afd08-9114-4a1c-866c-4ac656c75dc7.json new file mode 100644 index 0000000000000000000000000000000000000000..9ccb7577e083c05bf116b153de66a68b05c1802a --- /dev/null +++ b/data/vals_ai/poolside/laguna-xs.2/f23afd08-9114-4a1c-866c-4ac656c75dc7.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/poolside_laguna-xs.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "laguna-xs.2", + "id": "poolside/laguna-xs.2", + "developer": "poolside", + "additional_details": { + "vals_model_id": "poolside/laguna-xs.2", + "vals_provider": "Poolside" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 57.576, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "57.576", + "raw_stderr": "3.552", + "latency": "150.261", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 3.552, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 58.586, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "58.586", + "raw_stderr": "3.523", + "latency": "173.734", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 3.523, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:poolside/laguna-xs.2:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 58.081, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "58.081", + "raw_stderr": "2.501", + "latency": "161.998", + "temperature": "1", + "max_output_tokens": "32768", + "provider": "Poolside" + }, + "uncertainty": { + "standard_error": { + "value": 2.501, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "max_tokens": 32768, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/togethercomputer/llama-2-13b/5c52dc15-fa80-4412-96c7-eac71c8b81e4.json b/data/vals_ai/togethercomputer/llama-2-13b/5c52dc15-fa80-4412-96c7-eac71c8b81e4.json new file mode 100644 index 0000000000000000000000000000000000000000..f31bd5b304ffe0fa68eb1d99b49479ef0d4cf961 --- /dev/null +++ b/data/vals_ai/togethercomputer/llama-2-13b/5c52dc15-fa80-4412-96c7-eac71c8b81e4.json @@ -0,0 +1,408 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/togethercomputer_llama-2-13b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "llama-2-13b", + "id": "togethercomputer/llama-2-13b", + "developer": "togethercomputer", + "additional_details": { + "vals_model_id": "together/togethercomputer/llama-2-13b", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:together/togethercomputer/llama-2-13b:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 57.446, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "57.446", + "raw_stderr": "1.136", + "latency": "0.602", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.136, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:together/togethercomputer/llama-2-13b:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 51.063, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "51.063", + "raw_stderr": "0.169", + "latency": "1.101", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.169, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:together/togethercomputer/llama-2-13b:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 50.258, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "50.258", + "raw_stderr": "0.163", + "latency": "3.318", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.163, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:together/togethercomputer/llama-2-13b:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 53.696, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "53.696", + "raw_stderr": "0.733", + "latency": "1.673", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.733, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:together/togethercomputer/llama-2-13b:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 50.156, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "50.156", + "raw_stderr": "1.314", + "latency": "1.64", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.314, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:together/togethercomputer/llama-2-13b:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 59.559, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "59.559", + "raw_stderr": "3.216", + "latency": "1.703", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.216, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/togethercomputer/llama-2-7b/59973e38-07ad-4840-bcd2-9146bc743444.json b/data/vals_ai/togethercomputer/llama-2-7b/59973e38-07ad-4840-bcd2-9146bc743444.json new file mode 100644 index 0000000000000000000000000000000000000000..11c4889bea23c143603650556271d546d75fa26a --- /dev/null +++ b/data/vals_ai/togethercomputer/llama-2-7b/59973e38-07ad-4840-bcd2-9146bc743444.json @@ -0,0 +1,408 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/togethercomputer_llama-2-7b/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "llama-2-7b", + "id": "togethercomputer/llama-2-7b", + "developer": "togethercomputer", + "additional_details": { + "vals_model_id": "together/togethercomputer/llama-2-7b", + "vals_provider": "Together AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:together/togethercomputer/llama-2-7b:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 54.424, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "54.424", + "raw_stderr": "1.003", + "latency": "0.408", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.003, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:together/togethercomputer/llama-2-7b:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 48.545, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "48.545", + "raw_stderr": "0.128", + "latency": "0.742", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.128, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:together/togethercomputer/llama-2-7b:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 50.296, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "50.296", + "raw_stderr": "0.262", + "latency": "1.792", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.262, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:together/togethercomputer/llama-2-7b:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 51.869, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "51.869", + "raw_stderr": "0.63", + "latency": "0.984", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.63, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:together/togethercomputer/llama-2-7b:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 45.218, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "45.218", + "raw_stderr": "0.487", + "latency": "0.774", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.487, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:together/togethercomputer/llama-2-7b:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 60.864, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "60.864", + "raw_stderr": "2.929", + "latency": "1.204", + "max_output_tokens": "4096", + "provider": "Together AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.929, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "max_tokens": 4096, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-2-1212/3f439c93-e4e2-4d69-bf34-73d628775a6c.json b/data/vals_ai/xai/grok-2-1212/3f439c93-e4e2-4d69-bf34-73d628775a6c.json new file mode 100644 index 0000000000000000000000000000000000000000..17bb49f7e16b78f384623b315d72e3a3de825698 --- /dev/null +++ b/data/vals_ai/xai/grok-2-1212/3f439c93-e4e2-4d69-bf34-73d628775a6c.json @@ -0,0 +1,103 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/math500/xai_grok-2-1212/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MATH 500", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "math500", + "benchmark_name": "MATH 500", + "benchmark_updated": "2026-01-09", + "dataset_type": "public", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/math500", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-2-1212", + "id": "xai/grok-2-1212", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-2-1212", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "math500:overall:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.math500.overall", + "source_data": { + "dataset_name": "MATH 500 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/math500" + ], + "additional_details": { + "benchmark_slug": "math500", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/math500" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MATH 500 (Overall).", + "metric_id": "vals_ai.math500.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/math500" + } + }, + "score_details": { + "score": 78.4, + "details": { + "benchmark_slug": "math500", + "benchmark_name": "MATH 500", + "benchmark_updated": "2026-01-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "math", + "raw_score": "78.4", + "raw_stderr": "3.6", + "latency": "20.44", + "cost_per_test": "0.006277", + "temperature": "0", + "top_p": "0.95", + "max_output_tokens": "8192", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.6, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.0, + "top_p": 0.95, + "max_tokens": 8192, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-2-1212/49dcaf66-728c-4b81-8004-bd088ad4a715.json b/data/vals_ai/xai/grok-2-1212/49dcaf66-728c-4b81-8004-bd088ad4a715.json new file mode 100644 index 0000000000000000000000000000000000000000..1dfe9e9fd419e2847ffed1cb162c656d8d17fc25 --- /dev/null +++ b/data/vals_ai/xai/grok-2-1212/49dcaf66-728c-4b81-8004-bd088ad4a715.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/xai_grok-2-1212/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-2-1212", + "id": "xai/grok-2-1212", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-2-1212", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 52.02, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "52.02", + "raw_stderr": "3.55", + "latency": "20.466", + "cost_per_test": "0.011263", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.55, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 49.495, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "49.495", + "raw_stderr": "3.55", + "latency": "19.89", + "cost_per_test": "0.007736", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.55, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 50.758, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "50.758", + "raw_stderr": "2.51", + "latency": "20.178", + "cost_per_test": "0.0095", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.51, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-2-1212/977003aa-01c6-489f-9d8d-bbd9cf3e4e62.json b/data/vals_ai/xai/grok-2-1212/977003aa-01c6-489f-9d8d-bbd9cf3e4e62.json new file mode 100644 index 0000000000000000000000000000000000000000..ff9d81e559fda5de871e4ecfde84710d3a44eaea --- /dev/null +++ b/data/vals_ai/xai/grok-2-1212/977003aa-01c6-489f-9d8d-bbd9cf3e4e62.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/aime/xai_grok-2-1212/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - AIME", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "dataset_type": "public", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-2-1212", + "id": "xai/grok-2-1212", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-2-1212", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aime:aime_2024:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.aime.aime_2024", + "source_data": { + "dataset_name": "AIME - AIME 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/aime" + ], + "additional_details": { + "benchmark_slug": "aime", + "task_key": "aime_2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for AIME (AIME 2024).", + "metric_id": "vals_ai.aime.aime_2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "score_details": { + "score": 19.583, + "details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "task_key": "aime_2024", + "task_name": "AIME 2024", + "dataset_type": "public", + "industry": "math", + "raw_score": "19.583", + "raw_stderr": "1.327", + "latency": "58.028", + "cost_per_test": "0.014342", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.327, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "aime:aime_2025:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.aime.aime_2025", + "source_data": { + "dataset_name": "AIME - AIME 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/aime" + ], + "additional_details": { + "benchmark_slug": "aime", + "task_key": "aime_2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for AIME (AIME 2025).", + "metric_id": "vals_ai.aime.aime_2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "score_details": { + "score": 10.833, + "details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "task_key": "aime_2025", + "task_name": "AIME 2025", + "dataset_type": "public", + "industry": "math", + "raw_score": "10.833", + "raw_stderr": "0.833", + "latency": "57.739", + "cost_per_test": "0.015694", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.833, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "aime:overall:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.aime.overall", + "source_data": { + "dataset_name": "AIME - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/aime" + ], + "additional_details": { + "benchmark_slug": "aime", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for AIME (Overall).", + "metric_id": "vals_ai.aime.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "score_details": { + "score": 15.208, + "details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "math", + "raw_score": "15.208", + "raw_stderr": "0.784", + "latency": "57.884", + "cost_per_test": "0.015018", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.784, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-2-1212/9c348acf-2b98-4c23-a701-9091b9cec088.json b/data/vals_ai/xai/grok-2-1212/9c348acf-2b98-4c23-a701-9091b9cec088.json new file mode 100644 index 0000000000000000000000000000000000000000..1f92c6f025ae04daa516c1a2f67a73f1c6608186 --- /dev/null +++ b/data/vals_ai/xai/grok-2-1212/9c348acf-2b98-4c23-a701-9091b9cec088.json @@ -0,0 +1,288 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/xai_grok-2-1212/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-2-1212", + "id": "xai/grok-2-1212", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-2-1212", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 57.226, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.226", + "raw_stderr": "1.689", + "latency": "3.662", + "cost_per_test": "0", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.689, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 45.221, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "45.221", + "raw_stderr": "1.699", + "latency": "241.891", + "cost_per_test": "0.216047", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.699, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 51.126, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "51.126", + "raw_stderr": "0.981", + "latency": "87.134", + "cost_per_test": "0.115097", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.981, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 50.932, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "50.932", + "raw_stderr": "1.707", + "latency": "15.85", + "cost_per_test": "0.129243", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.707, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-2-1212/b99de1e4-b0c8-4865-842e-48855b9cab3e.json b/data/vals_ai/xai/grok-2-1212/b99de1e4-b0c8-4865-842e-48855b9cab3e.json new file mode 100644 index 0000000000000000000000000000000000000000..68b12036f1b210566456a40f8a1562f23510e130 --- /dev/null +++ b/data/vals_ai/xai/grok-2-1212/b99de1e4-b0c8-4865-842e-48855b9cab3e.json @@ -0,0 +1,1026 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/xai_grok-2-1212/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-2-1212", + "id": "xai/grok-2-1212", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-2-1212", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.053, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.053", + "raw_stderr": "1.294", + "latency": "7.621", + "cost_per_test": "0.005695", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.294, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.214, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.214", + "raw_stderr": "1.445", + "latency": "9.121", + "cost_per_test": "0.005639", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.445, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.71, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.71", + "raw_stderr": "1.217", + "latency": "16.668", + "cost_per_test": "0.008818", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.217, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.098, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.098", + "raw_stderr": "2.106", + "latency": "10.101", + "cost_per_test": "0.006592", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.106, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.924, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.924", + "raw_stderr": "1.352", + "latency": "5.024", + "cost_per_test": "0.004702", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.352, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 62.229, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "62.229", + "raw_stderr": "1.557", + "latency": "23.215", + "cost_per_test": "0.010247", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.557, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.961, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.961", + "raw_stderr": "1.534", + "latency": "3.946", + "cost_per_test": "0.003919", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.534, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 70.866, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.866", + "raw_stderr": "2.328", + "latency": "3.751", + "cost_per_test": "0.006042", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.328, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 59.401, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "59.401", + "raw_stderr": "1.48", + "latency": "5.726", + "cost_per_test": "0.00557", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.48, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.977, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.977", + "raw_stderr": "1.068", + "latency": "12.63", + "cost_per_test": "0.008117", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.068, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.623, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.623", + "raw_stderr": "1.392", + "latency": "3.619", + "cost_per_test": "0.003518", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.392, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.473, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.473", + "raw_stderr": "0.418", + "latency": "8.747", + "cost_per_test": "0.005966", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.418, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 70.341, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.341", + "raw_stderr": "2.045", + "latency": "4.414", + "cost_per_test": "0.003941", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.045, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.523, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.523", + "raw_stderr": "1.12", + "latency": "13.432", + "cost_per_test": "0.007152", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.12, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.704, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.704", + "raw_stderr": "1.369", + "latency": "3.194", + "cost_per_test": "0.003577", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.369, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-2-1212/e508ccfa-42d1-4f5d-b44b-f2917da23c13.json b/data/vals_ai/xai/grok-2-1212/e508ccfa-42d1-4f5d-b44b-f2917da23c13.json new file mode 100644 index 0000000000000000000000000000000000000000..2ba80bf4fdd3158a1451431d375b9a81819f57f8 --- /dev/null +++ b/data/vals_ai/xai/grok-2-1212/e508ccfa-42d1-4f5d-b44b-f2917da23c13.json @@ -0,0 +1,816 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mgsm/xai_grok-2-1212/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MGSM", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "dataset_type": "public", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-2-1212", + "id": "xai/grok-2-1212", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-2-1212", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mgsm:mgsm_bn:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.mgsm.mgsm_bn", + "source_data": { + "dataset_name": "MGSM - Bengali", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_bn", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Bengali).", + "metric_id": "vals_ai.mgsm.mgsm_bn.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 52.4, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_bn", + "task_name": "Bengali", + "dataset_type": "public", + "industry": "math", + "raw_score": "52.4", + "raw_stderr": "3.159", + "latency": "8.366", + "cost_per_test": "0.004885", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.159, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_de:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.mgsm.mgsm_de", + "source_data": { + "dataset_name": "MGSM - German", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_de", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (German).", + "metric_id": "vals_ai.mgsm.mgsm_de.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 89.6, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_de", + "task_name": "German", + "dataset_type": "public", + "industry": "math", + "raw_score": "89.6", + "raw_stderr": "1.931", + "latency": "4.947", + "cost_per_test": "0.002847", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.931, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_en:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.mgsm.mgsm_en", + "source_data": { + "dataset_name": "MGSM - English", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_en", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (English).", + "metric_id": "vals_ai.mgsm.mgsm_en.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 96.8, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_en", + "task_name": "English", + "dataset_type": "public", + "industry": "math", + "raw_score": "96.8", + "raw_stderr": "1.113", + "latency": "3.528", + "cost_per_test": "0.002058", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.113, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_es:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.mgsm.mgsm_es", + "source_data": { + "dataset_name": "MGSM - Spanish", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_es", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Spanish).", + "metric_id": "vals_ai.mgsm.mgsm_es.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 93.2, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_es", + "task_name": "Spanish", + "dataset_type": "public", + "industry": "math", + "raw_score": "93.2", + "raw_stderr": "1.592", + "latency": "4.291", + "cost_per_test": "0.002415", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.592, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_fr:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.mgsm.mgsm_fr", + "source_data": { + "dataset_name": "MGSM - French", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_fr", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (French).", + "metric_id": "vals_ai.mgsm.mgsm_fr.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 84.8, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_fr", + "task_name": "French", + "dataset_type": "public", + "industry": "math", + "raw_score": "84.8", + "raw_stderr": "2.271", + "latency": "4.484", + "cost_per_test": "0.002594", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.271, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_ja:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.mgsm.mgsm_ja", + "source_data": { + "dataset_name": "MGSM - Japanese", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_ja", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Japanese).", + "metric_id": "vals_ai.mgsm.mgsm_ja.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 88.0, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_ja", + "task_name": "Japanese", + "dataset_type": "public", + "industry": "math", + "raw_score": "88", + "raw_stderr": "2.055", + "latency": "5.667", + "cost_per_test": "0.003255", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.055, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_ru:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.mgsm.mgsm_ru", + "source_data": { + "dataset_name": "MGSM - Russian", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_ru", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Russian).", + "metric_id": "vals_ai.mgsm.mgsm_ru.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 92.4, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_ru", + "task_name": "Russian", + "dataset_type": "public", + "industry": "math", + "raw_score": "92.4", + "raw_stderr": "1.676", + "latency": "5.009", + "cost_per_test": "0.002874", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.676, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_sw:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.mgsm.mgsm_sw", + "source_data": { + "dataset_name": "MGSM - Swahili", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_sw", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Swahili).", + "metric_id": "vals_ai.mgsm.mgsm_sw.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 84.8, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_sw", + "task_name": "Swahili", + "dataset_type": "public", + "industry": "math", + "raw_score": "84.8", + "raw_stderr": "2.271", + "latency": "1.957", + "cost_per_test": "0.002114", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.271, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_te:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.mgsm.mgsm_te", + "source_data": { + "dataset_name": "MGSM - Telugu", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_te", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Telugu).", + "metric_id": "vals_ai.mgsm.mgsm_te.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 83.6, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_te", + "task_name": "Telugu", + "dataset_type": "public", + "industry": "math", + "raw_score": "83.6", + "raw_stderr": "2.342", + "latency": "18.976", + "cost_per_test": "0.011334", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.342, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_th:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.mgsm.mgsm_th", + "source_data": { + "dataset_name": "MGSM - Thai", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_th", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Thai).", + "metric_id": "vals_ai.mgsm.mgsm_th.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 92.4, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_th", + "task_name": "Thai", + "dataset_type": "public", + "industry": "math", + "raw_score": "92.4", + "raw_stderr": "1.676", + "latency": "7.354", + "cost_per_test": "0.004538", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.676, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_zh:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.mgsm.mgsm_zh", + "source_data": { + "dataset_name": "MGSM - Chinese", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_zh", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Chinese).", + "metric_id": "vals_ai.mgsm.mgsm_zh.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 89.6, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_zh", + "task_name": "Chinese", + "dataset_type": "public", + "industry": "math", + "raw_score": "89.6", + "raw_stderr": "1.931", + "latency": "4.409", + "cost_per_test": "0.002436", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.931, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:overall:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.mgsm.overall", + "source_data": { + "dataset_name": "MGSM - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Overall).", + "metric_id": "vals_ai.mgsm.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 86.146, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "math", + "raw_score": "86.146", + "raw_stderr": "2.001", + "latency": "6.272", + "cost_per_test": "0.003759", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.001, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-2-1212/f6ceed08-3b84-4f59-a2bb-3dd0e39bf02d.json b/data/vals_ai/xai/grok-2-1212/f6ceed08-3b84-4f59-a2bb-3dd0e39bf02d.json new file mode 100644 index 0000000000000000000000000000000000000000..05754fd3c06bffe87e061fa3806057ec2df27028 --- /dev/null +++ b/data/vals_ai/xai/grok-2-1212/f6ceed08-3b84-4f59-a2bb-3dd0e39bf02d.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/xai_grok-2-1212/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-2-1212", + "id": "xai/grok-2-1212", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-2-1212", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 85.714, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "85.714", + "raw_stderr": "1.95", + "latency": "1.554", + "cost_per_test": "0.002082", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.95, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 3.429, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "3.429", + "raw_stderr": "0.973", + "latency": "3.192", + "cost_per_test": "0.003925", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.973, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 26.893, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "26.893", + "raw_stderr": "2.266", + "latency": "2.387", + "cost_per_test": "0.003005", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.266, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 38.679, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "38.679", + "raw_stderr": "1.048", + "latency": "2.378", + "cost_per_test": "0.003004", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.048, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-2-1212/f78d4514-7c2e-4e0f-ab59-ef523c5fb41e.json b/data/vals_ai/xai/grok-2-1212/f78d4514-7c2e-4e0f-ab59-ef523c5fb41e.json new file mode 100644 index 0000000000000000000000000000000000000000..3bc835f24d2f4cc439c7c93f5feb5a20675422b4 --- /dev/null +++ b/data/vals_ai/xai/grok-2-1212/f78d4514-7c2e-4e0f-ab59-ef523c5fb41e.json @@ -0,0 +1,498 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medqa/xai_grok-2-1212/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedQA", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "dataset_type": "public", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-2-1212", + "id": "xai/grok-2-1212", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-2-1212", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medqa:asian:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.medqa.asian", + "source_data": { + "dataset_name": "MedQA - Asian", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "asian", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Asian).", + "metric_id": "vals_ai.medqa.asian.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 91.95, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "asian", + "task_name": "Asian", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "91.95", + "raw_stderr": "1.194", + "latency": "4.047", + "cost_per_test": "0.003272", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.194, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:black:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.medqa.black", + "source_data": { + "dataset_name": "MedQA - Black", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "black", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Black).", + "metric_id": "vals_ai.medqa.black.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 92.2, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "black", + "task_name": "Black", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "92.2", + "raw_stderr": "1.177", + "latency": "4.204", + "cost_per_test": "0.003354", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.177, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:hispanic:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.medqa.hispanic", + "source_data": { + "dataset_name": "MedQA - Hispanic", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "hispanic", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Hispanic).", + "metric_id": "vals_ai.medqa.hispanic.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 91.7, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "hispanic", + "task_name": "Hispanic", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "91.7", + "raw_stderr": "1.211", + "latency": "4.109", + "cost_per_test": "0.003321", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.211, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:indigenous:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.medqa.indigenous", + "source_data": { + "dataset_name": "MedQA - Indigenous", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "indigenous", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Indigenous).", + "metric_id": "vals_ai.medqa.indigenous.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 91.6, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "indigenous", + "task_name": "Indigenous", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "91.6", + "raw_stderr": "1.217", + "latency": "4.288", + "cost_per_test": "0.003412", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.217, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:overall:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.medqa.overall", + "source_data": { + "dataset_name": "MedQA - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Overall).", + "metric_id": "vals_ai.medqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 92.317, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "92.317", + "raw_stderr": "0.477", + "latency": "4.093", + "cost_per_test": "0.003276", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.477, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:unbiased:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.medqa.unbiased", + "source_data": { + "dataset_name": "MedQA - Unbiased", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "unbiased", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Unbiased).", + "metric_id": "vals_ai.medqa.unbiased.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 93.95, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "unbiased", + "task_name": "Unbiased", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "93.95", + "raw_stderr": "1.047", + "latency": "3.906", + "cost_per_test": "0.003053", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.047, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:white:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.medqa.white", + "source_data": { + "dataset_name": "MedQA - White", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "white", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (White).", + "metric_id": "vals_ai.medqa.white.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 92.5, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "white", + "task_name": "White", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "92.5", + "raw_stderr": "1.156", + "latency": "4.005", + "cost_per_test": "0.003244", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.156, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-2-1212/f9f1b688-db5e-45e7-8ea0-ad2c7a9f66e1.json b/data/vals_ai/xai/grok-2-1212/f9f1b688-db5e-45e7-8ea0-ad2c7a9f66e1.json new file mode 100644 index 0000000000000000000000000000000000000000..179d3f358cfa81841be297672eb9b8ae9e2bb448 --- /dev/null +++ b/data/vals_ai/xai/grok-2-1212/f9f1b688-db5e-45e7-8ea0-ad2c7a9f66e1.json @@ -0,0 +1,225 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/xai_grok-2-1212/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-2-1212", + "id": "xai/grok-2-1212", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-2-1212", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 54.865, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "54.865", + "raw_stderr": "1.423", + "latency": "10.621", + "cost_per_test": "0.006894", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.423, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 67.048, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.048", + "raw_stderr": "0.918", + "latency": "10.621", + "cost_per_test": "0.006894", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.918, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:grok/grok-2-1212:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 79.231, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "79.231", + "raw_stderr": "1.16", + "latency": "10.621", + "cost_per_test": "0.006894", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.16, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-2-vision-1212/02b14c89-5710-4999-b40a-1ed973ba31bb.json b/data/vals_ai/xai/grok-2-vision-1212/02b14c89-5710-4999-b40a-1ed973ba31bb.json new file mode 100644 index 0000000000000000000000000000000000000000..b245617bb18ffabcfdd02463c1c170b8f1eb0b7b --- /dev/null +++ b/data/vals_ai/xai/grok-2-vision-1212/02b14c89-5710-4999-b40a-1ed973ba31bb.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/xai_grok-2-vision-1212/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-2-vision-1212", + "id": "xai/grok-2-vision-1212", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-2-vision-1212", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:grok/grok-2-vision-1212:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 57.25, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "57.25", + "raw_stderr": "1.189", + "latency": "18.355", + "cost_per_test": "0.005402", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.189, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-2-vision-1212/e3dbbd29-79bf-48e4-bd3d-10dbe1fc3a53.json b/data/vals_ai/xai/grok-2-vision-1212/e3dbbd29-79bf-48e4-bd3d-10dbe1fc3a53.json new file mode 100644 index 0000000000000000000000000000000000000000..dde53ae90561593f89749323fe4d2311cc8eb860 --- /dev/null +++ b/data/vals_ai/xai/grok-2-vision-1212/e3dbbd29-79bf-48e4-bd3d-10dbe1fc3a53.json @@ -0,0 +1,225 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/xai_grok-2-vision-1212/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-2-vision-1212", + "id": "xai/grok-2-vision-1212", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-2-vision-1212", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:grok/grok-2-vision-1212:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 4.531, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "4.531", + "raw_stderr": "1.156", + "latency": "0", + "cost_per_test": "0.003715", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.156, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:grok/grok-2-vision-1212:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 3.458, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "3.458", + "raw_stderr": "0.719", + "latency": "0", + "cost_per_test": "0.003772", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.719, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:grok/grok-2-vision-1212:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 2.385, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "2.385", + "raw_stderr": "0.854", + "latency": "0", + "cost_per_test": "0.00383", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "0", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.854, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-3-mini-fast-high-reasoning/09d1d84a-332a-437f-a1f5-253b1e242749.json b/data/vals_ai/xai/grok-3-mini-fast-high-reasoning/09d1d84a-332a-437f-a1f5-253b1e242749.json new file mode 100644 index 0000000000000000000000000000000000000000..1c0471b16276c4c3f0fe6b390a924444ec5b4cd4 --- /dev/null +++ b/data/vals_ai/xai/grok-3-mini-fast-high-reasoning/09d1d84a-332a-437f-a1f5-253b1e242749.json @@ -0,0 +1,888 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mgsm/xai_grok-3-mini-fast-high-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MGSM", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "dataset_type": "public", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-3-mini-fast-high-reasoning", + "id": "xai/grok-3-mini-fast-high-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-3-mini-fast-high-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mgsm:mgsm_bn:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_bn", + "source_data": { + "dataset_name": "MGSM - Bengali", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_bn", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Bengali).", + "metric_id": "vals_ai.mgsm.mgsm_bn.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 88.8, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_bn", + "task_name": "Bengali", + "dataset_type": "public", + "industry": "math", + "raw_score": "88.8", + "raw_stderr": "1.995", + "latency": "11.998", + "cost_per_test": "0.001078", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.995, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_de:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_de", + "source_data": { + "dataset_name": "MGSM - German", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_de", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (German).", + "metric_id": "vals_ai.mgsm.mgsm_de.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 92.4, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_de", + "task_name": "German", + "dataset_type": "public", + "industry": "math", + "raw_score": "92.4", + "raw_stderr": "1.676", + "latency": "7.278", + "cost_per_test": "0.000686", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.676, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_en:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_en", + "source_data": { + "dataset_name": "MGSM - English", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_en", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (English).", + "metric_id": "vals_ai.mgsm.mgsm_en.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 92.0, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_en", + "task_name": "English", + "dataset_type": "public", + "industry": "math", + "raw_score": "92", + "raw_stderr": "1.716", + "latency": "6.89", + "cost_per_test": "0.000539", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.716, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_es:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_es", + "source_data": { + "dataset_name": "MGSM - Spanish", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_es", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Spanish).", + "metric_id": "vals_ai.mgsm.mgsm_es.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 92.4, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_es", + "task_name": "Spanish", + "dataset_type": "public", + "industry": "math", + "raw_score": "92.4", + "raw_stderr": "1.676", + "latency": "7.851", + "cost_per_test": "0.000639", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.676, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_fr:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_fr", + "source_data": { + "dataset_name": "MGSM - French", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_fr", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (French).", + "metric_id": "vals_ai.mgsm.mgsm_fr.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 86.0, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_fr", + "task_name": "French", + "dataset_type": "public", + "industry": "math", + "raw_score": "86", + "raw_stderr": "2.194", + "latency": "7.678", + "cost_per_test": "0.000676", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.194, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_ja:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_ja", + "source_data": { + "dataset_name": "MGSM - Japanese", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_ja", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Japanese).", + "metric_id": "vals_ai.mgsm.mgsm_ja.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 86.8, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_ja", + "task_name": "Japanese", + "dataset_type": "public", + "industry": "math", + "raw_score": "86.8", + "raw_stderr": "2.141", + "latency": "8.635", + "cost_per_test": "0.00077", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.141, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_ru:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_ru", + "source_data": { + "dataset_name": "MGSM - Russian", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_ru", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Russian).", + "metric_id": "vals_ai.mgsm.mgsm_ru.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 93.6, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_ru", + "task_name": "Russian", + "dataset_type": "public", + "industry": "math", + "raw_score": "93.6", + "raw_stderr": "1.548", + "latency": "6.071", + "cost_per_test": "0.000556", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.548, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_sw:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_sw", + "source_data": { + "dataset_name": "MGSM - Swahili", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_sw", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Swahili).", + "metric_id": "vals_ai.mgsm.mgsm_sw.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 92.0, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_sw", + "task_name": "Swahili", + "dataset_type": "public", + "industry": "math", + "raw_score": "92", + "raw_stderr": "1.716", + "latency": "16.42", + "cost_per_test": "0.000797", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.716, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_te:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_te", + "source_data": { + "dataset_name": "MGSM - Telugu", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_te", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Telugu).", + "metric_id": "vals_ai.mgsm.mgsm_te.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 87.2, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_te", + "task_name": "Telugu", + "dataset_type": "public", + "industry": "math", + "raw_score": "87.2", + "raw_stderr": "2.113", + "latency": "23.31", + "cost_per_test": "0.001628", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.113, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_th:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_th", + "source_data": { + "dataset_name": "MGSM - Thai", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_th", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Thai).", + "metric_id": "vals_ai.mgsm.mgsm_th.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 92.4, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_th", + "task_name": "Thai", + "dataset_type": "public", + "industry": "math", + "raw_score": "92.4", + "raw_stderr": "1.676", + "latency": "8.829", + "cost_per_test": "0.000855", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.676, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_zh:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_zh", + "source_data": { + "dataset_name": "MGSM - Chinese", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_zh", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Chinese).", + "metric_id": "vals_ai.mgsm.mgsm_zh.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 91.2, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_zh", + "task_name": "Chinese", + "dataset_type": "public", + "industry": "math", + "raw_score": "91.2", + "raw_stderr": "1.792", + "latency": "5.789", + "cost_per_test": "0.000549", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.792, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mgsm:overall:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.mgsm.overall", + "source_data": { + "dataset_name": "MGSM - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Overall).", + "metric_id": "vals_ai.mgsm.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 90.436, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "math", + "raw_score": "90.436", + "raw_stderr": "1.84", + "latency": "10.068", + "cost_per_test": "0.000798", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.84, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-3-mini-fast-high-reasoning/0c847455-1be0-42a0-bc3d-45425f0a7cd0.json b/data/vals_ai/xai/grok-3-mini-fast-high-reasoning/0c847455-1be0-42a0-bc3d-45425f0a7cd0.json new file mode 100644 index 0000000000000000000000000000000000000000..4efffdb779c534084c22b9e5667de311b69fccae --- /dev/null +++ b/data/vals_ai/xai/grok-3-mini-fast-high-reasoning/0c847455-1be0-42a0-bc3d-45425f0a7cd0.json @@ -0,0 +1,107 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/math500/xai_grok-3-mini-fast-high-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MATH 500", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "math500", + "benchmark_name": "MATH 500", + "benchmark_updated": "2026-01-09", + "dataset_type": "public", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/math500", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-3-mini-fast-high-reasoning", + "id": "xai/grok-3-mini-fast-high-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-3-mini-fast-high-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "math500:overall:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.math500.overall", + "source_data": { + "dataset_name": "MATH 500 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/math500" + ], + "additional_details": { + "benchmark_slug": "math500", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/math500" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MATH 500 (Overall).", + "metric_id": "vals_ai.math500.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/math500" + } + }, + "score_details": { + "score": 94.2, + "details": { + "benchmark_slug": "math500", + "benchmark_name": "MATH 500", + "benchmark_updated": "2026-01-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "math", + "raw_score": "94.2", + "raw_stderr": "1.045", + "latency": "22.766", + "cost_per_test": "0.001879", + "temperature": "0", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.045, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-3-mini-fast-high-reasoning/260119b0-a882-4756-bfa7-0eb470c4f9db.json b/data/vals_ai/xai/grok-3-mini-fast-high-reasoning/260119b0-a882-4756-bfa7-0eb470c4f9db.json new file mode 100644 index 0000000000000000000000000000000000000000..84f979b17051da30e846c269cbe9f74e203692d8 --- /dev/null +++ b/data/vals_ai/xai/grok-3-mini-fast-high-reasoning/260119b0-a882-4756-bfa7-0eb470c4f9db.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/aime/xai_grok-3-mini-fast-high-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - AIME", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "dataset_type": "public", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-3-mini-fast-high-reasoning", + "id": "xai/grok-3-mini-fast-high-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-3-mini-fast-high-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aime:aime_2024:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.aime.aime_2024", + "source_data": { + "dataset_name": "AIME - AIME 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/aime" + ], + "additional_details": { + "benchmark_slug": "aime", + "task_key": "aime_2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for AIME (AIME 2024).", + "metric_id": "vals_ai.aime.aime_2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "score_details": { + "score": 88.75, + "details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "task_key": "aime_2024", + "task_name": "AIME 2024", + "dataset_type": "public", + "industry": "math", + "raw_score": "88.75", + "raw_stderr": "0.877", + "latency": "94.102", + "cost_per_test": "0.007605", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.877, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "aime:aime_2025:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.aime.aime_2025", + "source_data": { + "dataset_name": "AIME - AIME 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/aime" + ], + "additional_details": { + "benchmark_slug": "aime", + "task_key": "aime_2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for AIME (AIME 2025).", + "metric_id": "vals_ai.aime.aime_2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "score_details": { + "score": 81.25, + "details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "task_key": "aime_2025", + "task_name": "AIME 2025", + "dataset_type": "public", + "industry": "math", + "raw_score": "81.25", + "raw_stderr": "0.61", + "latency": "110.39", + "cost_per_test": "0.008957", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.61, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "aime:overall:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.aime.overall", + "source_data": { + "dataset_name": "AIME - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/aime" + ], + "additional_details": { + "benchmark_slug": "aime", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for AIME (Overall).", + "metric_id": "vals_ai.aime.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "score_details": { + "score": 85.0, + "details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "math", + "raw_score": "85", + "raw_stderr": "0.534", + "latency": "102.246", + "cost_per_test": "0.008281", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.534, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-3-mini-fast-high-reasoning/3d18415e-6070-47ac-8871-20bb094210fd.json b/data/vals_ai/xai/grok-3-mini-fast-high-reasoning/3d18415e-6070-47ac-8871-20bb094210fd.json new file mode 100644 index 0000000000000000000000000000000000000000..b92a8f68926943a6c9c99d201c767407fc33b13c --- /dev/null +++ b/data/vals_ai/xai/grok-3-mini-fast-high-reasoning/3d18415e-6070-47ac-8871-20bb094210fd.json @@ -0,0 +1,240 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/xai_grok-3-mini-fast-high-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-3-mini-fast-high-reasoning", + "id": "xai/grok-3-mini-fast-high-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-3-mini-fast-high-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 60.18, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.18", + "raw_stderr": "1.4", + "latency": "25.447", + "cost_per_test": "0.002142", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.4, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 71.136, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.136", + "raw_stderr": "0.889", + "latency": "25.447", + "cost_per_test": "0.002142", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.889, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 82.093, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "82.093", + "raw_stderr": "1.096", + "latency": "25.447", + "cost_per_test": "0.002142", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.096, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-3-mini-fast-high-reasoning/5e0991a9-5dca-4d24-8ab9-47e1e9debdb6.json b/data/vals_ai/xai/grok-3-mini-fast-high-reasoning/5e0991a9-5dca-4d24-8ab9-47e1e9debdb6.json new file mode 100644 index 0000000000000000000000000000000000000000..25db8fa32576a5567d94560bf1349497c29dfe19 --- /dev/null +++ b/data/vals_ai/xai/grok-3-mini-fast-high-reasoning/5e0991a9-5dca-4d24-8ab9-47e1e9debdb6.json @@ -0,0 +1,308 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/xai_grok-3-mini-fast-high-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-3-mini-fast-high-reasoning", + "id": "xai/grok-3-mini-fast-high-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-3-mini-fast-high-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 65.851, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.851", + "raw_stderr": "1.619", + "latency": "7.748", + "cost_per_test": "0.001098", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.619, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 55.828, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "55.828", + "raw_stderr": "1.695", + "latency": "20.701", + "cost_per_test": "0.032067", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.695, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.111, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.111", + "raw_stderr": "0.957", + "latency": "14.386", + "cost_per_test": "0.017292", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.957, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.655, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.655", + "raw_stderr": "1.66", + "latency": "14.71", + "cost_per_test": "0.018712", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.66, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-3-mini-fast-high-reasoning/602b2e5a-a5be-40fe-9482-370dc5da0d96.json b/data/vals_ai/xai/grok-3-mini-fast-high-reasoning/602b2e5a-a5be-40fe-9482-370dc5da0d96.json new file mode 100644 index 0000000000000000000000000000000000000000..4c0eef7bfe2417aeecc0c51816c5f59ebaca9804 --- /dev/null +++ b/data/vals_ai/xai/grok-3-mini-fast-high-reasoning/602b2e5a-a5be-40fe-9482-370dc5da0d96.json @@ -0,0 +1,456 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/xai_grok-3-mini-fast-high-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-3-mini-fast-high-reasoning", + "id": "xai/grok-3-mini-fast-high-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-3-mini-fast-high-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.654, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.654", + "raw_stderr": "0.895", + "latency": "6.571", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.895, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.821, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.821", + "raw_stderr": "0.201", + "latency": "4.869", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.201, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 88.201, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.201", + "raw_stderr": "0.563", + "latency": "4.359", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.563, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.143, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.143", + "raw_stderr": "0.42", + "latency": "5.095", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.42, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.61, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.61", + "raw_stderr": "0.894", + "latency": "4.084", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.894, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.429, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.429", + "raw_stderr": "1.567", + "latency": "5.591", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.567, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-3-mini-fast-high-reasoning/b998b977-3023-456b-82f6-ee5dfcdebd46.json b/data/vals_ai/xai/grok-3-mini-fast-high-reasoning/b998b977-3023-456b-82f6-ee5dfcdebd46.json new file mode 100644 index 0000000000000000000000000000000000000000..bf16bfab2cce55c5bae1339ab25fb964efeb7af9 --- /dev/null +++ b/data/vals_ai/xai/grok-3-mini-fast-high-reasoning/b998b977-3023-456b-82f6-ee5dfcdebd46.json @@ -0,0 +1,249 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/xai_grok-3-mini-fast-high-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-3-mini-fast-high-reasoning", + "id": "xai/grok-3-mini-fast-high-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-3-mini-fast-high-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 80.303, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.303", + "raw_stderr": "2.826", + "latency": "40.258", + "cost_per_test": "0.004123", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.826, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 78.283, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.283", + "raw_stderr": "2.955", + "latency": "40.979", + "cost_per_test": "0.003637", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.955, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 79.293, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.293", + "raw_stderr": "2.044", + "latency": "40.618", + "cost_per_test": "0.00388", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.044, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-3-mini-fast-high-reasoning/bbbf4631-26a4-4385-867c-802a8bb81039.json b/data/vals_ai/xai/grok-3-mini-fast-high-reasoning/bbbf4631-26a4-4385-867c-802a8bb81039.json new file mode 100644 index 0000000000000000000000000000000000000000..64d8580cc5dc472e3809532df268147cfd5dde64 --- /dev/null +++ b/data/vals_ai/xai/grok-3-mini-fast-high-reasoning/bbbf4631-26a4-4385-867c-802a8bb81039.json @@ -0,0 +1,320 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/xai_grok-3-mini-fast-high-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-3-mini-fast-high-reasoning", + "id": "xai/grok-3-mini-fast-high-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-3-mini-fast-high-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.758, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.758", + "raw_stderr": "0.617", + "latency": "30.579", + "cost_per_test": "0.002727", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "40000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.617, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 40000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:hard:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 44.0, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "44", + "raw_stderr": "2.653", + "latency": "457.186", + "cost_per_test": "0.013734", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "40000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.653, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 40000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:medium:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 85.901, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "85.901", + "raw_stderr": "1.778", + "latency": "153.221", + "cost_per_test": "0.007709", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "40000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.778, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 40000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:overall:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 76.22, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "76.22", + "raw_stderr": "1.084", + "latency": "213.662", + "cost_per_test": "0.008057", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "40000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.084, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 40000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-3-mini-fast-high-reasoning/c5f6f393-e826-442c-9a66-5662bd5fe021.json b/data/vals_ai/xai/grok-3-mini-fast-high-reasoning/c5f6f393-e826-442c-9a66-5662bd5fe021.json new file mode 100644 index 0000000000000000000000000000000000000000..b1a5c6dcce5cee021a992b91ca042c90b6bee2a6 --- /dev/null +++ b/data/vals_ai/xai/grok-3-mini-fast-high-reasoning/c5f6f393-e826-442c-9a66-5662bd5fe021.json @@ -0,0 +1,533 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medqa/xai_grok-3-mini-fast-high-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedQA", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "dataset_type": "public", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-3-mini-fast-high-reasoning", + "id": "xai/grok-3-mini-fast-high-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-3-mini-fast-high-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medqa:asian:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.medqa.asian", + "source_data": { + "dataset_name": "MedQA - Asian", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "asian", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Asian).", + "metric_id": "vals_ai.medqa.asian.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 89.9, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "asian", + "task_name": "Asian", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "89.9", + "raw_stderr": "0.674", + "latency": "7.253", + "cost_per_test": "0.000769", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.674, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "medqa:black:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.medqa.black", + "source_data": { + "dataset_name": "MedQA - Black", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "black", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Black).", + "metric_id": "vals_ai.medqa.black.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 89.7, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "black", + "task_name": "Black", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "89.7", + "raw_stderr": "0.68", + "latency": "7.184", + "cost_per_test": "0.00078", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.68, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "medqa:hispanic:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.medqa.hispanic", + "source_data": { + "dataset_name": "MedQA - Hispanic", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "hispanic", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Hispanic).", + "metric_id": "vals_ai.medqa.hispanic.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 90.65, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "hispanic", + "task_name": "Hispanic", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "90.65", + "raw_stderr": "0.651", + "latency": "6.997", + "cost_per_test": "0.000761", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.651, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "medqa:indigenous:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.medqa.indigenous", + "source_data": { + "dataset_name": "MedQA - Indigenous", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "indigenous", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Indigenous).", + "metric_id": "vals_ai.medqa.indigenous.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 89.25, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "indigenous", + "task_name": "Indigenous", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "89.25", + "raw_stderr": "0.693", + "latency": "7.067", + "cost_per_test": "0.000773", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.693, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "medqa:overall:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.medqa.overall", + "source_data": { + "dataset_name": "MedQA - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Overall).", + "metric_id": "vals_ai.medqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 90.1, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "90.1", + "raw_stderr": "0.273", + "latency": "7.065", + "cost_per_test": "0.000765", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.273, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "medqa:unbiased:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.medqa.unbiased", + "source_data": { + "dataset_name": "MedQA - Unbiased", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "unbiased", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Unbiased).", + "metric_id": "vals_ai.medqa.unbiased.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 90.8, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "unbiased", + "task_name": "Unbiased", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "90.8", + "raw_stderr": "0.646", + "latency": "6.93", + "cost_per_test": "0.000744", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.646, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "medqa:white:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.medqa.white", + "source_data": { + "dataset_name": "MedQA - White", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "white", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (White).", + "metric_id": "vals_ai.medqa.white.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 90.3, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "white", + "task_name": "White", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "90.3", + "raw_stderr": "0.662", + "latency": "6.958", + "cost_per_test": "0.000761", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.662, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-3-mini-fast-high-reasoning/fcf89ca4-c8a5-49d2-98a3-c5e5365c7733.json b/data/vals_ai/xai/grok-3-mini-fast-high-reasoning/fcf89ca4-c8a5-49d2-98a3-c5e5365c7733.json new file mode 100644 index 0000000000000000000000000000000000000000..84b1b6de37af283b6f760c96313a093ed36a7d1b --- /dev/null +++ b/data/vals_ai/xai/grok-3-mini-fast-high-reasoning/fcf89ca4-c8a5-49d2-98a3-c5e5365c7733.json @@ -0,0 +1,1101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/xai_grok-3-mini-fast-high-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-3-mini-fast-high-reasoning", + "id": "xai/grok-3-mini-fast-high-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-3-mini-fast-high-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.934, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.934", + "raw_stderr": "1.072", + "latency": "9.654", + "cost_per_test": "0.00126", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.072, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.072, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.072", + "raw_stderr": "1.194", + "latency": "22.144", + "cost_per_test": "0.001932", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.194, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.134, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.134", + "raw_stderr": "0.925", + "latency": "27.757", + "cost_per_test": "0.002691", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.925, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.146, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.146", + "raw_stderr": "1.804", + "latency": "15.194", + "cost_per_test": "0.001848", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.804, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.019, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.019", + "raw_stderr": "1.194", + "latency": "8.779", + "cost_per_test": "0.001172", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.194, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.367, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.367", + "raw_stderr": "1.365", + "latency": "47.029", + "cost_per_test": "0.004227", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.365, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.406, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.406", + "raw_stderr": "1.485", + "latency": "8.07", + "cost_per_test": "0.001042", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.485, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 69.816, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.816", + "raw_stderr": "2.352", + "latency": "7.998", + "cost_per_test": "0.001357", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.352, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 63.669, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "63.669", + "raw_stderr": "1.449", + "latency": "10.826", + "cost_per_test": "0.001425", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.449, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.266, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.266", + "raw_stderr": "0.768", + "latency": "26.956", + "cost_per_test": "0.002248", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.768, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.545, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.545", + "raw_stderr": "1.327", + "latency": "7.928", + "cost_per_test": "0.000975", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.327, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.374, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.374", + "raw_stderr": "0.382", + "latency": "16.888", + "cost_per_test": "0.001758", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.382, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.95, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.95", + "raw_stderr": "1.94", + "latency": "8.91", + "cost_per_test": "0.001047", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.94, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.838, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.838", + "raw_stderr": "0.874", + "latency": "27.838", + "cost_per_test": "0.002409", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.874, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:grok/grok-3-mini-fast-high-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.078, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.078", + "raw_stderr": "1.387", + "latency": "7.35", + "cost_per_test": "0.000978", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.387, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-3-mini-fast-low-reasoning/2d494f5d-17c5-449f-b824-df9f5a7cc64a.json b/data/vals_ai/xai/grok-3-mini-fast-low-reasoning/2d494f5d-17c5-449f-b824-df9f5a7cc64a.json new file mode 100644 index 0000000000000000000000000000000000000000..f575ce0de20133bc9afb7d4a96823ac4b3d268c3 --- /dev/null +++ b/data/vals_ai/xai/grok-3-mini-fast-low-reasoning/2d494f5d-17c5-449f-b824-df9f5a7cc64a.json @@ -0,0 +1,1101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/xai_grok-3-mini-fast-low-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-3-mini-fast-low-reasoning", + "id": "xai/grok-3-mini-fast-low-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-3-mini-fast-low-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.958, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.958", + "raw_stderr": "1.122", + "latency": "6.336", + "cost_per_test": "0.000951", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.122, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.537, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.537", + "raw_stderr": "1.287", + "latency": "7.615", + "cost_per_test": "0.000997", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.287, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.456, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.456", + "raw_stderr": "0.984", + "latency": "14.197", + "cost_per_test": "0.001312", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.984, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.902, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.902", + "raw_stderr": "1.815", + "latency": "7.254", + "cost_per_test": "0.001104", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.815, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.308, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.308", + "raw_stderr": "1.219", + "latency": "5.538", + "cost_per_test": "0.000863", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.219, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.93, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.93", + "raw_stderr": "1.443", + "latency": "13.897", + "cost_per_test": "0.001665", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.443, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.817, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.817", + "raw_stderr": "1.518", + "latency": "5.44", + "cost_per_test": "0.000791", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.518, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 69.029, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.029", + "raw_stderr": "2.369", + "latency": "5.387", + "cost_per_test": "0.001109", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.369, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 60.854, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "60.854", + "raw_stderr": "1.471", + "latency": "6.139", + "cost_per_test": "0.000978", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.471, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.6, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.6", + "raw_stderr": "0.794", + "latency": "7.892", + "cost_per_test": "0.001214", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.794, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.68, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.68", + "raw_stderr": "1.347", + "latency": "4.975", + "cost_per_test": "0.000712", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.347, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.011, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.011", + "raw_stderr": "0.39", + "latency": "7.426", + "cost_per_test": "0.001026", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.39, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.551, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.551", + "raw_stderr": "1.924", + "latency": "5.384", + "cost_per_test": "0.000764", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.924, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.451, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.451", + "raw_stderr": "0.95", + "latency": "8.814", + "cost_per_test": "0.001151", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.95, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.078, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.078", + "raw_stderr": "1.387", + "latency": "5.093", + "cost_per_test": "0.000756", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.387, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-3-mini-fast-low-reasoning/422918d3-a523-4f7d-80e2-9abf4aab7064.json b/data/vals_ai/xai/grok-3-mini-fast-low-reasoning/422918d3-a523-4f7d-80e2-9abf4aab7064.json new file mode 100644 index 0000000000000000000000000000000000000000..ceb73e206ebfeda5749fd03017ebda69e5da8440 --- /dev/null +++ b/data/vals_ai/xai/grok-3-mini-fast-low-reasoning/422918d3-a523-4f7d-80e2-9abf4aab7064.json @@ -0,0 +1,308 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/xai_grok-3-mini-fast-low-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-3-mini-fast-low-reasoning", + "id": "xai/grok-3-mini-fast-low-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-3-mini-fast-low-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 67.016, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.016", + "raw_stderr": "1.605", + "latency": "5.678", + "cost_per_test": "0.000896", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.605, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 52.098, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.098", + "raw_stderr": "1.705", + "latency": "14.801", + "cost_per_test": "0.031819", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.705, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 59.479, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.479", + "raw_stderr": "0.96", + "latency": "10.516", + "cost_per_test": "0.01706", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.96, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 59.324, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.324", + "raw_stderr": "1.677", + "latency": "11.07", + "cost_per_test": "0.018464", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.677, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-3-mini-fast-low-reasoning/46310415-6f7b-455e-9082-c7767e190c08.json b/data/vals_ai/xai/grok-3-mini-fast-low-reasoning/46310415-6f7b-455e-9082-c7767e190c08.json new file mode 100644 index 0000000000000000000000000000000000000000..94d9eb3174c27d69c01fc3af4dfa7fd45bb4dd28 --- /dev/null +++ b/data/vals_ai/xai/grok-3-mini-fast-low-reasoning/46310415-6f7b-455e-9082-c7767e190c08.json @@ -0,0 +1,888 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mgsm/xai_grok-3-mini-fast-low-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MGSM", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "dataset_type": "public", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-3-mini-fast-low-reasoning", + "id": "xai/grok-3-mini-fast-low-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-3-mini-fast-low-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mgsm:mgsm_bn:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_bn", + "source_data": { + "dataset_name": "MGSM - Bengali", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_bn", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Bengali).", + "metric_id": "vals_ai.mgsm.mgsm_bn.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 88.4, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_bn", + "task_name": "Bengali", + "dataset_type": "public", + "industry": "math", + "raw_score": "88.4", + "raw_stderr": "2.025", + "latency": "6.591", + "cost_per_test": "0.000691", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.025, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_de:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_de", + "source_data": { + "dataset_name": "MGSM - German", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_de", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (German).", + "metric_id": "vals_ai.mgsm.mgsm_de.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 92.8, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_de", + "task_name": "German", + "dataset_type": "public", + "industry": "math", + "raw_score": "92.8", + "raw_stderr": "1.635", + "latency": "4.538", + "cost_per_test": "0.000449", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.635, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_en:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_en", + "source_data": { + "dataset_name": "MGSM - English", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_en", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (English).", + "metric_id": "vals_ai.mgsm.mgsm_en.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 96.0, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_en", + "task_name": "English", + "dataset_type": "public", + "industry": "math", + "raw_score": "96", + "raw_stderr": "1.239", + "latency": "3.399", + "cost_per_test": "0.000334", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.239, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_es:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_es", + "source_data": { + "dataset_name": "MGSM - Spanish", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_es", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Spanish).", + "metric_id": "vals_ai.mgsm.mgsm_es.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 90.8, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_es", + "task_name": "Spanish", + "dataset_type": "public", + "industry": "math", + "raw_score": "90.8", + "raw_stderr": "1.828", + "latency": "4.206", + "cost_per_test": "0.000417", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.828, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_fr:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_fr", + "source_data": { + "dataset_name": "MGSM - French", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_fr", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (French).", + "metric_id": "vals_ai.mgsm.mgsm_fr.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 85.6, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_fr", + "task_name": "French", + "dataset_type": "public", + "industry": "math", + "raw_score": "85.6", + "raw_stderr": "2.22", + "latency": "5.1", + "cost_per_test": "0.000443", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.22, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_ja:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_ja", + "source_data": { + "dataset_name": "MGSM - Japanese", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_ja", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Japanese).", + "metric_id": "vals_ai.mgsm.mgsm_ja.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 87.6, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_ja", + "task_name": "Japanese", + "dataset_type": "public", + "industry": "math", + "raw_score": "87.6", + "raw_stderr": "2.084", + "latency": "12.124", + "cost_per_test": "0.00048", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.084, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_ru:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_ru", + "source_data": { + "dataset_name": "MGSM - Russian", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_ru", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Russian).", + "metric_id": "vals_ai.mgsm.mgsm_ru.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 92.8, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_ru", + "task_name": "Russian", + "dataset_type": "public", + "industry": "math", + "raw_score": "92.8", + "raw_stderr": "1.635", + "latency": "4.466", + "cost_per_test": "0.000371", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.635, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_sw:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_sw", + "source_data": { + "dataset_name": "MGSM - Swahili", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_sw", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Swahili).", + "metric_id": "vals_ai.mgsm.mgsm_sw.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 90.8, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_sw", + "task_name": "Swahili", + "dataset_type": "public", + "industry": "math", + "raw_score": "90.8", + "raw_stderr": "1.828", + "latency": "5.377", + "cost_per_test": "0.000505", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.828, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_te:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_te", + "source_data": { + "dataset_name": "MGSM - Telugu", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_te", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Telugu).", + "metric_id": "vals_ai.mgsm.mgsm_te.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 86.8, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_te", + "task_name": "Telugu", + "dataset_type": "public", + "industry": "math", + "raw_score": "86.8", + "raw_stderr": "2.141", + "latency": "17.17", + "cost_per_test": "0.001032", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.141, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_th:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_th", + "source_data": { + "dataset_name": "MGSM - Thai", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_th", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Thai).", + "metric_id": "vals_ai.mgsm.mgsm_th.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 92.8, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_th", + "task_name": "Thai", + "dataset_type": "public", + "industry": "math", + "raw_score": "92.8", + "raw_stderr": "1.635", + "latency": "5.139", + "cost_per_test": "0.000516", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.635, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_zh:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_zh", + "source_data": { + "dataset_name": "MGSM - Chinese", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_zh", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Chinese).", + "metric_id": "vals_ai.mgsm.mgsm_zh.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 89.6, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_zh", + "task_name": "Chinese", + "dataset_type": "public", + "industry": "math", + "raw_score": "89.6", + "raw_stderr": "1.931", + "latency": "4.046", + "cost_per_test": "0.000356", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.931, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "mgsm:overall:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.mgsm.overall", + "source_data": { + "dataset_name": "MGSM - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Overall).", + "metric_id": "vals_ai.mgsm.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 90.364, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "math", + "raw_score": "90.364", + "raw_stderr": "1.836", + "latency": "6.56", + "cost_per_test": "0.000509", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.836, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-3-mini-fast-low-reasoning/48772310-61e5-475f-8db5-cbc653c3d37e.json b/data/vals_ai/xai/grok-3-mini-fast-low-reasoning/48772310-61e5-475f-8db5-cbc653c3d37e.json new file mode 100644 index 0000000000000000000000000000000000000000..55510b78914c82e58cccb0de6013eda6eeb863f0 --- /dev/null +++ b/data/vals_ai/xai/grok-3-mini-fast-low-reasoning/48772310-61e5-475f-8db5-cbc653c3d37e.json @@ -0,0 +1,240 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/xai_grok-3-mini-fast-low-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-3-mini-fast-low-reasoning", + "id": "xai/grok-3-mini-fast-low-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-3-mini-fast-low-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 61.979, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.979", + "raw_stderr": "1.388", + "latency": "11.728", + "cost_per_test": "0.00112", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.388, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 72.976, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "72.976", + "raw_stderr": "0.87", + "latency": "11.728", + "cost_per_test": "0.00112", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.87, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 83.974, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "83.974", + "raw_stderr": "1.049", + "latency": "11.728", + "cost_per_test": "0.00112", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.049, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-3-mini-fast-low-reasoning/523239ba-8c1d-496b-b9d9-18948dca6e19.json b/data/vals_ai/xai/grok-3-mini-fast-low-reasoning/523239ba-8c1d-496b-b9d9-18948dca6e19.json new file mode 100644 index 0000000000000000000000000000000000000000..c581b2501bdf0bebf8dd496b4249ac68d15cfb32 --- /dev/null +++ b/data/vals_ai/xai/grok-3-mini-fast-low-reasoning/523239ba-8c1d-496b-b9d9-18948dca6e19.json @@ -0,0 +1,107 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/math500/xai_grok-3-mini-fast-low-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MATH 500", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "math500", + "benchmark_name": "MATH 500", + "benchmark_updated": "2026-01-09", + "dataset_type": "public", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/math500", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-3-mini-fast-low-reasoning", + "id": "xai/grok-3-mini-fast-low-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-3-mini-fast-low-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "math500:overall:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.math500.overall", + "source_data": { + "dataset_name": "MATH 500 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/math500" + ], + "additional_details": { + "benchmark_slug": "math500", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/math500" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MATH 500 (Overall).", + "metric_id": "vals_ai.math500.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/math500" + } + }, + "score_details": { + "score": 70.2, + "details": { + "benchmark_slug": "math500", + "benchmark_name": "MATH 500", + "benchmark_updated": "2026-01-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "math", + "raw_score": "70.2", + "raw_stderr": "2.046", + "latency": "9.885", + "cost_per_test": "0.000893", + "temperature": "0", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.046, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-3-mini-fast-low-reasoning/6ab9c627-5170-428c-8028-cf4fad0c08d6.json b/data/vals_ai/xai/grok-3-mini-fast-low-reasoning/6ab9c627-5170-428c-8028-cf4fad0c08d6.json new file mode 100644 index 0000000000000000000000000000000000000000..d58629a737d7218de03787919a2b104d5b733b82 --- /dev/null +++ b/data/vals_ai/xai/grok-3-mini-fast-low-reasoning/6ab9c627-5170-428c-8028-cf4fad0c08d6.json @@ -0,0 +1,456 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/xai_grok-3-mini-fast-low-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-3-mini-fast-low-reasoning", + "id": "xai/grok-3-mini-fast-low-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-3-mini-fast-low-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 86.21, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.21", + "raw_stderr": "0.921", + "latency": "3.246", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.921, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.626, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.626", + "raw_stderr": "0.194", + "latency": "3.396", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.194, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.28, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.28", + "raw_stderr": "0.542", + "latency": "3.289", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.542, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.925, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.925", + "raw_stderr": "0.434", + "latency": "3.44", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.434, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.366, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.366", + "raw_stderr": "0.861", + "latency": "3.278", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.861, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 74.145, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "74.145", + "raw_stderr": "1.671", + "latency": "3.991", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.671, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-3-mini-fast-low-reasoning/a33055b2-cc3b-44d3-87f0-cc8fda0fc2c8.json b/data/vals_ai/xai/grok-3-mini-fast-low-reasoning/a33055b2-cc3b-44d3-87f0-cc8fda0fc2c8.json new file mode 100644 index 0000000000000000000000000000000000000000..8b5a5ad195504ad30cf315c6e35e14c61b472493 --- /dev/null +++ b/data/vals_ai/xai/grok-3-mini-fast-low-reasoning/a33055b2-cc3b-44d3-87f0-cc8fda0fc2c8.json @@ -0,0 +1,533 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medqa/xai_grok-3-mini-fast-low-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedQA", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "dataset_type": "public", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-3-mini-fast-low-reasoning", + "id": "xai/grok-3-mini-fast-low-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-3-mini-fast-low-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medqa:asian:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.medqa.asian", + "source_data": { + "dataset_name": "MedQA - Asian", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "asian", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Asian).", + "metric_id": "vals_ai.medqa.asian.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 88.25, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "asian", + "task_name": "Asian", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "88.25", + "raw_stderr": "0.72", + "latency": "4.855", + "cost_per_test": "0.000558", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.72, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "medqa:black:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.medqa.black", + "source_data": { + "dataset_name": "MedQA - Black", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "black", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Black).", + "metric_id": "vals_ai.medqa.black.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 89.05, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "black", + "task_name": "Black", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "89.05", + "raw_stderr": "0.698", + "latency": "4.894", + "cost_per_test": "0.000564", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.698, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "medqa:hispanic:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.medqa.hispanic", + "source_data": { + "dataset_name": "MedQA - Hispanic", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "hispanic", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Hispanic).", + "metric_id": "vals_ai.medqa.hispanic.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 88.3, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "hispanic", + "task_name": "Hispanic", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "88.3", + "raw_stderr": "0.719", + "latency": "4.956", + "cost_per_test": "0.000557", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.719, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "medqa:indigenous:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.medqa.indigenous", + "source_data": { + "dataset_name": "MedQA - Indigenous", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "indigenous", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Indigenous).", + "metric_id": "vals_ai.medqa.indigenous.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 87.7, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "indigenous", + "task_name": "Indigenous", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "87.7", + "raw_stderr": "0.734", + "latency": "5.04", + "cost_per_test": "0.000564", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.734, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "medqa:overall:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.medqa.overall", + "source_data": { + "dataset_name": "MedQA - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Overall).", + "metric_id": "vals_ai.medqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 88.65, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "88.65", + "raw_stderr": "0.29", + "latency": "4.883", + "cost_per_test": "0.000558", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.29, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "medqa:unbiased:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.medqa.unbiased", + "source_data": { + "dataset_name": "MedQA - Unbiased", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "unbiased", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Unbiased).", + "metric_id": "vals_ai.medqa.unbiased.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 89.4, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "unbiased", + "task_name": "Unbiased", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "89.4", + "raw_stderr": "0.688", + "latency": "4.757", + "cost_per_test": "0.000549", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.688, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "medqa:white:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.medqa.white", + "source_data": { + "dataset_name": "MedQA - White", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "white", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (White).", + "metric_id": "vals_ai.medqa.white.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 89.2, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "white", + "task_name": "White", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "89.2", + "raw_stderr": "0.694", + "latency": "4.796", + "cost_per_test": "0.000558", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.694, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-3-mini-fast-low-reasoning/d3536a14-e924-4f64-ac6b-234196d55d15.json b/data/vals_ai/xai/grok-3-mini-fast-low-reasoning/d3536a14-e924-4f64-ac6b-234196d55d15.json new file mode 100644 index 0000000000000000000000000000000000000000..c04f0ee347a1ef67ede49b1913ac22f1e03af96b --- /dev/null +++ b/data/vals_ai/xai/grok-3-mini-fast-low-reasoning/d3536a14-e924-4f64-ac6b-234196d55d15.json @@ -0,0 +1,249 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/xai_grok-3-mini-fast-low-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-3-mini-fast-low-reasoning", + "id": "xai/grok-3-mini-fast-low-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-3-mini-fast-low-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 69.697, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.697", + "raw_stderr": "3.266", + "latency": "14.615", + "cost_per_test": "0.001837", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.266, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 76.263, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.263", + "raw_stderr": "3.046", + "latency": "13.854", + "cost_per_test": "0.001361", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.046, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 72.98, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.98", + "raw_stderr": "2.233", + "latency": "14.234", + "cost_per_test": "0.001599", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.233, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-3-mini-fast-low-reasoning/f086d7b9-9692-4f9c-90e1-e510e1936b24.json b/data/vals_ai/xai/grok-3-mini-fast-low-reasoning/f086d7b9-9692-4f9c-90e1-e510e1936b24.json new file mode 100644 index 0000000000000000000000000000000000000000..a6665433bf2b15d16c3494dc5ddd3f1559a8f290 --- /dev/null +++ b/data/vals_ai/xai/grok-3-mini-fast-low-reasoning/f086d7b9-9692-4f9c-90e1-e510e1936b24.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/aime/xai_grok-3-mini-fast-low-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - AIME", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "dataset_type": "public", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-3-mini-fast-low-reasoning", + "id": "xai/grok-3-mini-fast-low-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-3-mini-fast-low-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aime:aime_2024:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.aime.aime_2024", + "source_data": { + "dataset_name": "AIME - AIME 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/aime" + ], + "additional_details": { + "benchmark_slug": "aime", + "task_key": "aime_2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for AIME (AIME 2024).", + "metric_id": "vals_ai.aime.aime_2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "score_details": { + "score": 73.75, + "details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "task_key": "aime_2024", + "task_name": "AIME 2024", + "dataset_type": "public", + "industry": "math", + "raw_score": "73.75", + "raw_stderr": "0.755", + "latency": "30.443", + "cost_per_test": "0.002841", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.755, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "aime:aime_2025:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.aime.aime_2025", + "source_data": { + "dataset_name": "AIME - AIME 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/aime" + ], + "additional_details": { + "benchmark_slug": "aime", + "task_key": "aime_2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for AIME (AIME 2025).", + "metric_id": "vals_ai.aime.aime_2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "score_details": { + "score": 67.5, + "details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "task_key": "aime_2025", + "task_name": "AIME 2025", + "dataset_type": "public", + "industry": "math", + "raw_score": "67.5", + "raw_stderr": "0.833", + "latency": "32.352", + "cost_per_test": "0.00302", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.833, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "aime:overall:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.aime.overall", + "source_data": { + "dataset_name": "AIME - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/aime" + ], + "additional_details": { + "benchmark_slug": "aime", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for AIME (Overall).", + "metric_id": "vals_ai.aime.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "score_details": { + "score": 70.625, + "details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "math", + "raw_score": "70.625", + "raw_stderr": "0.562", + "latency": "31.397", + "cost_per_test": "0.00293", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.562, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-3-mini-fast-low-reasoning/f2b965af-ba01-4af3-a4fc-96c8c2032645.json b/data/vals_ai/xai/grok-3-mini-fast-low-reasoning/f2b965af-ba01-4af3-a4fc-96c8c2032645.json new file mode 100644 index 0000000000000000000000000000000000000000..2395d1a197854536797615e9d7153400bfd42096 --- /dev/null +++ b/data/vals_ai/xai/grok-3-mini-fast-low-reasoning/f2b965af-ba01-4af3-a4fc-96c8c2032645.json @@ -0,0 +1,320 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/xai_grok-3-mini-fast-low-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-3-mini-fast-low-reasoning", + "id": "xai/grok-3-mini-fast-low-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-3-mini-fast-low-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.137, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.137", + "raw_stderr": "0.754", + "latency": "9.769", + "cost_per_test": "0.00096", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.754, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "lcb:hard:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 28.857, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "28.857", + "raw_stderr": "2.422", + "latency": "56.247", + "cost_per_test": "0.003407", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.422, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "lcb:medium:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 71.802, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "71.802", + "raw_stderr": "2.299", + "latency": "26.106", + "cost_per_test": "0.002094", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.299, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + }, + { + "evaluation_result_id": "lcb:overall:grok/grok-3-mini-fast-low-reasoning:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 66.265, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66.265", + "raw_stderr": "1.141", + "latency": "30.707", + "cost_per_test": "0.002154", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "reasoning_effort": "low", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.141, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "low" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-3/09f27983-8eda-4b14-bdc0-6b50d742faf6.json b/data/vals_ai/xai/grok-3/09f27983-8eda-4b14-bdc0-6b50d742faf6.json new file mode 100644 index 0000000000000000000000000000000000000000..c7cc733c0e3d774162ea3fe6d49abf9aeac02ac5 --- /dev/null +++ b/data/vals_ai/xai/grok-3/09f27983-8eda-4b14-bdc0-6b50d742faf6.json @@ -0,0 +1,432 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/xai_grok-3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-3", + "id": "xai/grok-3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:grok/grok-3:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 86.085, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.085", + "raw_stderr": "0.944", + "latency": "0.379", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.944, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:grok/grok-3:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.865, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.865", + "raw_stderr": "0.195", + "latency": "0.404", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.195, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:grok/grok-3:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.812, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.812", + "raw_stderr": "0.566", + "latency": "0.386", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.566, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:grok/grok-3:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.595, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.595", + "raw_stderr": "0.394", + "latency": "0.741", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.394, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:grok/grok-3:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.552, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.552", + "raw_stderr": "0.865", + "latency": "0.408", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.865, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:grok/grok-3:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 76.659, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "76.659", + "raw_stderr": "1.371", + "latency": "2.127", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.371, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-3/1576126d-52a4-4c3b-aa06-b766faa5bcbc.json b/data/vals_ai/xai/grok-3/1576126d-52a4-4c3b-aa06-b766faa5bcbc.json new file mode 100644 index 0000000000000000000000000000000000000000..6e0c121b29eba6822c3dfc4b14e4dd33f5569393 --- /dev/null +++ b/data/vals_ai/xai/grok-3/1576126d-52a4-4c3b-aa06-b766faa5bcbc.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/xai_grok-3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-3", + "id": "xai/grok-3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:grok/grok-3:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 58.508, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.508", + "raw_stderr": "1.682", + "latency": "5.488", + "cost_per_test": "0.007219", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "16384", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.682, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:grok/grok-3:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 57.11, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.11", + "raw_stderr": "1.69", + "latency": "49.7", + "cost_per_test": "0.317909", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "16384", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.69, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:grok/grok-3:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 59.713, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.713", + "raw_stderr": "0.965", + "latency": "24.354", + "cost_per_test": "0.169809", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "16384", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.965, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:grok/grok-3:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 63.52, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.52", + "raw_stderr": "1.643", + "latency": "17.873", + "cost_per_test": "0.184299", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "16384", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.643, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-3/2446198f-2d47-46c2-bd39-084f9d94c0c3.json b/data/vals_ai/xai/grok-3/2446198f-2d47-46c2-bd39-084f9d94c0c3.json new file mode 100644 index 0000000000000000000000000000000000000000..8b31cc836ad76f5e4c955f16df72681a0d77788c --- /dev/null +++ b/data/vals_ai/xai/grok-3/2446198f-2d47-46c2-bd39-084f9d94c0c3.json @@ -0,0 +1,505 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medqa/xai_grok-3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedQA", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "dataset_type": "public", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-3", + "id": "xai/grok-3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medqa:asian:grok/grok-3:score", + "evaluation_name": "vals_ai.medqa.asian", + "source_data": { + "dataset_name": "MedQA - Asian", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "asian", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Asian).", + "metric_id": "vals_ai.medqa.asian.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 83.7, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "asian", + "task_name": "Asian", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "83.7", + "raw_stderr": "0.826", + "latency": "8.814", + "cost_per_test": "0.006281", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.826, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:black:grok/grok-3:score", + "evaluation_name": "vals_ai.medqa.black", + "source_data": { + "dataset_name": "MedQA - Black", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "black", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Black).", + "metric_id": "vals_ai.medqa.black.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 84.3, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "black", + "task_name": "Black", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "84.3", + "raw_stderr": "0.814", + "latency": "8.077", + "cost_per_test": "0.006245", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.814, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:hispanic:grok/grok-3:score", + "evaluation_name": "vals_ai.medqa.hispanic", + "source_data": { + "dataset_name": "MedQA - Hispanic", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "hispanic", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Hispanic).", + "metric_id": "vals_ai.medqa.hispanic.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 83.2, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "hispanic", + "task_name": "Hispanic", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "83.2", + "raw_stderr": "0.836", + "latency": "7.415", + "cost_per_test": "0.006224", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.836, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:indigenous:grok/grok-3:score", + "evaluation_name": "vals_ai.medqa.indigenous", + "source_data": { + "dataset_name": "MedQA - Indigenous", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "indigenous", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Indigenous).", + "metric_id": "vals_ai.medqa.indigenous.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 82.75, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "indigenous", + "task_name": "Indigenous", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "82.75", + "raw_stderr": "0.845", + "latency": "10.571", + "cost_per_test": "0.006417", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.845, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:overall:grok/grok-3:score", + "evaluation_name": "vals_ai.medqa.overall", + "source_data": { + "dataset_name": "MedQA - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Overall).", + "metric_id": "vals_ai.medqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 83.85, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "83.85", + "raw_stderr": "0.336", + "latency": "7.536", + "cost_per_test": "0.006196", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.336, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:unbiased:grok/grok-3:score", + "evaluation_name": "vals_ai.medqa.unbiased", + "source_data": { + "dataset_name": "MedQA - Unbiased", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "unbiased", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Unbiased).", + "metric_id": "vals_ai.medqa.unbiased.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 85.35, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "unbiased", + "task_name": "Unbiased", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "85.35", + "raw_stderr": "0.791", + "latency": "4.502", + "cost_per_test": "0.005875", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.791, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:white:grok/grok-3:score", + "evaluation_name": "vals_ai.medqa.white", + "source_data": { + "dataset_name": "MedQA - White", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "white", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (White).", + "metric_id": "vals_ai.medqa.white.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 83.8, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "white", + "task_name": "White", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "83.8", + "raw_stderr": "0.824", + "latency": "5.838", + "cost_per_test": "0.006131", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.824, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-3/24733b36-7870-4e58-b82a-fa16fb9b1e88.json b/data/vals_ai/xai/grok-3/24733b36-7870-4e58-b82a-fa16fb9b1e88.json new file mode 100644 index 0000000000000000000000000000000000000000..ce2cd51e9d93ec8850480c0feb3f1c3c4d6879f4 --- /dev/null +++ b/data/vals_ai/xai/grok-3/24733b36-7870-4e58-b82a-fa16fb9b1e88.json @@ -0,0 +1,840 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mgsm/xai_grok-3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MGSM", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "dataset_type": "public", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-3", + "id": "xai/grok-3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mgsm:mgsm_bn:grok/grok-3:score", + "evaluation_name": "vals_ai.mgsm.mgsm_bn", + "source_data": { + "dataset_name": "MGSM - Bengali", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_bn", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Bengali).", + "metric_id": "vals_ai.mgsm.mgsm_bn.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 82.8, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_bn", + "task_name": "Bengali", + "dataset_type": "public", + "industry": "math", + "raw_score": "82.8", + "raw_stderr": "2.387", + "latency": "8.702", + "cost_per_test": "0.009055", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.387, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_de:grok/grok-3:score", + "evaluation_name": "vals_ai.mgsm.mgsm_de", + "source_data": { + "dataset_name": "MGSM - German", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_de", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (German).", + "metric_id": "vals_ai.mgsm.mgsm_de.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 92.0, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_de", + "task_name": "German", + "dataset_type": "public", + "industry": "math", + "raw_score": "92", + "raw_stderr": "1.716", + "latency": "5.091", + "cost_per_test": "0.004994", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.716, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_en:grok/grok-3:score", + "evaluation_name": "vals_ai.mgsm.mgsm_en", + "source_data": { + "dataset_name": "MGSM - English", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_en", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (English).", + "metric_id": "vals_ai.mgsm.mgsm_en.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 97.2, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_en", + "task_name": "English", + "dataset_type": "public", + "industry": "math", + "raw_score": "97.2", + "raw_stderr": "1.043", + "latency": "6.223", + "cost_per_test": "0.005651", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.043, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_es:grok/grok-3:score", + "evaluation_name": "vals_ai.mgsm.mgsm_es", + "source_data": { + "dataset_name": "MGSM - Spanish", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_es", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Spanish).", + "metric_id": "vals_ai.mgsm.mgsm_es.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 94.8, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_es", + "task_name": "Spanish", + "dataset_type": "public", + "industry": "math", + "raw_score": "94.8", + "raw_stderr": "1.404", + "latency": "3.67", + "cost_per_test": "0.004265", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.404, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_fr:grok/grok-3:score", + "evaluation_name": "vals_ai.mgsm.mgsm_fr", + "source_data": { + "dataset_name": "MGSM - French", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_fr", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (French).", + "metric_id": "vals_ai.mgsm.mgsm_fr.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 90.8, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_fr", + "task_name": "French", + "dataset_type": "public", + "industry": "math", + "raw_score": "90.8", + "raw_stderr": "1.828", + "latency": "4.564", + "cost_per_test": "0.00425", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.828, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_ja:grok/grok-3:score", + "evaluation_name": "vals_ai.mgsm.mgsm_ja", + "source_data": { + "dataset_name": "MGSM - Japanese", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_ja", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Japanese).", + "metric_id": "vals_ai.mgsm.mgsm_ja.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 86.4, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_ja", + "task_name": "Japanese", + "dataset_type": "public", + "industry": "math", + "raw_score": "86.4", + "raw_stderr": "2.168", + "latency": "5.354", + "cost_per_test": "0.005902", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.168, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_ru:grok/grok-3:score", + "evaluation_name": "vals_ai.mgsm.mgsm_ru", + "source_data": { + "dataset_name": "MGSM - Russian", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_ru", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Russian).", + "metric_id": "vals_ai.mgsm.mgsm_ru.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 95.6, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_ru", + "task_name": "Russian", + "dataset_type": "public", + "industry": "math", + "raw_score": "95.6", + "raw_stderr": "1.297", + "latency": "4.392", + "cost_per_test": "0.00478", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.297, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_sw:grok/grok-3:score", + "evaluation_name": "vals_ai.mgsm.mgsm_sw", + "source_data": { + "dataset_name": "MGSM - Swahili", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_sw", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Swahili).", + "metric_id": "vals_ai.mgsm.mgsm_sw.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 91.6, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_sw", + "task_name": "Swahili", + "dataset_type": "public", + "industry": "math", + "raw_score": "91.6", + "raw_stderr": "1.754", + "latency": "5.026", + "cost_per_test": "0.004118", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.754, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_te:grok/grok-3:score", + "evaluation_name": "vals_ai.mgsm.mgsm_te", + "source_data": { + "dataset_name": "MGSM - Telugu", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_te", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Telugu).", + "metric_id": "vals_ai.mgsm.mgsm_te.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 88.0, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_te", + "task_name": "Telugu", + "dataset_type": "public", + "industry": "math", + "raw_score": "88", + "raw_stderr": "2.055", + "latency": "11.171", + "cost_per_test": "0.01059", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.055, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_th:grok/grok-3:score", + "evaluation_name": "vals_ai.mgsm.mgsm_th", + "source_data": { + "dataset_name": "MGSM - Thai", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_th", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Thai).", + "metric_id": "vals_ai.mgsm.mgsm_th.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 93.2, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_th", + "task_name": "Thai", + "dataset_type": "public", + "industry": "math", + "raw_score": "93.2", + "raw_stderr": "1.592", + "latency": "5.966", + "cost_per_test": "0.00484", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.592, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_zh:grok/grok-3:score", + "evaluation_name": "vals_ai.mgsm.mgsm_zh", + "source_data": { + "dataset_name": "MGSM - Chinese", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_zh", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Chinese).", + "metric_id": "vals_ai.mgsm.mgsm_zh.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 92.4, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_zh", + "task_name": "Chinese", + "dataset_type": "public", + "industry": "math", + "raw_score": "92.4", + "raw_stderr": "1.676", + "latency": "3.98", + "cost_per_test": "0.003922", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.676, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:overall:grok/grok-3:score", + "evaluation_name": "vals_ai.mgsm.overall", + "source_data": { + "dataset_name": "MGSM - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Overall).", + "metric_id": "vals_ai.mgsm.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 91.346, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "math", + "raw_score": "91.346", + "raw_stderr": "1.72", + "latency": "5.831", + "cost_per_test": "0.00567", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.72, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-3/6124423c-86ad-4ab4-bc07-23260d90f65d.json b/data/vals_ai/xai/grok-3/6124423c-86ad-4ab4-bc07-23260d90f65d.json new file mode 100644 index 0000000000000000000000000000000000000000..a81cd34d202e2ed774f1d4f68fca944a3b8c6b04 --- /dev/null +++ b/data/vals_ai/xai/grok-3/6124423c-86ad-4ab4-bc07-23260d90f65d.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/xai_grok-3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-3", + "id": "xai/grok-3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:grok/grok-3:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 94.099, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "94.099", + "raw_stderr": "1.313", + "latency": "2.791", + "cost_per_test": "0.005439", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.313, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:grok/grok-3:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 13.429, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "13.429", + "raw_stderr": "1.823", + "latency": "8.513", + "cost_per_test": "0.013436", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.823, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:grok/grok-3:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 51.175, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "51.175", + "raw_stderr": "2.554", + "latency": "5.247", + "cost_per_test": "0.009192", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.554, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:grok/grok-3:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 52.901, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "52.901", + "raw_stderr": "1.134", + "latency": "5.517", + "cost_per_test": "0.009356", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.134, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-3/7fbcc2f5-1de8-46ca-afae-8be5faec85bc.json b/data/vals_ai/xai/grok-3/7fbcc2f5-1de8-46ca-afae-8be5faec85bc.json new file mode 100644 index 0000000000000000000000000000000000000000..946488a66eeec802353335fb8f919bf7307b0abe --- /dev/null +++ b/data/vals_ai/xai/grok-3/7fbcc2f5-1de8-46ca-afae-8be5faec85bc.json @@ -0,0 +1,103 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/math500/xai_grok-3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MATH 500", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "math500", + "benchmark_name": "MATH 500", + "benchmark_updated": "2026-01-09", + "dataset_type": "public", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/math500", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-3", + "id": "xai/grok-3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "math500:overall:grok/grok-3:score", + "evaluation_name": "vals_ai.math500.overall", + "source_data": { + "dataset_name": "MATH 500 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/math500" + ], + "additional_details": { + "benchmark_slug": "math500", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/math500" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MATH 500 (Overall).", + "metric_id": "vals_ai.math500.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/math500" + } + }, + "score_details": { + "score": 89.8, + "details": { + "benchmark_slug": "math500", + "benchmark_name": "MATH 500", + "benchmark_updated": "2026-01-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "math", + "raw_score": "89.8", + "raw_stderr": "1.354", + "latency": "9.085", + "cost_per_test": "0.012781", + "temperature": "0", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.354, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.0, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-3/91eeba20-55be-4552-8583-a1e2f89c6e05.json b/data/vals_ai/xai/grok-3/91eeba20-55be-4552-8583-a1e2f89c6e05.json new file mode 100644 index 0000000000000000000000000000000000000000..e7b337df8972bf71e35483caa47c6d563fba32e4 --- /dev/null +++ b/data/vals_ai/xai/grok-3/91eeba20-55be-4552-8583-a1e2f89c6e05.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/xai_grok-3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-3", + "id": "xai/grok-3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:grok/grok-3:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 71.717, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.717", + "raw_stderr": "3.201", + "latency": "25.907", + "cost_per_test": "0.0327", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.201, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:grok/grok-3:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 76.768, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.768", + "raw_stderr": "3.046", + "latency": "33.917", + "cost_per_test": "0.045785", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.046, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:grok/grok-3:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 74.242, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.242", + "raw_stderr": "2.209", + "latency": "29.912", + "cost_per_test": "0.039242", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.209, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-3/b8a853be-0b1d-48e4-ac02-ac620a89068d.json b/data/vals_ai/xai/grok-3/b8a853be-0b1d-48e4-ac02-ac620a89068d.json new file mode 100644 index 0000000000000000000000000000000000000000..96557df50babd32253924834008553394fff4fb2 --- /dev/null +++ b/data/vals_ai/xai/grok-3/b8a853be-0b1d-48e4-ac02-ac620a89068d.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/xai_grok-3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-3", + "id": "xai/grok-3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:grok/grok-3:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.145, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.145", + "raw_stderr": "1.207", + "latency": "14.358", + "cost_per_test": "0.015289", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.207, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:grok/grok-3:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.791, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.791", + "raw_stderr": "1.278", + "latency": "8.804", + "cost_per_test": "0.0131", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.278, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:grok/grok-3:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.749, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.749", + "raw_stderr": "1.008", + "latency": "18.732", + "cost_per_test": "0.020273", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.008, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:grok/grok-3:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.146, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.146", + "raw_stderr": "1.804", + "latency": "10.499", + "cost_per_test": "0.016725", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.804, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:grok/grok-3:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.834, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.834", + "raw_stderr": "1.235", + "latency": "8.42", + "cost_per_test": "0.013966", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.235, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:grok/grok-3:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 69.143, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.143", + "raw_stderr": "1.484", + "latency": "16.967", + "cost_per_test": "0.023894", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.484, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:grok/grok-3:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.362, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.362", + "raw_stderr": "1.44", + "latency": "8.855", + "cost_per_test": "0.013128", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.44, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:grok/grok-3:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 70.604, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.604", + "raw_stderr": "2.334", + "latency": "11.052", + "cost_per_test": "0.017408", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.334, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:grok/grok-3:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 60.218, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "60.218", + "raw_stderr": "1.475", + "latency": "11.861", + "cost_per_test": "0.019033", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.475, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:grok/grok-3:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.192, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.192", + "raw_stderr": "0.771", + "latency": "14.214", + "cost_per_test": "0.022041", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.771, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:grok/grok-3:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.221, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.221", + "raw_stderr": "1.335", + "latency": "8.568", + "cost_per_test": "0.011647", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.335, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:grok/grok-3:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.949, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.949", + "raw_stderr": "0.39", + "latency": "12.096", + "cost_per_test": "0.016622", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.39, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:grok/grok-3:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.148, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.148", + "raw_stderr": "1.96", + "latency": "13.985", + "cost_per_test": "0.013679", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.96, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:grok/grok-3:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.914, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.914", + "raw_stderr": "0.904", + "latency": "11.498", + "cost_per_test": "0.01956", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.904, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:grok/grok-3:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.825, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.825", + "raw_stderr": "1.421", + "latency": "11.534", + "cost_per_test": "0.012964", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.421, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-3/d20c288c-cb9d-4c8e-be0a-b7a9ffeb266f.json b/data/vals_ai/xai/grok-3/d20c288c-cb9d-4c8e-be0a-b7a9ffeb266f.json new file mode 100644 index 0000000000000000000000000000000000000000..997ef69c3d35ae34d103d15a2cb379f94234059d --- /dev/null +++ b/data/vals_ai/xai/grok-3/d20c288c-cb9d-4c8e-be0a-b7a9ffeb266f.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/aime/xai_grok-3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - AIME", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "dataset_type": "public", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-3", + "id": "xai/grok-3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aime:aime_2024:grok/grok-3:score", + "evaluation_name": "vals_ai.aime.aime_2024", + "source_data": { + "dataset_name": "AIME - AIME 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/aime" + ], + "additional_details": { + "benchmark_slug": "aime", + "task_key": "aime_2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for AIME (AIME 2024).", + "metric_id": "vals_ai.aime.aime_2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "score_details": { + "score": 60.417, + "details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "task_key": "aime_2024", + "task_name": "AIME 2024", + "dataset_type": "public", + "industry": "math", + "raw_score": "60.417", + "raw_stderr": "2.394", + "latency": "64.248", + "cost_per_test": "0.092816", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.394, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "aime:aime_2025:grok/grok-3:score", + "evaluation_name": "vals_ai.aime.aime_2025", + "source_data": { + "dataset_name": "AIME - AIME 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/aime" + ], + "additional_details": { + "benchmark_slug": "aime", + "task_key": "aime_2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for AIME (AIME 2025).", + "metric_id": "vals_ai.aime.aime_2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "score_details": { + "score": 57.083, + "details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "task_key": "aime_2025", + "task_name": "AIME 2025", + "dataset_type": "public", + "industry": "math", + "raw_score": "57.083", + "raw_stderr": "1.598", + "latency": "63.73", + "cost_per_test": "0.088521", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.598, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "aime:overall:grok/grok-3:score", + "evaluation_name": "vals_ai.aime.overall", + "source_data": { + "dataset_name": "AIME - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/aime" + ], + "additional_details": { + "benchmark_slug": "aime", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for AIME (Overall).", + "metric_id": "vals_ai.aime.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "score_details": { + "score": 58.75, + "details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "math", + "raw_score": "58.75", + "raw_stderr": "1.439", + "latency": "63.989", + "cost_per_test": "0.090668", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.439, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-3/f5fe46a9-f70b-4b77-8de8-be2e0df67a00.json b/data/vals_ai/xai/grok-3/f5fe46a9-f70b-4b77-8de8-be2e0df67a00.json new file mode 100644 index 0000000000000000000000000000000000000000..18a68322306f8ee152510102c84b67d5308635f9 --- /dev/null +++ b/data/vals_ai/xai/grok-3/f5fe46a9-f70b-4b77-8de8-be2e0df67a00.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/xai_grok-3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-3", + "id": "xai/grok-3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:grok/grok-3:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 63.859, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.859", + "raw_stderr": "1.374", + "latency": "13.058", + "cost_per_test": "0.017475", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.374, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:grok/grok-3:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 75.879, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "75.879", + "raw_stderr": "0.83", + "latency": "13.058", + "cost_per_test": "0.017475", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.83, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:grok/grok-3:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 87.899, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "87.899", + "raw_stderr": "0.933", + "latency": "13.058", + "cost_per_test": "0.017475", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "131000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.933, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 131000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-0709/1eb9dec8-a15f-4ed6-9caf-b0706581dd78.json b/data/vals_ai/xai/grok-4-0709/1eb9dec8-a15f-4ed6-9caf-b0706581dd78.json new file mode 100644 index 0000000000000000000000000000000000000000..9bfd81b4b3ec22e2f942dda2c6d0ec7a5521e6f3 --- /dev/null +++ b/data/vals_ai/xai/grok-4-0709/1eb9dec8-a15f-4ed6-9caf-b0706581dd78.json @@ -0,0 +1,103 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/math500/xai_grok-4-0709/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MATH 500", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "math500", + "benchmark_name": "MATH 500", + "benchmark_updated": "2026-01-09", + "dataset_type": "public", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/math500", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-0709", + "id": "xai/grok-4-0709", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-0709", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "math500:overall:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.math500.overall", + "source_data": { + "dataset_name": "MATH 500 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/math500" + ], + "additional_details": { + "benchmark_slug": "math500", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/math500" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MATH 500 (Overall).", + "metric_id": "vals_ai.math500.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/math500" + } + }, + "score_details": { + "score": 96.2, + "details": { + "benchmark_slug": "math500", + "benchmark_name": "MATH 500", + "benchmark_updated": "2026-01-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "math", + "raw_score": "96.2", + "raw_stderr": "0.855", + "latency": "27.261", + "cost_per_test": "0.037425", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.855, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-0709/23b28439-7652-452e-b947-fb04e2931474.json b/data/vals_ai/xai/grok-4-0709/23b28439-7652-452e-b947-fb04e2931474.json new file mode 100644 index 0000000000000000000000000000000000000000..86c0e96dd836d453b90769c7e574c544017fa023 --- /dev/null +++ b/data/vals_ai/xai/grok-4-0709/23b28439-7652-452e-b947-fb04e2931474.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/case_law_v2/xai_grok-4-0709/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CaseLaw v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "case_law_v2", + "benchmark_name": "CaseLaw v2", + "benchmark_updated": "2026-05-04", + "dataset_type": "private", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/case_law_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-0709", + "id": "xai/grok-4-0709", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-0709", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "case_law_v2:overall:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.case_law_v2.overall", + "source_data": { + "dataset_name": "CaseLaw v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "case_law_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/case_law_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CaseLaw v2 (Overall).", + "metric_id": "vals_ai.case_law_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/case_law_v2" + } + }, + "score_details": { + "score": 65.809, + "details": { + "benchmark_slug": "case_law_v2", + "benchmark_name": "CaseLaw v2", + "benchmark_updated": "2026-05-04", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "legal", + "raw_score": "65.809", + "raw_stderr": "0.214", + "latency": "32.335", + "cost_per_test": "0.158935", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.214, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-0709/27f0870a-a299-4b45-bbfa-3a4755c57b38.json b/data/vals_ai/xai/grok-4-0709/27f0870a-a299-4b45-bbfa-3a4755c57b38.json new file mode 100644 index 0000000000000000000000000000000000000000..5f8500e97fb7dc2a0d2bfd3e162d68833990899a --- /dev/null +++ b/data/vals_ai/xai/grok-4-0709/27f0870a-a299-4b45-bbfa-3a4755c57b38.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/xai_grok-4-0709/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-0709", + "id": "xai/grok-4-0709", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-0709", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 62.061, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.061", + "raw_stderr": "1.388", + "latency": "159.421", + "cost_per_test": "0.077989", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.388, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 65.086, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.086", + "raw_stderr": "0.962", + "latency": "159.421", + "cost_per_test": "0.077989", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.962, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 68.111, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.111", + "raw_stderr": "1.333", + "latency": "159.421", + "cost_per_test": "0.077989", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-0709/321356f3-59bd-44f1-9ee0-535e2923f6d4.json b/data/vals_ai/xai/grok-4-0709/321356f3-59bd-44f1-9ee0-535e2923f6d4.json new file mode 100644 index 0000000000000000000000000000000000000000..a2762a62b067d1d0ff3123ae5c43ef8ec8da8b4d --- /dev/null +++ b/data/vals_ai/xai/grok-4-0709/321356f3-59bd-44f1-9ee0-535e2923f6d4.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/xai_grok-4-0709/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-0709", + "id": "xai/grok-4-0709", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-0709", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.608, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.608", + "raw_stderr": "0.977", + "latency": "41.881", + "cost_per_test": "0.025477", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.977, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.508, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.508", + "raw_stderr": "0.992", + "latency": "124.718", + "cost_per_test": "0.054675", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.992, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.459, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.459", + "raw_stderr": "0.873", + "latency": "135.856", + "cost_per_test": "0.070869", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.873, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.585, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.585", + "raw_stderr": "1.683", + "latency": "46.115", + "cost_per_test": "0.050302", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.683, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.744, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.744", + "raw_stderr": "1.088", + "latency": "49.714", + "cost_per_test": "0.028301", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.088, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.456, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.456", + "raw_stderr": "1.222", + "latency": "257.063", + "cost_per_test": "0.147419", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.222, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.663, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.663", + "raw_stderr": "1.353", + "latency": "34.976", + "cost_per_test": "0.030109", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.353, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.591, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.591", + "raw_stderr": "2.201", + "latency": "50.819", + "cost_per_test": "0.032609", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.201, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.026, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.026", + "raw_stderr": "1.367", + "latency": "150.025", + "cost_per_test": "0.058366", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.367, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.822, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.822", + "raw_stderr": "0.786", + "latency": "57.184", + "cost_per_test": "0.051539", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.786, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.524, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.524", + "raw_stderr": "1.19", + "latency": "83.034", + "cost_per_test": "0.031888", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.19, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.304, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.304", + "raw_stderr": "0.349", + "latency": "89.255", + "cost_per_test": "0.05025", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.349, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.367, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.367", + "raw_stderr": "1.667", + "latency": "57.631", + "cost_per_test": "0.029648", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.667, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.069, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.069", + "raw_stderr": "0.83", + "latency": "120.658", + "cost_per_test": "0.069894", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.83, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.837, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.837", + "raw_stderr": "1.27", + "latency": "39.89", + "cost_per_test": "0.022401", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.27, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-0709/326f0276-49f3-4e7e-af3f-52010b4ee8ee.json b/data/vals_ai/xai/grok-4-0709/326f0276-49f3-4e7e-af3f-52010b4ee8ee.json new file mode 100644 index 0000000000000000000000000000000000000000..df0956b5a1fbe67aec8732d20af8b374e7d77792 --- /dev/null +++ b/data/vals_ai/xai/grok-4-0709/326f0276-49f3-4e7e-af3f-52010b4ee8ee.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/xai_grok-4-0709/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-0709", + "id": "xai/grok-4-0709", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-0709", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 99.068, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "99.068", + "raw_stderr": "0.535", + "latency": "50.479", + "cost_per_test": "0.032638", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 60.857, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "60.857", + "raw_stderr": "2.609", + "latency": "452.981", + "cost_per_test": "0.312336", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.609, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 89.817, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "89.817", + "raw_stderr": "1.545", + "latency": "184.746", + "cost_per_test": "0.122991", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.545, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 83.247, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "83.247", + "raw_stderr": "1.026", + "latency": "229.402", + "cost_per_test": "0.155988", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.026, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-0709/396d7623-faa6-4eeb-9f69-5e5cbf9d5996.json b/data/vals_ai/xai/grok-4-0709/396d7623-faa6-4eeb-9f69-5e5cbf9d5996.json new file mode 100644 index 0000000000000000000000000000000000000000..c13b5ab83032a22fb0727814c20f90ea28639888 --- /dev/null +++ b/data/vals_ai/xai/grok-4-0709/396d7623-faa6-4eeb-9f69-5e5cbf9d5996.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/xai_grok-4-0709/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-0709", + "id": "xai/grok-4-0709", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-0709", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 78.152, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "78.152", + "raw_stderr": "2.084", + "latency": "74.054", + "cost_per_test": "0.063955", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.084, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-0709/493c9750-8fd4-446c-b8ed-cb8f357daaf4.json b/data/vals_ai/xai/grok-4-0709/493c9750-8fd4-446c-b8ed-cb8f357daaf4.json new file mode 100644 index 0000000000000000000000000000000000000000..e9570dbefff3a872b63403bd6134f6f126c1b1cc --- /dev/null +++ b/data/vals_ai/xai/grok-4-0709/493c9750-8fd4-446c-b8ed-cb8f357daaf4.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/xai_grok-4-0709/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-0709", + "id": "xai/grok-4-0709", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-0709", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 87.879, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.879", + "raw_stderr": "2.319", + "latency": "119.477", + "cost_per_test": "0.152299", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.319, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 88.384, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.384", + "raw_stderr": "2.277", + "latency": "111.554", + "cost_per_test": "0.143643", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.277, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 88.132, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.132", + "raw_stderr": "1.625", + "latency": "115.516", + "cost_per_test": "0.147971", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.625, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-0709/50d74791-35f7-47ee-9a0d-de9d4df186a8.json b/data/vals_ai/xai/grok-4-0709/50d74791-35f7-47ee-9a0d-de9d4df186a8.json new file mode 100644 index 0000000000000000000000000000000000000000..fad480d3c59ae662b780d7ec0789a3c05830a6f9 --- /dev/null +++ b/data/vals_ai/xai/grok-4-0709/50d74791-35f7-47ee-9a0d-de9d4df186a8.json @@ -0,0 +1,371 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/xai_grok-4-0709/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-0709", + "id": "xai/grok-4-0709", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-0709", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 16.667, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "16.667", + "raw_stderr": "5.82", + "latency": "784.898", + "cost_per_test": "1.733504", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.82, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 51.724, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "51.724", + "raw_stderr": "3.099", + "latency": "647.212", + "cost_per_test": "1.61449", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.099, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 75.773, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "75.773", + "raw_stderr": "3.084", + "latency": "462.948", + "cost_per_test": "1.074998", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.084, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "757.453", + "cost_per_test": "1.61617", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 57.8, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "57.8", + "raw_stderr": "2.211", + "latency": "587.945", + "cost_per_test": "1.415174", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.211, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-0709/5314401c-1dec-4d8b-8575-339e4bf09dd3.json b/data/vals_ai/xai/grok-4-0709/5314401c-1dec-4d8b-8575-339e4bf09dd3.json new file mode 100644 index 0000000000000000000000000000000000000000..8581c34c558c423a171be7bdac1d23a34a71eda9 --- /dev/null +++ b/data/vals_ai/xai/grok-4-0709/5314401c-1dec-4d8b-8575-339e4bf09dd3.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/xai_grok-4-0709/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-0709", + "id": "xai/grok-4-0709", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-0709", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 32.909, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "32.909", + "raw_stderr": "1.325", + "latency": "97.415", + "cost_per_test": "0.114883", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.325, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 44.475, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "44.475", + "raw_stderr": "0.963", + "latency": "65.022", + "cost_per_test": "0.08575", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.963, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 56.041, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.041", + "raw_stderr": "1.399", + "latency": "32.629", + "cost_per_test": "0.056618", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.399, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-0709/58e56430-c742-4c29-b313-b1e005e2adec.json b/data/vals_ai/xai/grok-4-0709/58e56430-c742-4c29-b313-b1e005e2adec.json new file mode 100644 index 0000000000000000000000000000000000000000..9b6c87bc7cca9eb5efbfa06075a47c46fb3c84c0 --- /dev/null +++ b/data/vals_ai/xai/grok-4-0709/58e56430-c742-4c29-b313-b1e005e2adec.json @@ -0,0 +1,840 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mgsm/xai_grok-4-0709/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MGSM", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "dataset_type": "public", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-0709", + "id": "xai/grok-4-0709", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-0709", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mgsm:mgsm_bn:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.mgsm.mgsm_bn", + "source_data": { + "dataset_name": "MGSM - Bengali", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_bn", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Bengali).", + "metric_id": "vals_ai.mgsm.mgsm_bn.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 87.6, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_bn", + "task_name": "Bengali", + "dataset_type": "public", + "industry": "math", + "raw_score": "87.6", + "raw_stderr": "2.084", + "latency": "165.545", + "cost_per_test": "0.034715", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.084, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_de:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.mgsm.mgsm_de", + "source_data": { + "dataset_name": "MGSM - German", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_de", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (German).", + "metric_id": "vals_ai.mgsm.mgsm_de.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 92.0, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_de", + "task_name": "German", + "dataset_type": "public", + "industry": "math", + "raw_score": "92", + "raw_stderr": "1.716", + "latency": "160.431", + "cost_per_test": "0.023082", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.716, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_en:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.mgsm.mgsm_en", + "source_data": { + "dataset_name": "MGSM - English", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_en", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (English).", + "metric_id": "vals_ai.mgsm.mgsm_en.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 94.4, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_en", + "task_name": "English", + "dataset_type": "public", + "industry": "math", + "raw_score": "94.4", + "raw_stderr": "1.454", + "latency": "36.431", + "cost_per_test": "0.019169", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.454, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_es:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.mgsm.mgsm_es", + "source_data": { + "dataset_name": "MGSM - Spanish", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_es", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Spanish).", + "metric_id": "vals_ai.mgsm.mgsm_es.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 93.2, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_es", + "task_name": "Spanish", + "dataset_type": "public", + "industry": "math", + "raw_score": "93.2", + "raw_stderr": "1.592", + "latency": "157.382", + "cost_per_test": "0.022483", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.592, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_fr:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.mgsm.mgsm_fr", + "source_data": { + "dataset_name": "MGSM - French", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_fr", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (French).", + "metric_id": "vals_ai.mgsm.mgsm_fr.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 90.4, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_fr", + "task_name": "French", + "dataset_type": "public", + "industry": "math", + "raw_score": "90.4", + "raw_stderr": "1.863", + "latency": "164.804", + "cost_per_test": "0.024902", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.863, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_ja:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.mgsm.mgsm_ja", + "source_data": { + "dataset_name": "MGSM - Japanese", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_ja", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Japanese).", + "metric_id": "vals_ai.mgsm.mgsm_ja.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 88.8, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_ja", + "task_name": "Japanese", + "dataset_type": "public", + "industry": "math", + "raw_score": "88.8", + "raw_stderr": "1.995", + "latency": "48.936", + "cost_per_test": "0.02108", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.995, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_ru:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.mgsm.mgsm_ru", + "source_data": { + "dataset_name": "MGSM - Russian", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_ru", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Russian).", + "metric_id": "vals_ai.mgsm.mgsm_ru.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 94.0, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_ru", + "task_name": "Russian", + "dataset_type": "public", + "industry": "math", + "raw_score": "94", + "raw_stderr": "1.502", + "latency": "43.7", + "cost_per_test": "0.019564", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.502, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_sw:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.mgsm.mgsm_sw", + "source_data": { + "dataset_name": "MGSM - Swahili", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_sw", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Swahili).", + "metric_id": "vals_ai.mgsm.mgsm_sw.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 85.2, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_sw", + "task_name": "Swahili", + "dataset_type": "public", + "industry": "math", + "raw_score": "85.2", + "raw_stderr": "2.246", + "latency": "77.338", + "cost_per_test": "0.035632", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.246, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_te:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.mgsm.mgsm_te", + "source_data": { + "dataset_name": "MGSM - Telugu", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_te", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Telugu).", + "metric_id": "vals_ai.mgsm.mgsm_te.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 86.0, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_te", + "task_name": "Telugu", + "dataset_type": "public", + "industry": "math", + "raw_score": "86", + "raw_stderr": "2.194", + "latency": "206.881", + "cost_per_test": "0.043618", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.194, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_th:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.mgsm.mgsm_th", + "source_data": { + "dataset_name": "MGSM - Thai", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_th", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Thai).", + "metric_id": "vals_ai.mgsm.mgsm_th.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 96.0, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_th", + "task_name": "Thai", + "dataset_type": "public", + "industry": "math", + "raw_score": "96", + "raw_stderr": "1.239", + "latency": "180.431", + "cost_per_test": "0.025605", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.239, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_zh:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.mgsm.mgsm_zh", + "source_data": { + "dataset_name": "MGSM - Chinese", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_zh", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Chinese).", + "metric_id": "vals_ai.mgsm.mgsm_zh.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 92.4, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_zh", + "task_name": "Chinese", + "dataset_type": "public", + "industry": "math", + "raw_score": "92.4", + "raw_stderr": "1.676", + "latency": "40.934", + "cost_per_test": "0.017593", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.676, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:overall:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.mgsm.overall", + "source_data": { + "dataset_name": "MGSM - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Overall).", + "metric_id": "vals_ai.mgsm.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 90.909, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "math", + "raw_score": "90.909", + "raw_stderr": "1.778", + "latency": "116.619", + "cost_per_test": "0.026131", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.778, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-0709/761c0ae6-9efe-4ec2-bf1d-35bb18d0587e.json b/data/vals_ai/xai/grok-4-0709/761c0ae6-9efe-4ec2-bf1d-35bb18d0587e.json new file mode 100644 index 0000000000000000000000000000000000000000..ed951f8876a1914d8cee706c837363d38834058a --- /dev/null +++ b/data/vals_ai/xai/grok-4-0709/761c0ae6-9efe-4ec2-bf1d-35bb18d0587e.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2/xai_grok-4-0709/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.0", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2", + "benchmark_name": "Terminal-Bench 2.0", + "benchmark_updated": "2026-06-04", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-0709", + "id": "xai/grok-4-0709", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-0709", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2:easy:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.terminal-bench-2.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.0 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.0 (Easy).", + "metric_id": "vals_ai.terminal-bench-2.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "score_details": { + "score": 75.0, + "details": { + "benchmark_slug": "terminal-bench-2", + "benchmark_name": "Terminal-Bench 2.0", + "benchmark_updated": "2026-06-04", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "75", + "raw_stderr": "25", + "latency": "19695.363", + "cost_per_test": "12.298802", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 25.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2:hard:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.terminal-bench-2.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.0 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.0 (Hard).", + "metric_id": "vals_ai.terminal-bench-2.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "score_details": { + "score": 6.667, + "details": { + "benchmark_slug": "terminal-bench-2", + "benchmark_name": "Terminal-Bench 2.0", + "benchmark_updated": "2026-06-04", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "6.667", + "raw_stderr": "4.632", + "latency": "2626.048", + "cost_per_test": "1.63984", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.632, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2:medium:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.terminal-bench-2.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.0 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.0 (Medium).", + "metric_id": "vals_ai.terminal-bench-2.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "score_details": { + "score": 36.364, + "details": { + "benchmark_slug": "terminal-bench-2", + "benchmark_name": "Terminal-Bench 2.0", + "benchmark_updated": "2026-06-04", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "36.364", + "raw_stderr": "6.546", + "latency": "1432.39", + "cost_per_test": "0.894458", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.546, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2:overall:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.terminal-bench-2.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.0 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.0 (Overall).", + "metric_id": "vals_ai.terminal-bench-2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "score_details": { + "score": 28.09, + "details": { + "benchmark_slug": "terminal-bench-2", + "benchmark_name": "Terminal-Bench 2.0", + "benchmark_updated": "2026-06-04", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "28.09", + "raw_stderr": "4.791", + "latency": "885.185", + "cost_per_test": "0.552755", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.791, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-0709/87d72904-5a5c-48cc-8b40-5b74fede214d.json b/data/vals_ai/xai/grok-4-0709/87d72904-5a5c-48cc-8b40-5b74fede214d.json new file mode 100644 index 0000000000000000000000000000000000000000..e16bae6ff96742ad9ea6a5bdcb7762daa671893f --- /dev/null +++ b/data/vals_ai/xai/grok-4-0709/87d72904-5a5c-48cc-8b40-5b74fede214d.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medcode/xai_grok-4-0709/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedCode", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medcode", + "benchmark_name": "MedCode", + "benchmark_updated": "2026-07-08", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-0709", + "id": "xai/grok-4-0709", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-0709", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medcode:overall:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.medcode.overall", + "source_data": { + "dataset_name": "MedCode - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medcode", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedCode (Overall).", + "metric_id": "vals_ai.medcode.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode" + } + }, + "score_details": { + "score": 38.078, + "details": { + "benchmark_slug": "medcode", + "benchmark_name": "MedCode", + "benchmark_updated": "2026-07-08", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "38.078", + "raw_stderr": "2.206", + "latency": "89.002", + "cost_per_test": "0.034103", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.206, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-0709/9264e6f1-416f-4151-80a5-19ac476f644b.json b/data/vals_ai/xai/grok-4-0709/9264e6f1-416f-4151-80a5-19ac476f644b.json new file mode 100644 index 0000000000000000000000000000000000000000..e6acd03b69383bdc0b96eafc5ef8379d5e347663 --- /dev/null +++ b/data/vals_ai/xai/grok-4-0709/9264e6f1-416f-4151-80a5-19ac476f644b.json @@ -0,0 +1,505 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medqa/xai_grok-4-0709/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedQA", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "dataset_type": "public", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-0709", + "id": "xai/grok-4-0709", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-0709", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medqa:asian:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.medqa.asian", + "source_data": { + "dataset_name": "MedQA - Asian", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "asian", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Asian).", + "metric_id": "vals_ai.medqa.asian.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 91.8, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "asian", + "task_name": "Asian", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "91.8", + "raw_stderr": "0.614", + "latency": "66.998", + "cost_per_test": "0.026976", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.614, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:black:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.medqa.black", + "source_data": { + "dataset_name": "MedQA - Black", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "black", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Black).", + "metric_id": "vals_ai.medqa.black.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 91.85, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "black", + "task_name": "Black", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "91.85", + "raw_stderr": "0.612", + "latency": "67.636", + "cost_per_test": "0.027553", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.612, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:hispanic:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.medqa.hispanic", + "source_data": { + "dataset_name": "MedQA - Hispanic", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "hispanic", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Hispanic).", + "metric_id": "vals_ai.medqa.hispanic.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 92.0, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "hispanic", + "task_name": "Hispanic", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "92", + "raw_stderr": "0.607", + "latency": "65.196", + "cost_per_test": "0.026197", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.607, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:indigenous:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.medqa.indigenous", + "source_data": { + "dataset_name": "MedQA - Indigenous", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "indigenous", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Indigenous).", + "metric_id": "vals_ai.medqa.indigenous.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 92.4, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "indigenous", + "task_name": "Indigenous", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "92.4", + "raw_stderr": "0.593", + "latency": "63.94", + "cost_per_test": "0.025788", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.593, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:overall:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.medqa.overall", + "source_data": { + "dataset_name": "MedQA - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Overall).", + "metric_id": "vals_ai.medqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 92.492, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "92.492", + "raw_stderr": "0.24", + "latency": "64.247", + "cost_per_test": "0.025753", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.24, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:unbiased:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.medqa.unbiased", + "source_data": { + "dataset_name": "MedQA - Unbiased", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "unbiased", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Unbiased).", + "metric_id": "vals_ai.medqa.unbiased.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 93.7, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "unbiased", + "task_name": "Unbiased", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "93.7", + "raw_stderr": "0.543", + "latency": "58.874", + "cost_per_test": "0.023165", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.543, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:white:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.medqa.white", + "source_data": { + "dataset_name": "MedQA - White", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "white", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (White).", + "metric_id": "vals_ai.medqa.white.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 93.2, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "white", + "task_name": "White", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "93.2", + "raw_stderr": "0.563", + "latency": "62.836", + "cost_per_test": "0.024839", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.563, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-0709/aaf8c777-3a62-4332-af9b-6961518fd69e.json b/data/vals_ai/xai/grok-4-0709/aaf8c777-3a62-4332-af9b-6961518fd69e.json new file mode 100644 index 0000000000000000000000000000000000000000..7715028a65dc1a4a88609179645d9463b5a04707 --- /dev/null +++ b/data/vals_ai/xai/grok-4-0709/aaf8c777-3a62-4332-af9b-6961518fd69e.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/xai_grok-4-0709/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-0709", + "id": "xai/grok-4-0709", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-0709", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 65.967, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.967", + "raw_stderr": "1.618", + "latency": "11.944", + "cost_per_test": "0.021639", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.618, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 62.821, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.821", + "raw_stderr": "1.65", + "latency": "43.442", + "cost_per_test": "0.345687", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.65, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 66.045, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.045", + "raw_stderr": "0.932", + "latency": "29.733", + "cost_per_test": "0.190665", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.932, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 69.347, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "69.347", + "raw_stderr": "1.574", + "latency": "33.813", + "cost_per_test": "0.204668", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.574, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-0709/b7044fd6-02c2-4cc8-912f-8ba37c3238c4.json b/data/vals_ai/xai/grok-4-0709/b7044fd6-02c2-4cc8-912f-8ba37c3238c4.json new file mode 100644 index 0000000000000000000000000000000000000000..4cd1a572e8f9ecc141518c8e41654f10872a0357 --- /dev/null +++ b/data/vals_ai/xai/grok-4-0709/b7044fd6-02c2-4cc8-912f-8ba37c3238c4.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/xai_grok-4-0709/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-0709", + "id": "xai/grok-4-0709", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-0709", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 23.167, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "23.167", + "raw_stderr": "11.985", + "latency": "4824.866", + "cost_per_test": "1.953434", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 11.985, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 29.167, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "29.167", + "raw_stderr": "6.269", + "latency": "3706.897", + "cost_per_test": "3.434538", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.269, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 26.167, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "26.167", + "raw_stderr": "6.763", + "latency": "4265.882", + "cost_per_test": "2.693986", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.763, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-0709/ba68316b-5292-4800-a318-bd812e94b73b.json b/data/vals_ai/xai/grok-4-0709/ba68316b-5292-4800-a318-bd812e94b73b.json new file mode 100644 index 0000000000000000000000000000000000000000..67ea7d328bc36ed22e5b96bd1c8c00489864935c --- /dev/null +++ b/data/vals_ai/xai/grok-4-0709/ba68316b-5292-4800-a318-bd812e94b73b.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/aime/xai_grok-4-0709/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - AIME", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "dataset_type": "public", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-0709", + "id": "xai/grok-4-0709", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-0709", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aime:aime_2024:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.aime.aime_2024", + "source_data": { + "dataset_name": "AIME - AIME 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/aime" + ], + "additional_details": { + "benchmark_slug": "aime", + "task_key": "aime_2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for AIME (AIME 2024).", + "metric_id": "vals_ai.aime.aime_2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "score_details": { + "score": 93.333, + "details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "task_key": "aime_2024", + "task_name": "AIME 2024", + "dataset_type": "public", + "industry": "math", + "raw_score": "93.333", + "raw_stderr": "3.333", + "latency": "103.985", + "cost_per_test": "0.141831", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "aime:aime_2025:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.aime.aime_2025", + "source_data": { + "dataset_name": "AIME - AIME 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/aime" + ], + "additional_details": { + "benchmark_slug": "aime", + "task_key": "aime_2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for AIME (AIME 2025).", + "metric_id": "vals_ai.aime.aime_2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "score_details": { + "score": 87.778, + "details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "task_key": "aime_2025", + "task_name": "AIME 2025", + "dataset_type": "public", + "industry": "math", + "raw_score": "87.778", + "raw_stderr": "1.111", + "latency": "162.471", + "cost_per_test": "0.206967", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.111, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "aime:overall:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.aime.overall", + "source_data": { + "dataset_name": "AIME - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/aime" + ], + "additional_details": { + "benchmark_slug": "aime", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for AIME (Overall).", + "metric_id": "vals_ai.aime.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "score_details": { + "score": 90.556, + "details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "math", + "raw_score": "90.556", + "raw_stderr": "1.757", + "latency": "133.228", + "cost_per_test": "0.174399", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.757, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-0709/d9a7d6f0-b7e1-4b5b-a299-b93f65d3eacb.json b/data/vals_ai/xai/grok-4-0709/d9a7d6f0-b7e1-4b5b-a299-b93f65d3eacb.json new file mode 100644 index 0000000000000000000000000000000000000000..3397561d57d045f7589af7a7b3e12a13175f2e71 --- /dev/null +++ b/data/vals_ai/xai/grok-4-0709/d9a7d6f0-b7e1-4b5b-a299-b93f65d3eacb.json @@ -0,0 +1,432 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/xai_grok-4-0709/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-0709", + "id": "xai/grok-4-0709", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-0709", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.31, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.31", + "raw_stderr": "0.836", + "latency": "30.771", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.836, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.675, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.675", + "raw_stderr": "0.181", + "latency": "24.861", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.181, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 89.425, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.425", + "raw_stderr": "0.553", + "latency": "15.489", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.553, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.192, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.192", + "raw_stderr": "0.469", + "latency": "27.346", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.469, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.683, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.683", + "raw_stderr": "0.847", + "latency": "12.169", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.847, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 73.869, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "73.869", + "raw_stderr": "1.937", + "latency": "53.442", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.937, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-0709/e7da2216-07b1-430c-a9d4-d763375f4121.json b/data/vals_ai/xai/grok-4-0709/e7da2216-07b1-430c-a9d4-d763375f4121.json new file mode 100644 index 0000000000000000000000000000000000000000..1761e930fb6f4fc38fa64469a70e23c2b8436296 --- /dev/null +++ b/data/vals_ai/xai/grok-4-0709/e7da2216-07b1-430c-a9d4-d763375f4121.json @@ -0,0 +1,103 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/xai_grok-4-0709/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-0709", + "id": "xai/grok-4-0709", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-0709", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 76.27, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.27", + "raw_stderr": "1.022", + "latency": "102.831", + "cost_per_test": "0.097579", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.022, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-0709/f8725bb6-c3e5-4b0c-afe6-be7afd5c3540.json b/data/vals_ai/xai/grok-4-0709/f8725bb6-c3e5-4b0c-afe6-be7afd5c3540.json new file mode 100644 index 0000000000000000000000000000000000000000..efc6f2eff22defe80593f76c841b36055c9a3a7a --- /dev/null +++ b/data/vals_ai/xai/grok-4-0709/f8725bb6-c3e5-4b0c-afe6-be7afd5c3540.json @@ -0,0 +1,356 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/xai_grok-4-0709/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-0709", + "id": "xai/grok-4-0709", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-0709", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 15.079, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "15.079", + "raw_stderr": "4.491", + "latency": "245.263", + "cost_per_test": "0.344283", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.491, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 25.0, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "25", + "raw_stderr": "5.546", + "latency": "220.944", + "cost_per_test": "0.367405", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.546, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 29.938, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "29.938", + "raw_stderr": "6.125", + "latency": "169.15", + "cost_per_test": "1.176585", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.125, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 35.223, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "35.223", + "raw_stderr": "5.365", + "latency": "172.123", + "cost_per_test": "0.503586", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.365, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:grok/grok-4-0709:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 25.101, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "25.101", + "raw_stderr": "2.976", + "latency": "212.777", + "cost_per_test": "0.405091", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.976, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-non-reasoning/13eaeafc-2582-45ca-9836-59522c1c2dbe.json b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/13eaeafc-2582-45ca-9836-59522c1c2dbe.json new file mode 100644 index 0000000000000000000000000000000000000000..81e6e1a8aed123a2b734664786ec965c2010518f --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/13eaeafc-2582-45ca-9836-59522c1c2dbe.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medcode/xai_grok-4-1-fast-non-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedCode", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medcode", + "benchmark_name": "MedCode", + "benchmark_updated": "2026-07-08", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-non-reasoning", + "id": "xai/grok-4-1-fast-non-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-non-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medcode:overall:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.medcode.overall", + "source_data": { + "dataset_name": "MedCode - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medcode", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedCode (Overall).", + "metric_id": "vals_ai.medcode.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode" + } + }, + "score_details": { + "score": 28.349, + "details": { + "benchmark_slug": "medcode", + "benchmark_name": "MedCode", + "benchmark_updated": "2026-07-08", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "28.349", + "raw_stderr": "1.921", + "latency": "4.496", + "cost_per_test": "0.002193", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.921, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-non-reasoning/178f8404-8d4e-446a-b552-2376dddaafe4.json b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/178f8404-8d4e-446a-b552-2376dddaafe4.json new file mode 100644 index 0000000000000000000000000000000000000000..bb26ffef6d17c1b1f5d4ffb1cd43a24909a4106b --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/178f8404-8d4e-446a-b552-2376dddaafe4.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2/xai_grok-4-1-fast-non-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.0", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2", + "benchmark_name": "Terminal-Bench 2.0", + "benchmark_updated": "2026-06-04", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-non-reasoning", + "id": "xai/grok-4-1-fast-non-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-non-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2:easy:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.terminal-bench-2.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.0 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.0 (Easy).", + "metric_id": "vals_ai.terminal-bench-2.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "score_details": { + "score": 50.0, + "details": { + "benchmark_slug": "terminal-bench-2", + "benchmark_name": "Terminal-Bench 2.0", + "benchmark_updated": "2026-06-04", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "50", + "raw_stderr": "28.868", + "latency": "5633.071", + "cost_per_test": "0.730044", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 28.868, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2:hard:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.terminal-bench-2.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.0 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.0 (Hard).", + "metric_id": "vals_ai.terminal-bench-2.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "score_details": { + "score": 6.667, + "details": { + "benchmark_slug": "terminal-bench-2", + "benchmark_name": "Terminal-Bench 2.0", + "benchmark_updated": "2026-06-04", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "6.667", + "raw_stderr": "4.632", + "latency": "751.076", + "cost_per_test": "0.097339", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.632, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2:medium:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.terminal-bench-2.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.0 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.0 (Medium).", + "metric_id": "vals_ai.terminal-bench-2.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "score_details": { + "score": 21.818, + "details": { + "benchmark_slug": "terminal-bench-2", + "benchmark_name": "Terminal-Bench 2.0", + "benchmark_updated": "2026-06-04", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "21.818", + "raw_stderr": "5.62", + "latency": "409.678", + "cost_per_test": "0.053094", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.62, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2:overall:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.terminal-bench-2.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.0 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.0 (Overall).", + "metric_id": "vals_ai.terminal-bench-2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "score_details": { + "score": 17.978, + "details": { + "benchmark_slug": "terminal-bench-2", + "benchmark_name": "Terminal-Bench 2.0", + "benchmark_updated": "2026-06-04", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "17.978", + "raw_stderr": "4.093", + "latency": "253.172", + "cost_per_test": "0.032811", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.093, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-non-reasoning/215ec7b8-6a23-489a-a383-bf1ed2e47b98.json b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/215ec7b8-6a23-489a-a383-bf1ed2e47b98.json new file mode 100644 index 0000000000000000000000000000000000000000..eb1aac39db6a9a571b0ebf4b3aee90aa683fd233 --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/215ec7b8-6a23-489a-a383-bf1ed2e47b98.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/xai_grok-4-1-fast-non-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-non-reasoning", + "id": "xai/grok-4-1-fast-non-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-non-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 64.646, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "64.646", + "raw_stderr": "3.397", + "latency": "8.363", + "cost_per_test": "0.000498", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.397, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 65.657, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "65.657", + "raw_stderr": "3.386", + "latency": "8.956", + "cost_per_test": "0.000392", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.386, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 65.152, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "65.152", + "raw_stderr": "2.398", + "latency": "8.66", + "cost_per_test": "0.000445", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.398, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-non-reasoning/4e060ba0-f650-481e-8299-2ce6a5c1341a.json b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/4e060ba0-f650-481e-8299-2ce6a5c1341a.json new file mode 100644 index 0000000000000000000000000000000000000000..f9a2b8cb5126261e3cfe640316b1e6c71c05416a --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/4e060ba0-f650-481e-8299-2ce6a5c1341a.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/aime/xai_grok-4-1-fast-non-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - AIME", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "dataset_type": "public", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-non-reasoning", + "id": "xai/grok-4-1-fast-non-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-non-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aime:aime_2024:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.aime.aime_2024", + "source_data": { + "dataset_name": "AIME - AIME 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/aime" + ], + "additional_details": { + "benchmark_slug": "aime", + "task_key": "aime_2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for AIME (AIME 2024).", + "metric_id": "vals_ai.aime.aime_2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "score_details": { + "score": 20.833, + "details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "task_key": "aime_2024", + "task_name": "AIME 2024", + "dataset_type": "public", + "industry": "math", + "raw_score": "20.833", + "raw_stderr": "2.8", + "latency": "12.902", + "cost_per_test": "0.00059", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.8, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "aime:aime_2025:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.aime.aime_2025", + "source_data": { + "dataset_name": "AIME - AIME 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/aime" + ], + "additional_details": { + "benchmark_slug": "aime", + "task_key": "aime_2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for AIME (AIME 2025).", + "metric_id": "vals_ai.aime.aime_2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "task_key": "aime_2025", + "task_name": "AIME 2025", + "dataset_type": "public", + "industry": "math", + "raw_score": "33.333", + "raw_stderr": "2.673", + "latency": "21.828", + "cost_per_test": "0.000987", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.673, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "aime:overall:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.aime.overall", + "source_data": { + "dataset_name": "AIME - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/aime" + ], + "additional_details": { + "benchmark_slug": "aime", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for AIME (Overall).", + "metric_id": "vals_ai.aime.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "score_details": { + "score": 27.083, + "details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "math", + "raw_score": "27.083", + "raw_stderr": "1.935", + "latency": "17.366", + "cost_per_test": "0.000788", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.935, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-non-reasoning/55dcd8bf-befd-4424-89f7-25f3278fe762.json b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/55dcd8bf-befd-4424-89f7-25f3278fe762.json new file mode 100644 index 0000000000000000000000000000000000000000..b6499fc0e413c54e48b33a9ec979ab1c8382dcdf --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/55dcd8bf-befd-4424-89f7-25f3278fe762.json @@ -0,0 +1,103 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/xai_grok-4-1-fast-non-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-non-reasoning", + "id": "xai/grok-4-1-fast-non-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-non-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 63.699, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "63.699", + "raw_stderr": "1.156", + "latency": "51.426", + "cost_per_test": "0.000628", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.156, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-non-reasoning/56249465-cc66-4a52-b47f-39e64bb19b4f.json b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/56249465-cc66-4a52-b47f-39e64bb19b4f.json new file mode 100644 index 0000000000000000000000000000000000000000..e90fe001ce882970d497e9b567cc9dfd48c82c07 --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/56249465-cc66-4a52-b47f-39e64bb19b4f.json @@ -0,0 +1,840 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mgsm/xai_grok-4-1-fast-non-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MGSM", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "dataset_type": "public", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-non-reasoning", + "id": "xai/grok-4-1-fast-non-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-non-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mgsm:mgsm_bn:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_bn", + "source_data": { + "dataset_name": "MGSM - Bengali", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_bn", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Bengali).", + "metric_id": "vals_ai.mgsm.mgsm_bn.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 79.6, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_bn", + "task_name": "Bengali", + "dataset_type": "public", + "industry": "math", + "raw_score": "79.6", + "raw_stderr": "2.549", + "latency": "6.105", + "cost_per_test": "0.000242", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.549, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_de:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_de", + "source_data": { + "dataset_name": "MGSM - German", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_de", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (German).", + "metric_id": "vals_ai.mgsm.mgsm_de.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 91.2, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_de", + "task_name": "German", + "dataset_type": "public", + "industry": "math", + "raw_score": "91.2", + "raw_stderr": "1.792", + "latency": "3.637", + "cost_per_test": "0.000148", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.792, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_en:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_en", + "source_data": { + "dataset_name": "MGSM - English", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_en", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (English).", + "metric_id": "vals_ai.mgsm.mgsm_en.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 96.0, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_en", + "task_name": "English", + "dataset_type": "public", + "industry": "math", + "raw_score": "96", + "raw_stderr": "1.239", + "latency": "2.11", + "cost_per_test": "8.9e-05", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.239, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_es:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_es", + "source_data": { + "dataset_name": "MGSM - Spanish", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_es", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Spanish).", + "metric_id": "vals_ai.mgsm.mgsm_es.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 89.2, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_es", + "task_name": "Spanish", + "dataset_type": "public", + "industry": "math", + "raw_score": "89.2", + "raw_stderr": "1.963", + "latency": "2.626", + "cost_per_test": "0.000111", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.963, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_fr:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_fr", + "source_data": { + "dataset_name": "MGSM - French", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_fr", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (French).", + "metric_id": "vals_ai.mgsm.mgsm_fr.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 82.4, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_fr", + "task_name": "French", + "dataset_type": "public", + "industry": "math", + "raw_score": "82.4", + "raw_stderr": "2.408", + "latency": "2.998", + "cost_per_test": "0.000121", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.408, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_ja:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_ja", + "source_data": { + "dataset_name": "MGSM - Japanese", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_ja", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Japanese).", + "metric_id": "vals_ai.mgsm.mgsm_ja.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 85.2, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_ja", + "task_name": "Japanese", + "dataset_type": "public", + "industry": "math", + "raw_score": "85.2", + "raw_stderr": "2.246", + "latency": "3.418", + "cost_per_test": "0.000137", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.246, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_ru:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_ru", + "source_data": { + "dataset_name": "MGSM - Russian", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_ru", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Russian).", + "metric_id": "vals_ai.mgsm.mgsm_ru.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 92.4, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_ru", + "task_name": "Russian", + "dataset_type": "public", + "industry": "math", + "raw_score": "92.4", + "raw_stderr": "1.676", + "latency": "3.01", + "cost_per_test": "0.000124", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.676, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_sw:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_sw", + "source_data": { + "dataset_name": "MGSM - Swahili", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_sw", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Swahili).", + "metric_id": "vals_ai.mgsm.mgsm_sw.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 84.8, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_sw", + "task_name": "Swahili", + "dataset_type": "public", + "industry": "math", + "raw_score": "84.8", + "raw_stderr": "2.271", + "latency": "3.751", + "cost_per_test": "0.000134", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.271, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_te:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_te", + "source_data": { + "dataset_name": "MGSM - Telugu", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_te", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Telugu).", + "metric_id": "vals_ai.mgsm.mgsm_te.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 84.4, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_te", + "task_name": "Telugu", + "dataset_type": "public", + "industry": "math", + "raw_score": "84.4", + "raw_stderr": "2.295", + "latency": "6.731", + "cost_per_test": "0.000261", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.295, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_th:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_th", + "source_data": { + "dataset_name": "MGSM - Thai", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_th", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Thai).", + "metric_id": "vals_ai.mgsm.mgsm_th.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 89.2, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_th", + "task_name": "Thai", + "dataset_type": "public", + "industry": "math", + "raw_score": "89.2", + "raw_stderr": "1.963", + "latency": "3.532", + "cost_per_test": "0.000145", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.963, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_zh:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_zh", + "source_data": { + "dataset_name": "MGSM - Chinese", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_zh", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Chinese).", + "metric_id": "vals_ai.mgsm.mgsm_zh.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 88.8, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_zh", + "task_name": "Chinese", + "dataset_type": "public", + "industry": "math", + "raw_score": "88.8", + "raw_stderr": "1.995", + "latency": "2.594", + "cost_per_test": "0.000109", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.995, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:overall:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mgsm.overall", + "source_data": { + "dataset_name": "MGSM - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Overall).", + "metric_id": "vals_ai.mgsm.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 87.564, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "math", + "raw_score": "87.564", + "raw_stderr": "2.036", + "latency": "3.683", + "cost_per_test": "0.000147", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.036, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-non-reasoning/926126b7-697a-42f4-9394-633ce3af72be.json b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/926126b7-697a-42f4-9394-633ce3af72be.json new file mode 100644 index 0000000000000000000000000000000000000000..6c3ab96dba53890aa2199c5232be2d4a64154c97 --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/926126b7-697a-42f4-9394-633ce3af72be.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/xai_grok-4-1-fast-non-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-non-reasoning", + "id": "xai/grok-4-1-fast-non-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-non-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 77.464, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "77.464", + "raw_stderr": "2.04", + "latency": "19.865", + "cost_per_test": "0.001782", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.04, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-non-reasoning/a61f0616-d23f-4413-88f3-91b735359097.json b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/a61f0616-d23f-4413-88f3-91b735359097.json new file mode 100644 index 0000000000000000000000000000000000000000..34565d539204fa45d80e03efa14a95467af01134 --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/a61f0616-d23f-4413-88f3-91b735359097.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/xai_grok-4-1-fast-non-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-non-reasoning", + "id": "xai/grok-4-1-fast-non-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-non-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 7.667, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "7.667", + "raw_stderr": "5.13", + "latency": "4647.63", + "cost_per_test": "0.047983", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.13, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 7.667, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "7.667", + "raw_stderr": "3.82", + "latency": "10702.953", + "cost_per_test": "0.091116", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.82, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 7.667, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "7.667", + "raw_stderr": "3.198", + "latency": "7675.292", + "cost_per_test": "0.06955", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.198, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-non-reasoning/afdb8710-8b69-432c-a59d-468428cf8ae6.json b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/afdb8710-8b69-432c-a59d-468428cf8ae6.json new file mode 100644 index 0000000000000000000000000000000000000000..b3b6efdd73b6079c490f795f78d59bafb60fe89a --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/afdb8710-8b69-432c-a59d-468428cf8ae6.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/xai_grok-4-1-fast-non-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-non-reasoning", + "id": "xai/grok-4-1-fast-non-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-non-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.448, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.448", + "raw_stderr": "1.237", + "latency": "2.646", + "cost_per_test": "0.000211", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.237, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.834, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.834", + "raw_stderr": "1.454", + "latency": "3.11", + "cost_per_test": "0.000205", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.454, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.297, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.297", + "raw_stderr": "1.245", + "latency": "4.232", + "cost_per_test": "0.000291", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.245, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.537, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.537", + "raw_stderr": "2.052", + "latency": "4.302", + "cost_per_test": "0.000237", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.052, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.583, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.583", + "raw_stderr": "1.305", + "latency": "1.975", + "cost_per_test": "0.000171", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.305, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 62.126, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "62.126", + "raw_stderr": "1.558", + "latency": "6.197", + "cost_per_test": "0.000363", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.558, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.594, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.594", + "raw_stderr": "1.541", + "latency": "1.964", + "cost_per_test": "0.000149", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.541, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 66.142, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.142", + "raw_stderr": "2.424", + "latency": "2.078", + "cost_per_test": "0.000239", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.424, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 59.582, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "59.582", + "raw_stderr": "1.479", + "latency": "2.596", + "cost_per_test": "0.000217", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.479, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.051, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.051", + "raw_stderr": "1.066", + "latency": "3.881", + "cost_per_test": "0.000279", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.066, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.351, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.351", + "raw_stderr": "1.437", + "latency": "1.454", + "cost_per_test": "0.000127", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.437, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.214, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.214", + "raw_stderr": "0.419", + "latency": "3.098", + "cost_per_test": "0.000217", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.419, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 72.946, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.946", + "raw_stderr": "1.989", + "latency": "2.103", + "cost_per_test": "0.000151", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.989, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.06, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.06", + "raw_stderr": "1.148", + "latency": "4.979", + "cost_per_test": "0.00025", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.148, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.451, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.451", + "raw_stderr": "1.404", + "latency": "1.854", + "cost_per_test": "0.000151", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.404, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-non-reasoning/b7d8b7aa-77ce-49a3-a97f-0b30f23693bd.json b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/b7d8b7aa-77ce-49a3-a97f-0b30f23693bd.json new file mode 100644 index 0000000000000000000000000000000000000000..c6f2db42c2907f555430fd6634849ff306b68b41 --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/b7d8b7aa-77ce-49a3-a97f-0b30f23693bd.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/xai_grok-4-1-fast-non-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-non-reasoning", + "id": "xai/grok-4-1-fast-non-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-non-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 44.072, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "44.072", + "raw_stderr": "1.42", + "latency": "9.277", + "cost_per_test": "0.000372", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.42, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 61.856, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.856", + "raw_stderr": "0.914", + "latency": "9.277", + "cost_per_test": "0.000372", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.914, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 79.64, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "79.64", + "raw_stderr": "1.151", + "latency": "9.277", + "cost_per_test": "0.000372", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.151, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-non-reasoning/ce85375b-736f-44dd-b4ee-f02f124a1faa.json b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/ce85375b-736f-44dd-b4ee-f02f124a1faa.json new file mode 100644 index 0000000000000000000000000000000000000000..b4a57353953728b58eda3138e1b5ce2bfd1a14dc --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/ce85375b-736f-44dd-b4ee-f02f124a1faa.json @@ -0,0 +1,356 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/xai_grok-4-1-fast-non-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-non-reasoning", + "id": "xai/grok-4-1-fast-non-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-non-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 3.175, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "3.175", + "raw_stderr": "2.2", + "latency": "5.926", + "cost_per_test": "0.000744", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.2, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 11.112, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "11.112", + "raw_stderr": "4.009", + "latency": "5.85", + "cost_per_test": "0.001684", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.009, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 19.136, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "19.136", + "raw_stderr": "5.117", + "latency": "18.058", + "cost_per_test": "0.010491", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.117, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 21.457, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "21.457", + "raw_stderr": "4.611", + "latency": "6.047", + "cost_per_test": "0.001468", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.611, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 11.915, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "11.915", + "raw_stderr": "2.165", + "latency": "5.941", + "cost_per_test": "0.001299", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.165, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-non-reasoning/d7cd1eb9-a2fd-4d84-b440-79d06b77099d.json b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/d7cd1eb9-a2fd-4d84-b440-79d06b77099d.json new file mode 100644 index 0000000000000000000000000000000000000000..ba8d21fa4c7fec1b7fff666bcfb19c490316e2d8 --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/d7cd1eb9-a2fd-4d84-b440-79d06b77099d.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/xai_grok-4-1-fast-non-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-non-reasoning", + "id": "xai/grok-4-1-fast-non-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-non-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 40.461, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "40.461", + "raw_stderr": "1.384", + "latency": "11.723", + "cost_per_test": "0.000997", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.384, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 34.26, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "34.26", + "raw_stderr": "0.938", + "latency": "8.886", + "cost_per_test": "0.001018", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.938, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 28.06, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "28.06", + "raw_stderr": "1.267", + "latency": "6.049", + "cost_per_test": "0.00104", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.267, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-non-reasoning/dca3b04e-9137-4029-930f-2ae18434eb15.json b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/dca3b04e-9137-4029-930f-2ae18434eb15.json new file mode 100644 index 0000000000000000000000000000000000000000..5b68c197c33b75dcf1e2b185bf643d184b33562f --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/dca3b04e-9137-4029-930f-2ae18434eb15.json @@ -0,0 +1,438 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/xai_grok-4-1-fast-non-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-non-reasoning", + "id": "xai/grok-4-1-fast-non-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-non-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.804, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.804", + "raw_stderr": "0.945", + "latency": "0.589", + "cost_per_test": "4.7e-05", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.945, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.668, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.668", + "raw_stderr": "0.214", + "latency": "1.594", + "cost_per_test": "9e-05", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.214, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 85.753, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.753", + "raw_stderr": "0.643", + "latency": "3.897", + "cost_per_test": "0.000159", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.643, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.113, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.113", + "raw_stderr": "0.46", + "latency": "1.757", + "cost_per_test": "9.1e-05", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.46, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.595, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.595", + "raw_stderr": "1.025", + "latency": "1.036", + "cost_per_test": "8.5e-05", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.025, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 70.747, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "70.747", + "raw_stderr": "1.696", + "latency": "1.67", + "cost_per_test": "7.2e-05", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.696, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-non-reasoning/f6802a9b-c185-49c6-9c2d-f85540553a2e.json b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/f6802a9b-c185-49c6-9c2d-f85540553a2e.json new file mode 100644 index 0000000000000000000000000000000000000000..834ee8538f8c1ac38e1b82589c98368e6ac2edc0 --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/f6802a9b-c185-49c6-9c2d-f85540553a2e.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/xai_grok-4-1-fast-non-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-non-reasoning", + "id": "xai/grok-4-1-fast-non-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-non-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 88.82, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "88.82", + "raw_stderr": "1.756", + "latency": "3.173", + "cost_per_test": "0.000199", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.756, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 7.714, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "7.714", + "raw_stderr": "1.426", + "latency": "17.265", + "cost_per_test": "0.000716", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.426, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 31.332, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "31.332", + "raw_stderr": "2.37", + "latency": "15.447", + "cost_per_test": "0.000392", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.37, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 42.622, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "42.622", + "raw_stderr": "1.092", + "latency": "11.962", + "cost_per_test": "0.000436", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.092, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-non-reasoning/f940332e-f89c-420f-a9b1-21f32e41a96b.json b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/f940332e-f89c-420f-a9b1-21f32e41a96b.json new file mode 100644 index 0000000000000000000000000000000000000000..63120f9d5cbc19db74b166da46021d39fb738c6b --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/f940332e-f89c-420f-a9b1-21f32e41a96b.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/xai_grok-4-1-fast-non-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-non-reasoning", + "id": "xai/grok-4-1-fast-non-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-non-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 46.853, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "46.853", + "raw_stderr": "1.704", + "latency": "19.906", + "cost_per_test": "0.03445", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.704, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 64.103, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.103", + "raw_stderr": "1.638", + "latency": "12.831", + "cost_per_test": "0.000347", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.638, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 52.486, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.486", + "raw_stderr": "0.971", + "latency": "13.465", + "cost_per_test": "0.015705", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.971, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 46.503, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "46.503", + "raw_stderr": "1.703", + "latency": "7.659", + "cost_per_test": "0.012319", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.703, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-non-reasoning/fc84feb6-8d4e-4be4-b092-5b8b670b6d51.json b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/fc84feb6-8d4e-4be4-b092-5b8b670b6d51.json new file mode 100644 index 0000000000000000000000000000000000000000..bb9b843a68f6a3ff23aa4fd81c6d10c0eb4a94a1 --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-non-reasoning/fc84feb6-8d4e-4be4-b092-5b8b670b6d51.json @@ -0,0 +1,505 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medqa/xai_grok-4-1-fast-non-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedQA", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "dataset_type": "public", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-non-reasoning", + "id": "xai/grok-4-1-fast-non-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-non-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medqa:asian:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.medqa.asian", + "source_data": { + "dataset_name": "MedQA - Asian", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "asian", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Asian).", + "metric_id": "vals_ai.medqa.asian.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 76.3, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "asian", + "task_name": "Asian", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "76.3", + "raw_stderr": "0.951", + "latency": "7.339", + "cost_per_test": "0.000165", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.951, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:black:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.medqa.black", + "source_data": { + "dataset_name": "MedQA - Black", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "black", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Black).", + "metric_id": "vals_ai.medqa.black.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 75.55, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "black", + "task_name": "Black", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "75.55", + "raw_stderr": "0.961", + "latency": "5.003", + "cost_per_test": "0.000168", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.961, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:hispanic:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.medqa.hispanic", + "source_data": { + "dataset_name": "MedQA - Hispanic", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "hispanic", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Hispanic).", + "metric_id": "vals_ai.medqa.hispanic.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 75.45, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "hispanic", + "task_name": "Hispanic", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "75.45", + "raw_stderr": "0.962", + "latency": "7.301", + "cost_per_test": "0.000166", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.962, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:indigenous:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.medqa.indigenous", + "source_data": { + "dataset_name": "MedQA - Indigenous", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "indigenous", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Indigenous).", + "metric_id": "vals_ai.medqa.indigenous.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 75.35, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "indigenous", + "task_name": "Indigenous", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "75.35", + "raw_stderr": "0.964", + "latency": "7.393", + "cost_per_test": "0.000167", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.964, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:overall:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.medqa.overall", + "source_data": { + "dataset_name": "MedQA - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Overall).", + "metric_id": "vals_ai.medqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 76.025, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "76.025", + "raw_stderr": "0.39", + "latency": "6.552", + "cost_per_test": "0.000165", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.39, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:unbiased:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.medqa.unbiased", + "source_data": { + "dataset_name": "MedQA - Unbiased", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "unbiased", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Unbiased).", + "metric_id": "vals_ai.medqa.unbiased.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 77.5, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "unbiased", + "task_name": "Unbiased", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "77.5", + "raw_stderr": "0.934", + "latency": "4.928", + "cost_per_test": "0.000159", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.934, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:white:grok/grok-4-1-fast-non-reasoning:score", + "evaluation_name": "vals_ai.medqa.white", + "source_data": { + "dataset_name": "MedQA - White", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "white", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (White).", + "metric_id": "vals_ai.medqa.white.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 76.0, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "white", + "task_name": "White", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "76", + "raw_stderr": "0.955", + "latency": "7.348", + "cost_per_test": "0.000166", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.955, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-reasoning/045c3f5b-34b2-4fb5-a3e6-fe525d592689.json b/data/vals_ai/xai/grok-4-1-fast-reasoning/045c3f5b-34b2-4fb5-a3e6-fe525d592689.json new file mode 100644 index 0000000000000000000000000000000000000000..54b4064708fcf1c0ce3ce57ee9485cd5f563b15b --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-reasoning/045c3f5b-34b2-4fb5-a3e6-fe525d592689.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/case_law_v2/xai_grok-4-1-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CaseLaw v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "case_law_v2", + "benchmark_name": "CaseLaw v2", + "benchmark_updated": "2026-05-04", + "dataset_type": "private", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/case_law_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-reasoning", + "id": "xai/grok-4-1-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "case_law_v2:overall:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.case_law_v2.overall", + "source_data": { + "dataset_name": "CaseLaw v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "case_law_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/case_law_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CaseLaw v2 (Overall).", + "metric_id": "vals_ai.case_law_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/case_law_v2" + } + }, + "score_details": { + "score": 60.45, + "details": { + "benchmark_slug": "case_law_v2", + "benchmark_name": "CaseLaw v2", + "benchmark_updated": "2026-05-04", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "legal", + "raw_score": "60.45", + "raw_stderr": "0.809", + "latency": "19.094", + "cost_per_test": "0.011453", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.809, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-reasoning/08d35a4c-06c3-46ba-8e7e-5fea608e4993.json b/data/vals_ai/xai/grok-4-1-fast-reasoning/08d35a4c-06c3-46ba-8e7e-5fea608e4993.json new file mode 100644 index 0000000000000000000000000000000000000000..597c1c2c30631cdedcbe8c6286be9577f41b058a --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-reasoning/08d35a4c-06c3-46ba-8e7e-5fea608e4993.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/xai_grok-4-1-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-reasoning", + "id": "xai/grok-4-1-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 83.838, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.838", + "raw_stderr": "2.616", + "latency": "70.139", + "cost_per_test": "0.003089", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.616, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 84.848, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.848", + "raw_stderr": "2.616", + "latency": "66.823", + "cost_per_test": "0.002743", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.616, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 84.343, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.343", + "raw_stderr": "1.85", + "latency": "68.481", + "cost_per_test": "0.002916", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.85, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-reasoning/0c17df63-90cb-4367-9efe-a406debda771.json b/data/vals_ai/xai/grok-4-1-fast-reasoning/0c17df63-90cb-4367-9efe-a406debda771.json new file mode 100644 index 0000000000000000000000000000000000000000..2bbc45a28bf68b265c804e9d20bcc7a386f88af2 --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-reasoning/0c17df63-90cb-4367-9efe-a406debda771.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2/xai_grok-4-1-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.0", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2", + "benchmark_name": "Terminal-Bench 2.0", + "benchmark_updated": "2026-06-04", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-reasoning", + "id": "xai/grok-4-1-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2:easy:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.terminal-bench-2.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.0 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.0 (Easy).", + "metric_id": "vals_ai.terminal-bench-2.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "score_details": { + "score": 75.0, + "details": { + "benchmark_slug": "terminal-bench-2", + "benchmark_name": "Terminal-Bench 2.0", + "benchmark_updated": "2026-06-04", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "75", + "raw_stderr": "25", + "latency": "10661.252", + "cost_per_test": "1.235461", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 25.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2:hard:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.terminal-bench-2.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.0 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.0 (Hard).", + "metric_id": "vals_ai.terminal-bench-2.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "score_details": { + "score": 3.333, + "details": { + "benchmark_slug": "terminal-bench-2", + "benchmark_name": "Terminal-Bench 2.0", + "benchmark_updated": "2026-06-04", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "3.333", + "raw_stderr": "3.333", + "latency": "1421.5", + "cost_per_test": "0.164728", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2:medium:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.terminal-bench-2.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.0 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.0 (Medium).", + "metric_id": "vals_ai.terminal-bench-2.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "score_details": { + "score": 32.727, + "details": { + "benchmark_slug": "terminal-bench-2", + "benchmark_name": "Terminal-Bench 2.0", + "benchmark_updated": "2026-06-04", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "32.727", + "raw_stderr": "6.385", + "latency": "775.364", + "cost_per_test": "0.089852", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.385, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2:overall:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.terminal-bench-2.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.0 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.0 (Overall).", + "metric_id": "vals_ai.terminal-bench-2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "score_details": { + "score": 24.719, + "details": { + "benchmark_slug": "terminal-bench-2", + "benchmark_name": "Terminal-Bench 2.0", + "benchmark_updated": "2026-06-04", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "24.719", + "raw_stderr": "4.599", + "latency": "479.157", + "cost_per_test": "0.055526", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.599, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-reasoning/1b3d1deb-d488-4a91-802c-475fa6d288cd.json b/data/vals_ai/xai/grok-4-1-fast-reasoning/1b3d1deb-d488-4a91-802c-475fa6d288cd.json new file mode 100644 index 0000000000000000000000000000000000000000..a2c0282701608d1c7a7b10bb5dac7d907fbc8f17 --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-reasoning/1b3d1deb-d488-4a91-802c-475fa6d288cd.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/xai_grok-4-1-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-reasoning", + "id": "xai/grok-4-1-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.026, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.026", + "raw_stderr": "0.951", + "latency": "13.658", + "cost_per_test": "0.00067", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.951, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.72, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.72", + "raw_stderr": "1.126", + "latency": "24.113", + "cost_per_test": "0.001053", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.126, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.633, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.633", + "raw_stderr": "0.978", + "latency": "37.247", + "cost_per_test": "0.001572", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.978, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.561, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.561", + "raw_stderr": "1.63", + "latency": "21.213", + "cost_per_test": "0.000995", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.63, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.678, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.678", + "raw_stderr": "1.131", + "latency": "13.023", + "cost_per_test": "0.000626", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.131, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.773, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.773", + "raw_stderr": "1.29", + "latency": "71.805", + "cost_per_test": "0.003066", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.29, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.218, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.218", + "raw_stderr": "1.419", + "latency": "15.775", + "cost_per_test": "0.000726", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.419, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.853, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.853", + "raw_stderr": "2.193", + "latency": "15.211", + "cost_per_test": "0.000752", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.193, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.117, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.117", + "raw_stderr": "1.366", + "latency": "42.667", + "cost_per_test": "0.001114", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.366, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.6, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.6", + "raw_stderr": "0.794", + "latency": "29.009", + "cost_per_test": "0.001272", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.794, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.061, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.061", + "raw_stderr": "1.289", + "latency": "15.741", + "cost_per_test": "0.000682", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.289, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.18, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.18", + "raw_stderr": "0.357", + "latency": "25.469", + "cost_per_test": "0.001078", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.357, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.162, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.162", + "raw_stderr": "1.75", + "latency": "14.651", + "cost_per_test": "0.00065", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.75, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.53, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.53", + "raw_stderr": "0.849", + "latency": "30.45", + "cost_per_test": "0.001349", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.849, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.589, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.589", + "raw_stderr": "1.243", + "latency": "12.005", + "cost_per_test": "0.00056", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.243, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-reasoning/37ad21db-a2a3-4e57-834c-524961a98dc0.json b/data/vals_ai/xai/grok-4-1-fast-reasoning/37ad21db-a2a3-4e57-834c-524961a98dc0.json new file mode 100644 index 0000000000000000000000000000000000000000..2891b7a55847aaaa25c71c1ad20401cadbcb288e --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-reasoning/37ad21db-a2a3-4e57-834c-524961a98dc0.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/xai_grok-4-1-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-reasoning", + "id": "xai/grok-4-1-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 65.501, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.501", + "raw_stderr": "1.623", + "latency": "32.826", + "cost_per_test": "0.033844", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.623, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 64.802, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.802", + "raw_stderr": "1.63", + "latency": "33.846", + "cost_per_test": "0.000351", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.63, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 65.967, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.967", + "raw_stderr": "0.934", + "latency": "28.411", + "cost_per_test": "0.015643", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.934, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 67.599, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.599", + "raw_stderr": "1.598", + "latency": "18.561", + "cost_per_test": "0.012735", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.598, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-reasoning/6280d54e-2900-4ea9-bf3b-ce4025899eff.json b/data/vals_ai/xai/grok-4-1-fast-reasoning/6280d54e-2900-4ea9-bf3b-ce4025899eff.json new file mode 100644 index 0000000000000000000000000000000000000000..5e5bfb88966e71d03cfb9199fbabeca0944bc314 --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-reasoning/6280d54e-2900-4ea9-bf3b-ce4025899eff.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medcode/xai_grok-4-1-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedCode", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medcode", + "benchmark_name": "MedCode", + "benchmark_updated": "2026-07-08", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-reasoning", + "id": "xai/grok-4-1-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medcode:overall:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.medcode.overall", + "source_data": { + "dataset_name": "MedCode - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medcode", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedCode (Overall).", + "metric_id": "vals_ai.medcode.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode" + } + }, + "score_details": { + "score": 28.08, + "details": { + "benchmark_slug": "medcode", + "benchmark_name": "MedCode", + "benchmark_updated": "2026-07-08", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "28.08", + "raw_stderr": "1.992", + "latency": "46.5", + "cost_per_test": "0.002108", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.992, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-reasoning/771ebafe-2678-4cff-ac40-859d595fdb9f.json b/data/vals_ai/xai/grok-4-1-fast-reasoning/771ebafe-2678-4cff-ac40-859d595fdb9f.json new file mode 100644 index 0000000000000000000000000000000000000000..a1ab55f966db5af11329c98298dc214af0fca4e8 --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-reasoning/771ebafe-2678-4cff-ac40-859d595fdb9f.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/xai_grok-4-1-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-reasoning", + "id": "xai/grok-4-1-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 0.5, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0.5", + "raw_stderr": "0.456", + "latency": "1364.023", + "cost_per_test": "0.01707", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.456, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 5.667, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "5.667", + "raw_stderr": "3.469", + "latency": "1702.552", + "cost_per_test": "0.042869", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.469, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 3.084, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "3.084", + "raw_stderr": "1.749", + "latency": "1533.288", + "cost_per_test": "0.02997", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.749, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-reasoning/87f277eb-fbe5-43d0-b570-d9115b914ec9.json b/data/vals_ai/xai/grok-4-1-fast-reasoning/87f277eb-fbe5-43d0-b570-d9115b914ec9.json new file mode 100644 index 0000000000000000000000000000000000000000..9c0c561020873973998463336c3748cfb643edd8 --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-reasoning/87f277eb-fbe5-43d0-b570-d9115b914ec9.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/aime/xai_grok-4-1-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - AIME", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "dataset_type": "public", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-reasoning", + "id": "xai/grok-4-1-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aime:aime_2024:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.aime.aime_2024", + "source_data": { + "dataset_name": "AIME - AIME 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/aime" + ], + "additional_details": { + "benchmark_slug": "aime", + "task_key": "aime_2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for AIME (AIME 2024).", + "metric_id": "vals_ai.aime.aime_2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "score_details": { + "score": 93.333, + "details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "task_key": "aime_2024", + "task_name": "AIME 2024", + "dataset_type": "public", + "industry": "math", + "raw_score": "93.333", + "raw_stderr": "0.891", + "latency": "83.322", + "cost_per_test": "0.00373", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.891, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "aime:aime_2025:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.aime.aime_2025", + "source_data": { + "dataset_name": "AIME - AIME 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/aime" + ], + "additional_details": { + "benchmark_slug": "aime", + "task_key": "aime_2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for AIME (AIME 2025).", + "metric_id": "vals_ai.aime.aime_2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "score_details": { + "score": 90.417, + "details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "task_key": "aime_2025", + "task_name": "AIME 2025", + "dataset_type": "public", + "industry": "math", + "raw_score": "90.417", + "raw_stderr": "0.755", + "latency": "100.673", + "cost_per_test": "0.004458", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.755, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "aime:overall:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.aime.overall", + "source_data": { + "dataset_name": "AIME - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/aime" + ], + "additional_details": { + "benchmark_slug": "aime", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for AIME (Overall).", + "metric_id": "vals_ai.aime.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "score_details": { + "score": 91.875, + "details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "math", + "raw_score": "91.875", + "raw_stderr": "0.584", + "latency": "91.997", + "cost_per_test": "0.004094", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.584, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-reasoning/8b5dbaca-8fb9-4b18-9803-cc9706d99e8f.json b/data/vals_ai/xai/grok-4-1-fast-reasoning/8b5dbaca-8fb9-4b18-9803-cc9706d99e8f.json new file mode 100644 index 0000000000000000000000000000000000000000..601a5a2cf793f2241db80b7952a75877f67c1c37 --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-reasoning/8b5dbaca-8fb9-4b18-9803-cc9706d99e8f.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/xai_grok-4-1-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-reasoning", + "id": "xai/grok-4-1-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 4.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "4", + "raw_stderr": "1.96", + "latency": "1421.041", + "cost_per_test": "0.096578", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.96, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-reasoning/9342a7e3-e60c-4984-b808-9e220821a54e.json b/data/vals_ai/xai/grok-4-1-fast-reasoning/9342a7e3-e60c-4984-b808-9e220821a54e.json new file mode 100644 index 0000000000000000000000000000000000000000..0a3123688768400917dee855036475dc3b35e03e --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-reasoning/9342a7e3-e60c-4984-b808-9e220821a54e.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/public-benefits-bench/xai_grok-4-1-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Public Benefits Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "social mobility", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-reasoning", + "id": "xai/grok-4-1-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "public-benefits-bench:overall:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.public-benefits-bench.overall", + "source_data": { + "dataset_name": "Public Benefits Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Public Benefits Bench v1.1 (Overall).", + "metric_id": "vals_ai.public-benefits-bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "score_details": { + "score": 44.79, + "details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "social mobility", + "raw_score": "44.79", + "raw_stderr": "1.293", + "latency": "256.569", + "cost_per_test": "0.031221", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.293, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-reasoning/a5c39c85-d598-4c6e-b50e-003008a491da.json b/data/vals_ai/xai/grok-4-1-fast-reasoning/a5c39c85-d598-4c6e-b50e-003008a491da.json new file mode 100644 index 0000000000000000000000000000000000000000..d2f9c2eea70194ffb3b7db3081aa58881598c952 --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-reasoning/a5c39c85-d598-4c6e-b50e-003008a491da.json @@ -0,0 +1,505 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medqa/xai_grok-4-1-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedQA", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "dataset_type": "public", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-reasoning", + "id": "xai/grok-4-1-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medqa:asian:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.medqa.asian", + "source_data": { + "dataset_name": "MedQA - Asian", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "asian", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Asian).", + "metric_id": "vals_ai.medqa.asian.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 91.6, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "asian", + "task_name": "Asian", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "91.6", + "raw_stderr": "0.62", + "latency": "24.336", + "cost_per_test": "0.000591", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.62, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:black:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.medqa.black", + "source_data": { + "dataset_name": "MedQA - Black", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "black", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Black).", + "metric_id": "vals_ai.medqa.black.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 92.3, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "black", + "task_name": "Black", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "92.3", + "raw_stderr": "0.596", + "latency": "14.992", + "cost_per_test": "0.000589", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.596, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:hispanic:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.medqa.hispanic", + "source_data": { + "dataset_name": "MedQA - Hispanic", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "hispanic", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Hispanic).", + "metric_id": "vals_ai.medqa.hispanic.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 91.75, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "hispanic", + "task_name": "Hispanic", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "91.75", + "raw_stderr": "0.615", + "latency": "13.495", + "cost_per_test": "0.000578", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.615, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:indigenous:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.medqa.indigenous", + "source_data": { + "dataset_name": "MedQA - Indigenous", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "indigenous", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Indigenous).", + "metric_id": "vals_ai.medqa.indigenous.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 91.55, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "indigenous", + "task_name": "Indigenous", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "91.55", + "raw_stderr": "0.622", + "latency": "14.818", + "cost_per_test": "0.000582", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.622, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:overall:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.medqa.overall", + "source_data": { + "dataset_name": "MedQA - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Overall).", + "metric_id": "vals_ai.medqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 92.083, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "92.083", + "raw_stderr": "0.246", + "latency": "17.636", + "cost_per_test": "0.000571", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.246, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:unbiased:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.medqa.unbiased", + "source_data": { + "dataset_name": "MedQA - Unbiased", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "unbiased", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Unbiased).", + "metric_id": "vals_ai.medqa.unbiased.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 92.75, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "unbiased", + "task_name": "Unbiased", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "92.75", + "raw_stderr": "0.58", + "latency": "14.838", + "cost_per_test": "0.000534", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.58, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:white:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.medqa.white", + "source_data": { + "dataset_name": "MedQA - White", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "white", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (White).", + "metric_id": "vals_ai.medqa.white.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 92.55, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "white", + "task_name": "White", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "92.55", + "raw_stderr": "0.587", + "latency": "23.339", + "cost_per_test": "0.000552", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.587, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-reasoning/ae0917b9-3edc-40ec-8d17-f0f2f0bdd915.json b/data/vals_ai/xai/grok-4-1-fast-reasoning/ae0917b9-3edc-40ec-8d17-f0f2f0bdd915.json new file mode 100644 index 0000000000000000000000000000000000000000..5b0a0b46d7412f253cddb0216104fb6b3ee4545e --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-reasoning/ae0917b9-3edc-40ec-8d17-f0f2f0bdd915.json @@ -0,0 +1,356 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/public-benefits-bench-v1/xai_grok-4-1-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Public Benefits Bench v1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "public-benefits-bench-v1", + "benchmark_name": "Public Benefits Bench v1", + "benchmark_updated": "2026-06-09", + "dataset_type": "private", + "industry": "social mobility", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench-v1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-reasoning", + "id": "xai/grok-4-1-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "public-benefits-bench-v1:both:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.public-benefits-bench-v1.both", + "source_data": { + "dataset_name": "Public Benefits Bench v1 - Both", + "source_type": "other", + "additional_details": { + "benchmark_slug": "public-benefits-bench-v1", + "task_key": "both", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench-v1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Public Benefits Bench v1 (Both).", + "metric_id": "vals_ai.public-benefits-bench-v1.both.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench-v1" + } + }, + "score_details": { + "score": 44.317, + "details": { + "benchmark_slug": "public-benefits-bench-v1", + "benchmark_name": "Public Benefits Bench v1", + "benchmark_updated": "2026-06-09", + "task_key": "both", + "task_name": "Both", + "dataset_type": "private", + "industry": "social mobility", + "raw_score": "44.317", + "raw_stderr": "1.292", + "latency": "185.115", + "cost_per_test": "0.064873", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.292, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "public-benefits-bench-v1:multi_turn:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.public-benefits-bench-v1.multi_turn", + "source_data": { + "dataset_name": "Public Benefits Bench v1 - Multi Turn", + "source_type": "other", + "additional_details": { + "benchmark_slug": "public-benefits-bench-v1", + "task_key": "multi_turn", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench-v1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Public Benefits Bench v1 (Multi Turn).", + "metric_id": "vals_ai.public-benefits-bench-v1.multi_turn.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench-v1" + } + }, + "score_details": { + "score": 41.678, + "details": { + "benchmark_slug": "public-benefits-bench-v1", + "benchmark_name": "Public Benefits Bench v1", + "benchmark_updated": "2026-06-09", + "task_key": "multi_turn", + "task_name": "Multi Turn", + "dataset_type": "private", + "industry": "social mobility", + "raw_score": "41.678", + "raw_stderr": "1.282", + "latency": "110.415", + "cost_per_test": "0.005668", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.282, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "public-benefits-bench-v1:neither:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.public-benefits-bench-v1.neither", + "source_data": { + "dataset_name": "Public Benefits Bench v1 - Neither", + "source_type": "other", + "additional_details": { + "benchmark_slug": "public-benefits-bench-v1", + "task_key": "neither", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench-v1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Public Benefits Bench v1 (Neither).", + "metric_id": "vals_ai.public-benefits-bench-v1.neither.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench-v1" + } + }, + "score_details": { + "score": 33.018, + "details": { + "benchmark_slug": "public-benefits-bench-v1", + "benchmark_name": "Public Benefits Bench v1", + "benchmark_updated": "2026-06-09", + "task_key": "neither", + "task_name": "Neither", + "dataset_type": "private", + "industry": "social mobility", + "raw_score": "33.018", + "raw_stderr": "1.223", + "latency": "15.402", + "cost_per_test": "0.001004", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.223, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "public-benefits-bench-v1:overall:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.public-benefits-bench-v1.overall", + "source_data": { + "dataset_name": "Public Benefits Bench v1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "public-benefits-bench-v1", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench-v1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Public Benefits Bench v1 (Overall).", + "metric_id": "vals_ai.public-benefits-bench-v1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench-v1" + } + }, + "score_details": { + "score": 44.317, + "details": { + "benchmark_slug": "public-benefits-bench-v1", + "benchmark_name": "Public Benefits Bench v1", + "benchmark_updated": "2026-06-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "social mobility", + "raw_score": "44.317", + "raw_stderr": "1.292", + "latency": "185.115", + "cost_per_test": "0.064873", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.292, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "public-benefits-bench-v1:web_search:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.public-benefits-bench-v1.web_search", + "source_data": { + "dataset_name": "Public Benefits Bench v1 - Web Search", + "source_type": "other", + "additional_details": { + "benchmark_slug": "public-benefits-bench-v1", + "task_key": "web_search", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench-v1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Public Benefits Bench v1 (Web Search).", + "metric_id": "vals_ai.public-benefits-bench-v1.web_search.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench-v1" + } + }, + "score_details": { + "score": 33.085, + "details": { + "benchmark_slug": "public-benefits-bench-v1", + "benchmark_name": "Public Benefits Bench v1", + "benchmark_updated": "2026-06-09", + "task_key": "web_search", + "task_name": "Web Search", + "dataset_type": "private", + "industry": "social mobility", + "raw_score": "33.085", + "raw_stderr": "1.224", + "latency": "31.551", + "cost_per_test": "0.004028", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.224, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-reasoning/b23e28e2-7fe4-4c31-97fe-a5e375251fbf.json b/data/vals_ai/xai/grok-4-1-fast-reasoning/b23e28e2-7fe4-4c31-97fe-a5e375251fbf.json new file mode 100644 index 0000000000000000000000000000000000000000..dcd2f67ac7f8a6a354a81543ade50b2a4e25817d --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-reasoning/b23e28e2-7fe4-4c31-97fe-a5e375251fbf.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/xai_grok-4-1-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-reasoning", + "id": "xai/grok-4-1-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.758, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.758", + "raw_stderr": "0.617", + "latency": "18.116", + "cost_per_test": "0.000707", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.617, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 58.571, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "58.571", + "raw_stderr": "2.633", + "latency": "209.035", + "cost_per_test": "0.008272", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.633, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 84.595, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "84.595", + "raw_stderr": "1.845", + "latency": "83.197", + "cost_per_test": "0.002696", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.845, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 80.641, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "80.641", + "raw_stderr": "1.091", + "latency": "103.449", + "cost_per_test": "0.003892", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.091, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-reasoning/b409428b-8c93-4b61-943e-42f8466f88fe.json b/data/vals_ai/xai/grok-4-1-fast-reasoning/b409428b-8c93-4b61-943e-42f8466f88fe.json new file mode 100644 index 0000000000000000000000000000000000000000..d5660bc94dd604f4cc26a1b35070b2cbd00faeff --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-reasoning/b409428b-8c93-4b61-943e-42f8466f88fe.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/xai_grok-4-1-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-reasoning", + "id": "xai/grok-4-1-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 55.564, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "55.564", + "raw_stderr": "1.401", + "latency": "72.456", + "cost_per_test": "0.00324", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.401, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 42.607, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "42.607", + "raw_stderr": "0.952", + "latency": "47.45", + "cost_per_test": "0.002438", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.952, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 29.65, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "29.65", + "raw_stderr": "1.288", + "latency": "22.445", + "cost_per_test": "0.001637", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.288, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-reasoning/bbbd75b3-eff7-4979-a38a-fef065380d19.json b/data/vals_ai/xai/grok-4-1-fast-reasoning/bbbd75b3-eff7-4979-a38a-fef065380d19.json new file mode 100644 index 0000000000000000000000000000000000000000..dd7a392bad0f9f5f62d03174d971e2f97459f71f --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-reasoning/bbbd75b3-eff7-4979-a38a-fef065380d19.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/xai_grok-4-1-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-reasoning", + "id": "xai/grok-4-1-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 61.652, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.652", + "raw_stderr": "1.39", + "latency": "30.758", + "cost_per_test": "0.001333", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.39, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 73.14, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "73.14", + "raw_stderr": "0.865", + "latency": "30.758", + "cost_per_test": "0.001333", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.865, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 84.628, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "84.628", + "raw_stderr": "1.031", + "latency": "30.758", + "cost_per_test": "0.001333", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.031, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-reasoning/c8007da9-a665-4977-9151-154cc6d7d137.json b/data/vals_ai/xai/grok-4-1-fast-reasoning/c8007da9-a665-4977-9151-154cc6d7d137.json new file mode 100644 index 0000000000000000000000000000000000000000..d408e9729781db4b85814ae4c38a37b5a5f4a4ee --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-reasoning/c8007da9-a665-4977-9151-154cc6d7d137.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/xai_grok-4-1-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-reasoning", + "id": "xai/grok-4-1-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 78.732, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "78.732", + "raw_stderr": "1.866", + "latency": "39.292", + "cost_per_test": "0.002387", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.866, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-reasoning/ceca3d95-346c-435f-973d-7919746cf72c.json b/data/vals_ai/xai/grok-4-1-fast-reasoning/ceca3d95-346c-435f-973d-7919746cf72c.json new file mode 100644 index 0000000000000000000000000000000000000000..91e0cdd080a42382427ef46b11174a98aacbc8ad --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-reasoning/ceca3d95-346c-435f-973d-7919746cf72c.json @@ -0,0 +1,356 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/xai_grok-4-1-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-reasoning", + "id": "xai/grok-4-1-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 30.159, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "30.159", + "raw_stderr": "5.759", + "latency": "121.01", + "cost_per_test": "0.001074", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.759, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 12.302, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "12.302", + "raw_stderr": "4.174", + "latency": "61.786", + "cost_per_test": "0.001806", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.174, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "33.333", + "raw_stderr": "6.225", + "latency": "81.528", + "cost_per_test": "0.011822", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.225, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 51.533, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "51.533", + "raw_stderr": "5.614", + "latency": "48.364", + "cost_per_test": "0.001538", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.614, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 31.331, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "31.331", + "raw_stderr": "3.02", + "latency": "77.053", + "cost_per_test": "0.001473", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.02, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-reasoning/d02ccc06-1ff4-4663-9a10-670644740688.json b/data/vals_ai/xai/grok-4-1-fast-reasoning/d02ccc06-1ff4-4663-9a10-670644740688.json new file mode 100644 index 0000000000000000000000000000000000000000..8d13538e1555c7e46de28cd8ac9a120fa4f98fef --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-reasoning/d02ccc06-1ff4-4663-9a10-670644740688.json @@ -0,0 +1,103 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/xai_grok-4-1-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-reasoning", + "id": "xai/grok-4-1-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 72.659, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.659", + "raw_stderr": "1.072", + "latency": "96.354", + "cost_per_test": "0.00056", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.072, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-reasoning/d07e71e8-77ad-4022-9866-c47265bab84f.json b/data/vals_ai/xai/grok-4-1-fast-reasoning/d07e71e8-77ad-4022-9866-c47265bab84f.json new file mode 100644 index 0000000000000000000000000000000000000000..20f193a492dd3cdfbb2ffbdc08f95af6175d00ef --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-reasoning/d07e71e8-77ad-4022-9866-c47265bab84f.json @@ -0,0 +1,371 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/xai_grok-4-1-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-reasoning", + "id": "xai/grok-4-1-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 7.143, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "7.143", + "raw_stderr": "4.022", + "latency": "174.975", + "cost_per_test": "0.035708", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.022, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.716, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.716", + "raw_stderr": "2.932", + "latency": "149.895", + "cost_per_test": "0.030252", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.932, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 59.794, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "59.794", + "raw_stderr": "3.529", + "latency": "115.167", + "cost_per_test": "0.022111", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.529, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "132.713", + "cost_per_test": "0.024384", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 41.4, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "41.4", + "raw_stderr": "2.205", + "latency": "138.424", + "cost_per_test": "0.027516", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.205, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-reasoning/dbd29342-31d1-4d4e-9541-32cd897aedc1.json b/data/vals_ai/xai/grok-4-1-fast-reasoning/dbd29342-31d1-4d4e-9541-32cd897aedc1.json new file mode 100644 index 0000000000000000000000000000000000000000..18d0b05568ec604410173ad3e07b87cac3902563 --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-reasoning/dbd29342-31d1-4d4e-9541-32cd897aedc1.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/xai_grok-4-1-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-reasoning", + "id": "xai/grok-4-1-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 1.2, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "1.2", + "raw_stderr": "1.2", + "latency": "527.562", + "cost_per_test": "0.209444", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.2, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-reasoning/e9860324-5f7b-4b82-9373-a8c984034a11.json b/data/vals_ai/xai/grok-4-1-fast-reasoning/e9860324-5f7b-4b82-9373-a8c984034a11.json new file mode 100644 index 0000000000000000000000000000000000000000..3ad462448c31104090d1f5baedc7e163e6eddafa --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-reasoning/e9860324-5f7b-4b82-9373-a8c984034a11.json @@ -0,0 +1,438 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/xai_grok-4-1-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-reasoning", + "id": "xai/grok-4-1-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 86.698, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.698", + "raw_stderr": "0.711", + "latency": "8.025", + "cost_per_test": "0.000348", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.711, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.296, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.296", + "raw_stderr": "0.19", + "latency": "11.467", + "cost_per_test": "0.000427", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.19, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 89.256, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.256", + "raw_stderr": "0.583", + "latency": "7.239", + "cost_per_test": "0.000422", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.583, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.451, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.451", + "raw_stderr": "0.457", + "latency": "8.579", + "cost_per_test": "0.000384", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.457, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.152, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.152", + "raw_stderr": "1.212", + "latency": "7.275", + "cost_per_test": "0.00033", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.212, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 75.853, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "75.853", + "raw_stderr": "1.691", + "latency": "8.888", + "cost_per_test": "0.000393", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.691, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-1-fast-reasoning/f9d95204-4432-4e1b-99b7-dc355ffc0352.json b/data/vals_ai/xai/grok-4-1-fast-reasoning/f9d95204-4432-4e1b-99b7-dc355ffc0352.json new file mode 100644 index 0000000000000000000000000000000000000000..c2b5aa751969dd7dc9345d436d92e7f7c8a70656 --- /dev/null +++ b/data/vals_ai/xai/grok-4-1-fast-reasoning/f9d95204-4432-4e1b-99b7-dc355ffc0352.json @@ -0,0 +1,840 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mgsm/xai_grok-4-1-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MGSM", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "dataset_type": "public", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-1-fast-reasoning", + "id": "xai/grok-4-1-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-1-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mgsm:mgsm_bn:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_bn", + "source_data": { + "dataset_name": "MGSM - Bengali", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_bn", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Bengali).", + "metric_id": "vals_ai.mgsm.mgsm_bn.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 87.6, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_bn", + "task_name": "Bengali", + "dataset_type": "public", + "industry": "math", + "raw_score": "87.6", + "raw_stderr": "2.084", + "latency": "20.59", + "cost_per_test": "0.000735", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.084, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_de:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_de", + "source_data": { + "dataset_name": "MGSM - German", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_de", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (German).", + "metric_id": "vals_ai.mgsm.mgsm_de.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 90.4, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_de", + "task_name": "German", + "dataset_type": "public", + "industry": "math", + "raw_score": "90.4", + "raw_stderr": "1.863", + "latency": "11.03", + "cost_per_test": "0.000448", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.863, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_en:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_en", + "source_data": { + "dataset_name": "MGSM - English", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_en", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (English).", + "metric_id": "vals_ai.mgsm.mgsm_en.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 95.2, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_en", + "task_name": "English", + "dataset_type": "public", + "industry": "math", + "raw_score": "95.2", + "raw_stderr": "1.352", + "latency": "6.779", + "cost_per_test": "0.000298", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.352, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_es:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_es", + "source_data": { + "dataset_name": "MGSM - Spanish", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_es", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Spanish).", + "metric_id": "vals_ai.mgsm.mgsm_es.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 86.0, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_es", + "task_name": "Spanish", + "dataset_type": "public", + "industry": "math", + "raw_score": "86", + "raw_stderr": "2.194", + "latency": "9.599", + "cost_per_test": "0.00039", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.194, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_fr:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_fr", + "source_data": { + "dataset_name": "MGSM - French", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_fr", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (French).", + "metric_id": "vals_ai.mgsm.mgsm_fr.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 82.0, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_fr", + "task_name": "French", + "dataset_type": "public", + "industry": "math", + "raw_score": "82", + "raw_stderr": "2.43", + "latency": "11.581", + "cost_per_test": "0.000474", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.43, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_ja:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_ja", + "source_data": { + "dataset_name": "MGSM - Japanese", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_ja", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Japanese).", + "metric_id": "vals_ai.mgsm.mgsm_ja.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 87.2, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_ja", + "task_name": "Japanese", + "dataset_type": "public", + "industry": "math", + "raw_score": "87.2", + "raw_stderr": "2.113", + "latency": "10.958", + "cost_per_test": "0.000432", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.113, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_ru:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_ru", + "source_data": { + "dataset_name": "MGSM - Russian", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_ru", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Russian).", + "metric_id": "vals_ai.mgsm.mgsm_ru.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 93.6, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_ru", + "task_name": "Russian", + "dataset_type": "public", + "industry": "math", + "raw_score": "93.6", + "raw_stderr": "1.548", + "latency": "10.677", + "cost_per_test": "0.000401", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.548, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_sw:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_sw", + "source_data": { + "dataset_name": "MGSM - Swahili", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_sw", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Swahili).", + "metric_id": "vals_ai.mgsm.mgsm_sw.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 93.2, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_sw", + "task_name": "Swahili", + "dataset_type": "public", + "industry": "math", + "raw_score": "93.2", + "raw_stderr": "1.592", + "latency": "17.427", + "cost_per_test": "0.00066", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.592, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_te:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_te", + "source_data": { + "dataset_name": "MGSM - Telugu", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_te", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Telugu).", + "metric_id": "vals_ai.mgsm.mgsm_te.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 85.2, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_te", + "task_name": "Telugu", + "dataset_type": "public", + "industry": "math", + "raw_score": "85.2", + "raw_stderr": "2.246", + "latency": "27.775", + "cost_per_test": "0.001001", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.246, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_th:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_th", + "source_data": { + "dataset_name": "MGSM - Thai", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_th", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Thai).", + "metric_id": "vals_ai.mgsm.mgsm_th.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 93.2, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_th", + "task_name": "Thai", + "dataset_type": "public", + "industry": "math", + "raw_score": "93.2", + "raw_stderr": "1.592", + "latency": "13.831", + "cost_per_test": "0.000533", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.592, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_zh:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_zh", + "source_data": { + "dataset_name": "MGSM - Chinese", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_zh", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Chinese).", + "metric_id": "vals_ai.mgsm.mgsm_zh.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 91.2, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_zh", + "task_name": "Chinese", + "dataset_type": "public", + "industry": "math", + "raw_score": "91.2", + "raw_stderr": "1.792", + "latency": "8.268", + "cost_per_test": "0.000321", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.792, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:overall:grok/grok-4-1-fast-reasoning:score", + "evaluation_name": "vals_ai.mgsm.overall", + "source_data": { + "dataset_name": "MGSM - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Overall).", + "metric_id": "vals_ai.mgsm.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 89.527, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "math", + "raw_score": "89.527", + "raw_stderr": "1.892", + "latency": "13.501", + "cost_per_test": "0.000518", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.892, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-non-reasoning/3c894fbd-94de-4e4c-aa60-e48726f4eb8f.json b/data/vals_ai/xai/grok-4-fast-non-reasoning/3c894fbd-94de-4e4c-aa60-e48726f4eb8f.json new file mode 100644 index 0000000000000000000000000000000000000000..830e9d4f41cbe07d08b49827fd45c7938dfce3fd --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-non-reasoning/3c894fbd-94de-4e4c-aa60-e48726f4eb8f.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/xai_grok-4-fast-non-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-non-reasoning", + "id": "xai/grok-4-fast-non-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-non-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 64.802, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.802", + "raw_stderr": "1.63", + "latency": "6.835", + "cost_per_test": "0.001137", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.63, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 54.079, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "54.079", + "raw_stderr": "1.701", + "latency": "13.932", + "cost_per_test": "0.07579", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.701, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 58.392, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.392", + "raw_stderr": "0.967", + "latency": "9.37", + "cost_per_test": "0.039974", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.967, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 56.294, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.294", + "raw_stderr": "1.693", + "latency": "7.342", + "cost_per_test": "0.042996", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.693, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-non-reasoning/4f65e1c4-f861-4adc-ab16-ece734fed8a2.json b/data/vals_ai/xai/grok-4-fast-non-reasoning/4f65e1c4-f861-4adc-ab16-ece734fed8a2.json new file mode 100644 index 0000000000000000000000000000000000000000..62de8398995061e9b0ae5e488a26da84bc02f2aa --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-non-reasoning/4f65e1c4-f861-4adc-ab16-ece734fed8a2.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/xai_grok-4-fast-non-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-non-reasoning", + "id": "xai/grok-4-fast-non-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-non-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 67.172, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "67.172", + "raw_stderr": "3.337", + "latency": "4.507", + "cost_per_test": "0.001771", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.337, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 57.071, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "57.071", + "raw_stderr": "3.531", + "latency": "2.828", + "cost_per_test": "0.000426", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.531, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 62.122, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "62.122", + "raw_stderr": "2.429", + "latency": "3.668", + "cost_per_test": "0.001098", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.429, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-non-reasoning/4fb3b662-991d-45d0-bc0d-3063cf0ddd94.json b/data/vals_ai/xai/grok-4-fast-non-reasoning/4fb3b662-991d-45d0-bc0d-3063cf0ddd94.json new file mode 100644 index 0000000000000000000000000000000000000000..7e391580a496299327adecd13e292580d8ca4dc4 --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-non-reasoning/4fb3b662-991d-45d0-bc0d-3063cf0ddd94.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/xai_grok-4-fast-non-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-non-reasoning", + "id": "xai/grok-4-fast-non-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-non-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 6.167, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "6.167", + "raw_stderr": "5.629", + "latency": "307.059", + "cost_per_test": "0.173256", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.629, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 1.5, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "1.5", + "raw_stderr": "0.874", + "latency": "848.205", + "cost_per_test": "0.080245", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.874, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 3.834, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "3.834", + "raw_stderr": "2.848", + "latency": "577.632", + "cost_per_test": "0.12675", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.848, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-non-reasoning/60353131-d2bc-4cd2-8438-bc3ba33cf9e2.json b/data/vals_ai/xai/grok-4-fast-non-reasoning/60353131-d2bc-4cd2-8438-bc3ba33cf9e2.json new file mode 100644 index 0000000000000000000000000000000000000000..13c3bfc89d6cff6691099adb4161cf32e93014e9 --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-non-reasoning/60353131-d2bc-4cd2-8438-bc3ba33cf9e2.json @@ -0,0 +1,840 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mgsm/xai_grok-4-fast-non-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MGSM", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "dataset_type": "public", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-non-reasoning", + "id": "xai/grok-4-fast-non-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-non-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mgsm:mgsm_bn:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_bn", + "source_data": { + "dataset_name": "MGSM - Bengali", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_bn", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Bengali).", + "metric_id": "vals_ai.mgsm.mgsm_bn.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 78.4, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_bn", + "task_name": "Bengali", + "dataset_type": "public", + "industry": "math", + "raw_score": "78.4", + "raw_stderr": "2.603", + "latency": "2.409", + "cost_per_test": "0.000531", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.603, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_de:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_de", + "source_data": { + "dataset_name": "MGSM - German", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_de", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (German).", + "metric_id": "vals_ai.mgsm.mgsm_de.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 91.2, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_de", + "task_name": "German", + "dataset_type": "public", + "industry": "math", + "raw_score": "91.2", + "raw_stderr": "1.792", + "latency": "1.441", + "cost_per_test": "0.0003", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.792, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_en:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_en", + "source_data": { + "dataset_name": "MGSM - English", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_en", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (English).", + "metric_id": "vals_ai.mgsm.mgsm_en.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 95.2, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_en", + "task_name": "English", + "dataset_type": "public", + "industry": "math", + "raw_score": "95.2", + "raw_stderr": "1.352", + "latency": "1.452", + "cost_per_test": "0.000228", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.352, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_es:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_es", + "source_data": { + "dataset_name": "MGSM - Spanish", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_es", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Spanish).", + "metric_id": "vals_ai.mgsm.mgsm_es.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 92.0, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_es", + "task_name": "Spanish", + "dataset_type": "public", + "industry": "math", + "raw_score": "92", + "raw_stderr": "1.716", + "latency": "1.783", + "cost_per_test": "0.00028", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.716, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_fr:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_fr", + "source_data": { + "dataset_name": "MGSM - French", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_fr", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (French).", + "metric_id": "vals_ai.mgsm.mgsm_fr.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 83.6, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_fr", + "task_name": "French", + "dataset_type": "public", + "industry": "math", + "raw_score": "83.6", + "raw_stderr": "2.342", + "latency": "1.27", + "cost_per_test": "0.000279", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.342, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_ja:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_ja", + "source_data": { + "dataset_name": "MGSM - Japanese", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_ja", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Japanese).", + "metric_id": "vals_ai.mgsm.mgsm_ja.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 84.8, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_ja", + "task_name": "Japanese", + "dataset_type": "public", + "industry": "math", + "raw_score": "84.8", + "raw_stderr": "2.271", + "latency": "2.867", + "cost_per_test": "0.000284", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.271, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_ru:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_ru", + "source_data": { + "dataset_name": "MGSM - Russian", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_ru", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Russian).", + "metric_id": "vals_ai.mgsm.mgsm_ru.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 92.8, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_ru", + "task_name": "Russian", + "dataset_type": "public", + "industry": "math", + "raw_score": "92.8", + "raw_stderr": "1.635", + "latency": "1.34", + "cost_per_test": "0.000286", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.635, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_sw:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_sw", + "source_data": { + "dataset_name": "MGSM - Swahili", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_sw", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Swahili).", + "metric_id": "vals_ai.mgsm.mgsm_sw.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 84.0, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_sw", + "task_name": "Swahili", + "dataset_type": "public", + "industry": "math", + "raw_score": "84", + "raw_stderr": "2.319", + "latency": "1.626", + "cost_per_test": "0.000321", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.319, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_te:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_te", + "source_data": { + "dataset_name": "MGSM - Telugu", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_te", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Telugu).", + "metric_id": "vals_ai.mgsm.mgsm_te.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 85.2, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_te", + "task_name": "Telugu", + "dataset_type": "public", + "industry": "math", + "raw_score": "85.2", + "raw_stderr": "2.246", + "latency": "2.29", + "cost_per_test": "0.0006", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.246, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_th:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_th", + "source_data": { + "dataset_name": "MGSM - Thai", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_th", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Thai).", + "metric_id": "vals_ai.mgsm.mgsm_th.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 91.6, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_th", + "task_name": "Thai", + "dataset_type": "public", + "industry": "math", + "raw_score": "91.6", + "raw_stderr": "1.754", + "latency": "2.243", + "cost_per_test": "0.00036", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.754, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_zh:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_zh", + "source_data": { + "dataset_name": "MGSM - Chinese", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_zh", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Chinese).", + "metric_id": "vals_ai.mgsm.mgsm_zh.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 89.2, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_zh", + "task_name": "Chinese", + "dataset_type": "public", + "industry": "math", + "raw_score": "89.2", + "raw_stderr": "1.963", + "latency": "1.201", + "cost_per_test": "0.00025", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.963, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:overall:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mgsm.overall", + "source_data": { + "dataset_name": "MGSM - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Overall).", + "metric_id": "vals_ai.mgsm.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 88.0, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "math", + "raw_score": "88", + "raw_stderr": "1.999", + "latency": "1.811", + "cost_per_test": "0.000338", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.999, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-non-reasoning/68fbd378-6bb8-4ea4-b24f-c4feb56cec8f.json b/data/vals_ai/xai/grok-4-fast-non-reasoning/68fbd378-6bb8-4ea4-b24f-c4feb56cec8f.json new file mode 100644 index 0000000000000000000000000000000000000000..dfe2108a40065595f4b7c933414046685ed4a6f6 --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-non-reasoning/68fbd378-6bb8-4ea4-b24f-c4feb56cec8f.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/xai_grok-4-fast-non-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-non-reasoning", + "id": "xai/grok-4-fast-non-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-non-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 79.722, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "79.722", + "raw_stderr": "1.871", + "latency": "8.482", + "cost_per_test": "0.002056", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.871, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-non-reasoning/75b60c4a-b964-4ba8-8f09-66004720ad47.json b/data/vals_ai/xai/grok-4-fast-non-reasoning/75b60c4a-b964-4ba8-8f09-66004720ad47.json new file mode 100644 index 0000000000000000000000000000000000000000..85b7ceb0e92d7d04c6fe9d239df6516d12014a8a --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-non-reasoning/75b60c4a-b964-4ba8-8f09-66004720ad47.json @@ -0,0 +1,432 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/xai_grok-4-fast-non-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-non-reasoning", + "id": "xai/grok-4-fast-non-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-non-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.733, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.733", + "raw_stderr": "0.963", + "latency": "0.766", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.963, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.668, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.668", + "raw_stderr": "0.2", + "latency": "1.146", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.2, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.671, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.671", + "raw_stderr": "0.693", + "latency": "1.002", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.693, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.396, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.396", + "raw_stderr": "0.472", + "latency": "0.933", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.472, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.446, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.446", + "raw_stderr": "0.892", + "latency": "0.559", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.892, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 70.464, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "70.464", + "raw_stderr": "1.822", + "latency": "1.19", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.822, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-non-reasoning/78caca5e-58a7-4511-afa1-9862f2d1c4c5.json b/data/vals_ai/xai/grok-4-fast-non-reasoning/78caca5e-58a7-4511-afa1-9862f2d1c4c5.json new file mode 100644 index 0000000000000000000000000000000000000000..41f5e9d4e2c8fb0635202026e0c7c6f838443792 --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-non-reasoning/78caca5e-58a7-4511-afa1-9862f2d1c4c5.json @@ -0,0 +1,505 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medqa/xai_grok-4-fast-non-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedQA", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "dataset_type": "public", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-non-reasoning", + "id": "xai/grok-4-fast-non-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-non-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medqa:asian:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.medqa.asian", + "source_data": { + "dataset_name": "MedQA - Asian", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "asian", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Asian).", + "metric_id": "vals_ai.medqa.asian.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 75.05, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "asian", + "task_name": "Asian", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "75.05", + "raw_stderr": "0.968", + "latency": "2.377", + "cost_per_test": "0.000482", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.968, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:black:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.medqa.black", + "source_data": { + "dataset_name": "MedQA - Black", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "black", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Black).", + "metric_id": "vals_ai.medqa.black.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 74.75, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "black", + "task_name": "Black", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "74.75", + "raw_stderr": "0.972", + "latency": "2.398", + "cost_per_test": "0.000484", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.972, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:hispanic:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.medqa.hispanic", + "source_data": { + "dataset_name": "MedQA - Hispanic", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "hispanic", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Hispanic).", + "metric_id": "vals_ai.medqa.hispanic.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 75.25, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "hispanic", + "task_name": "Hispanic", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "75.25", + "raw_stderr": "0.965", + "latency": "2.396", + "cost_per_test": "0.000483", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.965, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:indigenous:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.medqa.indigenous", + "source_data": { + "dataset_name": "MedQA - Indigenous", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "indigenous", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Indigenous).", + "metric_id": "vals_ai.medqa.indigenous.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 74.6, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "indigenous", + "task_name": "Indigenous", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "74.6", + "raw_stderr": "0.973", + "latency": "2.483", + "cost_per_test": "0.000485", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.973, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:overall:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.medqa.overall", + "source_data": { + "dataset_name": "MedQA - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Overall).", + "metric_id": "vals_ai.medqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 75.358, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "75.358", + "raw_stderr": "0.393", + "latency": "2.416", + "cost_per_test": "0.000481", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.393, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:unbiased:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.medqa.unbiased", + "source_data": { + "dataset_name": "MedQA - Unbiased", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "unbiased", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Unbiased).", + "metric_id": "vals_ai.medqa.unbiased.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 77.1, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "unbiased", + "task_name": "Unbiased", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "77.1", + "raw_stderr": "0.94", + "latency": "2.283", + "cost_per_test": "0.000467", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.94, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:white:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.medqa.white", + "source_data": { + "dataset_name": "MedQA - White", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "white", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (White).", + "metric_id": "vals_ai.medqa.white.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 75.4, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "white", + "task_name": "White", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "75.4", + "raw_stderr": "0.963", + "latency": "2.555", + "cost_per_test": "0.000484", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.963, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-non-reasoning/7ac58cc4-19c7-4997-8160-dad4f801856b.json b/data/vals_ai/xai/grok-4-fast-non-reasoning/7ac58cc4-19c7-4997-8160-dad4f801856b.json new file mode 100644 index 0000000000000000000000000000000000000000..0129db2415c78124954ce68a0f277b420467187a --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-non-reasoning/7ac58cc4-19c7-4997-8160-dad4f801856b.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/xai_grok-4-fast-non-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-non-reasoning", + "id": "xai/grok-4-fast-non-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-non-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.798, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.798", + "raw_stderr": "1.341", + "latency": "1.597", + "cost_per_test": "0.001092", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.341, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 72.877, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.877", + "raw_stderr": "1.583", + "latency": "2.094", + "cost_per_test": "0.000895", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.583, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 70.406, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.406", + "raw_stderr": "1.357", + "latency": "1.581", + "cost_per_test": "0.001271", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.357, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 67.64, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "67.64", + "raw_stderr": "2.308", + "latency": "1.945", + "cost_per_test": "0.001211", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.308, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.318, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.318", + "raw_stderr": "1.418", + "latency": "2.074", + "cost_per_test": "0.001013", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.418, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 58.308, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "58.308", + "raw_stderr": "1.584", + "latency": "2.258", + "cost_per_test": "0.001118", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.584, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 67.237, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "67.237", + "raw_stderr": "1.641", + "latency": "1.761", + "cost_per_test": "0.00089", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.641, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 63.089, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "63.089", + "raw_stderr": "2.469", + "latency": "2.238", + "cost_per_test": "0.001623", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.469, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 57.856, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "57.856", + "raw_stderr": "1.488", + "latency": "1.884", + "cost_per_test": "0.001188", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.488, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 67.876, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "67.876", + "raw_stderr": "1.27", + "latency": "1.575", + "cost_per_test": "0.001346", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.27, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 73.052, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.052", + "raw_stderr": "1.46", + "latency": "4.498", + "cost_per_test": "0.000764", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.46, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 70.342, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.342", + "raw_stderr": "0.431", + "latency": "2.135", + "cost_per_test": "0.001084", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.431, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 69.756, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.756", + "raw_stderr": "1.435", + "latency": "2.738", + "cost_per_test": "0.000827", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.435, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 75.751, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.751", + "raw_stderr": "1.189", + "latency": "2.071", + "cost_per_test": "0.001043", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.189, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.82, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.82", + "raw_stderr": "1.471", + "latency": "1.579", + "cost_per_test": "0.000896", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.471, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-non-reasoning/8044a4b7-fc6b-433a-b7a4-b5b3394571c0.json b/data/vals_ai/xai/grok-4-fast-non-reasoning/8044a4b7-fc6b-433a-b7a4-b5b3394571c0.json new file mode 100644 index 0000000000000000000000000000000000000000..36206136fdeac25d885bd3404e9f3d0201322358 --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-non-reasoning/8044a4b7-fc6b-433a-b7a4-b5b3394571c0.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medcode/xai_grok-4-fast-non-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedCode", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medcode", + "benchmark_name": "MedCode", + "benchmark_updated": "2026-07-08", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-non-reasoning", + "id": "xai/grok-4-fast-non-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-non-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medcode:overall:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.medcode.overall", + "source_data": { + "dataset_name": "MedCode - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medcode", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedCode (Overall).", + "metric_id": "vals_ai.medcode.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode" + } + }, + "score_details": { + "score": 30.036, + "details": { + "benchmark_slug": "medcode", + "benchmark_name": "MedCode", + "benchmark_updated": "2026-07-08", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "30.036", + "raw_stderr": "1.974", + "latency": "18.211", + "cost_per_test": "0.002149", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.974, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-non-reasoning/83594b67-2a7a-464f-b2c8-ce6942b0a63c.json b/data/vals_ai/xai/grok-4-fast-non-reasoning/83594b67-2a7a-464f-b2c8-ce6942b0a63c.json new file mode 100644 index 0000000000000000000000000000000000000000..2a8f89a3e9549bad6cee0adc602cc6604cd971cc --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-non-reasoning/83594b67-2a7a-464f-b2c8-ce6942b0a63c.json @@ -0,0 +1,356 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/xai_grok-4-fast-non-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-non-reasoning", + "id": "xai/grok-4-fast-non-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-non-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 12.698, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "12.698", + "raw_stderr": "4.178", + "latency": "5.622", + "cost_per_test": "0.006058", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.178, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 6.35, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "6.35", + "raw_stderr": "3.125", + "latency": "8.586", + "cost_per_test": "0.008768", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.125, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 27.16, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "27.16", + "raw_stderr": "5.909", + "latency": "21.745", + "cost_per_test": "0.022455", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.909, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 28.918, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "28.918", + "raw_stderr": "5.093", + "latency": "6.518", + "cost_per_test": "0.014611", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.093, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 15.989, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "15.989", + "raw_stderr": "2.43", + "latency": "6.909", + "cost_per_test": "0.009812", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.43, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-non-reasoning/a50e4125-4eae-4c44-aa1d-7a173a94ded7.json b/data/vals_ai/xai/grok-4-fast-non-reasoning/a50e4125-4eae-4c44-aa1d-7a173a94ded7.json new file mode 100644 index 0000000000000000000000000000000000000000..58dda1ce73c84798f9305bdda5656256ba47b2aa --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-non-reasoning/a50e4125-4eae-4c44-aa1d-7a173a94ded7.json @@ -0,0 +1,103 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/xai_grok-4-fast-non-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-non-reasoning", + "id": "xai/grok-4-fast-non-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-non-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 63.41, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "63.41", + "raw_stderr": "1.158", + "latency": "7.209", + "cost_per_test": "0.001555", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.158, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-non-reasoning/ac4125a2-8fa2-4302-9150-4ec9616cfab4.json b/data/vals_ai/xai/grok-4-fast-non-reasoning/ac4125a2-8fa2-4302-9150-4ec9616cfab4.json new file mode 100644 index 0000000000000000000000000000000000000000..88d3dffe7bdbe4e4097e76196fb2180b28a7931a --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-non-reasoning/ac4125a2-8fa2-4302-9150-4ec9616cfab4.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/xai_grok-4-fast-non-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-non-reasoning", + "id": "xai/grok-4-fast-non-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-non-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 88.199, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "88.199", + "raw_stderr": "1.798", + "latency": "2.443", + "cost_per_test": "0.00056", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.798, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 8.571, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "8.571", + "raw_stderr": "1.496", + "latency": "14.244", + "cost_per_test": "0.001507", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.496, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 41.514, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "41.514", + "raw_stderr": "2.518", + "latency": "10.186", + "cost_per_test": "0.000968", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.518, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 46.095, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "46.095", + "raw_stderr": "1.146", + "latency": "8.958", + "cost_per_test": "0.001012", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.146, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-non-reasoning/f54a6d7a-cffd-4f48-b018-a46b0967ad3f.json b/data/vals_ai/xai/grok-4-fast-non-reasoning/f54a6d7a-cffd-4f48-b018-a46b0967ad3f.json new file mode 100644 index 0000000000000000000000000000000000000000..8d901c1e5fee51eb28b741f397065975f4d3262d --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-non-reasoning/f54a6d7a-cffd-4f48-b018-a46b0967ad3f.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/aime/xai_grok-4-fast-non-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - AIME", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "dataset_type": "public", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-non-reasoning", + "id": "xai/grok-4-fast-non-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-non-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aime:aime_2024:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.aime.aime_2024", + "source_data": { + "dataset_name": "AIME - AIME 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/aime" + ], + "additional_details": { + "benchmark_slug": "aime", + "task_key": "aime_2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for AIME (AIME 2024).", + "metric_id": "vals_ai.aime.aime_2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "score_details": { + "score": 20.833, + "details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "task_key": "aime_2024", + "task_name": "AIME 2024", + "dataset_type": "public", + "industry": "math", + "raw_score": "20.833", + "raw_stderr": "0.833", + "latency": "6.92", + "cost_per_test": "0.00097", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.833, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "aime:aime_2025:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.aime.aime_2025", + "source_data": { + "dataset_name": "AIME - AIME 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/aime" + ], + "additional_details": { + "benchmark_slug": "aime", + "task_key": "aime_2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for AIME (AIME 2025).", + "metric_id": "vals_ai.aime.aime_2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "score_details": { + "score": 45.833, + "details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "task_key": "aime_2025", + "task_name": "AIME 2025", + "dataset_type": "public", + "industry": "math", + "raw_score": "45.833", + "raw_stderr": "1.967", + "latency": "4.837", + "cost_per_test": "0.00064", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.967, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "aime:overall:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.aime.overall", + "source_data": { + "dataset_name": "AIME - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/aime" + ], + "additional_details": { + "benchmark_slug": "aime", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for AIME (Overall).", + "metric_id": "vals_ai.aime.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "math", + "raw_score": "33.333", + "raw_stderr": "1.068", + "latency": "5.879", + "cost_per_test": "0.000805", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.068, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-non-reasoning/fc5d17b2-4ef2-4062-a825-7ee070d80d6a.json b/data/vals_ai/xai/grok-4-fast-non-reasoning/fc5d17b2-4ef2-4062-a825-7ee070d80d6a.json new file mode 100644 index 0000000000000000000000000000000000000000..09f66b5a10ab1aa2de37439accaeab13ee344d22 --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-non-reasoning/fc5d17b2-4ef2-4062-a825-7ee070d80d6a.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/xai_grok-4-fast-non-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-non-reasoning", + "id": "xai/grok-4-fast-non-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-non-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 25.199, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "25.199", + "raw_stderr": "1.224", + "latency": "4.617", + "cost_per_test": "0.003464", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.224, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 25.318, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "25.318", + "raw_stderr": "0.867", + "latency": "4.508", + "cost_per_test": "0.003483", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.867, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 25.437, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "25.437", + "raw_stderr": "1.228", + "latency": "4.399", + "cost_per_test": "0.003502", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.228, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-non-reasoning/feaede51-39fb-4fc1-975c-2e4dfb7ba1eb.json b/data/vals_ai/xai/grok-4-fast-non-reasoning/feaede51-39fb-4fc1-975c-2e4dfb7ba1eb.json new file mode 100644 index 0000000000000000000000000000000000000000..921c5ae9aae0a8ce6b6084d9352f117fc2015654 --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-non-reasoning/feaede51-39fb-4fc1-975c-2e4dfb7ba1eb.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/xai_grok-4-fast-non-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-non-reasoning", + "id": "xai/grok-4-fast-non-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-non-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 59.559, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.559", + "raw_stderr": "1.249", + "latency": "4.25", + "cost_per_test": "0.000571", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.249, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 71.581, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.581", + "raw_stderr": "0.783", + "latency": "4.25", + "cost_per_test": "0.000571", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.783, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:grok/grok-4-fast-non-reasoning:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 83.603, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "83.603", + "raw_stderr": "0.943", + "latency": "4.25", + "cost_per_test": "0.000571", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.943, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-reasoning/1695366b-da76-4d51-9536-da3ed92d6cce.json b/data/vals_ai/xai/grok-4-fast-reasoning/1695366b-da76-4d51-9536-da3ed92d6cce.json new file mode 100644 index 0000000000000000000000000000000000000000..26a8bfc99624b06ace64b4491eeb10fce16ff1bb --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-reasoning/1695366b-da76-4d51-9536-da3ed92d6cce.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/xai_grok-4-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-reasoning", + "id": "xai/grok-4-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 68.244, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.244", + "raw_stderr": "1.185", + "latency": "10.184", + "cost_per_test": "0.001268", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.185, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 75.697, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "75.697", + "raw_stderr": "0.76", + "latency": "10.184", + "cost_per_test": "0.001268", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.76, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 83.15, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "83.15", + "raw_stderr": "0.953", + "latency": "10.184", + "cost_per_test": "0.001268", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.953, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-reasoning/40f6a8f5-3602-47b6-91b2-e2ff1cf203ff.json b/data/vals_ai/xai/grok-4-fast-reasoning/40f6a8f5-3602-47b6-91b2-e2ff1cf203ff.json new file mode 100644 index 0000000000000000000000000000000000000000..f26f3ffa0785ac3b56ef5fac6e9736fb0b70aa29 --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-reasoning/40f6a8f5-3602-47b6-91b2-e2ff1cf203ff.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/xai_grok-4-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-reasoning", + "id": "xai/grok-4-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.656, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.656", + "raw_stderr": "1.087", + "latency": "4.835", + "cost_per_test": "0.001417", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.087, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 69.328, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "69.328", + "raw_stderr": "1.642", + "latency": "7.248", + "cost_per_test": "0.001573", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.642, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.855, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.855", + "raw_stderr": "1.254", + "latency": "10.218", + "cost_per_test": "0.002341", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.254, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.244, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.244", + "raw_stderr": "1.966", + "latency": "7.272", + "cost_per_test": "0.001847", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.966, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.242, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.242", + "raw_stderr": "1.254", + "latency": "4.536", + "cost_per_test": "0.001322", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.254, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.083, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.083", + "raw_stderr": "1.283", + "latency": "22.173", + "cost_per_test": "0.00358", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.283, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.572, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.572", + "raw_stderr": "1.523", + "latency": "5.016", + "cost_per_test": "0.001232", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.523, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.016, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.016", + "raw_stderr": "2.247", + "latency": "4.846", + "cost_per_test": "0.001969", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.247, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 68.483, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "68.483", + "raw_stderr": "1.4", + "latency": "8.045", + "cost_per_test": "0.001762", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.4, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.601, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.601", + "raw_stderr": "0.865", + "latency": "7.219", + "cost_per_test": "0.002105", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.865, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.654, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.654", + "raw_stderr": "1.324", + "latency": "4.164", + "cost_per_test": "0.001093", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.324, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.704, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.704", + "raw_stderr": "0.393", + "latency": "7.343", + "cost_per_test": "0.001742", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.393, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.958, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.958", + "raw_stderr": "1.825", + "latency": "3.994", + "cost_per_test": "0.001108", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.825, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.836, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.836", + "raw_stderr": "0.938", + "latency": "9.383", + "cost_per_test": "0.001892", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.938, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.333, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.333", + "raw_stderr": "1.319", + "latency": "3.847", + "cost_per_test": "0.001148", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.319, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-reasoning/419d20e3-36ed-4052-8d84-9a0258c9f952.json b/data/vals_ai/xai/grok-4-fast-reasoning/419d20e3-36ed-4052-8d84-9a0258c9f952.json new file mode 100644 index 0000000000000000000000000000000000000000..9c697537fab0ef786d5d85639fd21d55db9084fd --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-reasoning/419d20e3-36ed-4052-8d84-9a0258c9f952.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/xai_grok-4-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-reasoning", + "id": "xai/grok-4-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1448.855", + "cost_per_test": "0.133206", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-reasoning/43ca9b03-d1f7-49be-b37f-d59a44671960.json b/data/vals_ai/xai/grok-4-fast-reasoning/43ca9b03-d1f7-49be-b37f-d59a44671960.json new file mode 100644 index 0000000000000000000000000000000000000000..08de73955d8b352ef7371349886eae4e1a987075 --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-reasoning/43ca9b03-d1f7-49be-b37f-d59a44671960.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/aime/xai_grok-4-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - AIME", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "dataset_type": "public", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-reasoning", + "id": "xai/grok-4-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aime:aime_2024:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.aime.aime_2024", + "source_data": { + "dataset_name": "AIME - AIME 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/aime" + ], + "additional_details": { + "benchmark_slug": "aime", + "task_key": "aime_2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for AIME (AIME 2024).", + "metric_id": "vals_ai.aime.aime_2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "score_details": { + "score": 94.167, + "details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "task_key": "aime_2024", + "task_name": "AIME 2024", + "dataset_type": "public", + "industry": "math", + "raw_score": "94.167", + "raw_stderr": "0.833", + "latency": "31.664", + "cost_per_test": "0.003984", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.833, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "aime:aime_2025:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.aime.aime_2025", + "source_data": { + "dataset_name": "AIME - AIME 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/aime" + ], + "additional_details": { + "benchmark_slug": "aime", + "task_key": "aime_2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for AIME (AIME 2025).", + "metric_id": "vals_ai.aime.aime_2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "score_details": { + "score": 88.333, + "details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "task_key": "aime_2025", + "task_name": "AIME 2025", + "dataset_type": "public", + "industry": "math", + "raw_score": "88.333", + "raw_stderr": "1.89", + "latency": "37.454", + "cost_per_test": "0.00465", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.89, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "aime:overall:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.aime.overall", + "source_data": { + "dataset_name": "AIME - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/aime" + ], + "additional_details": { + "benchmark_slug": "aime", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for AIME (Overall).", + "metric_id": "vals_ai.aime.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "score_details": { + "score": 91.25, + "details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "math", + "raw_score": "91.25", + "raw_stderr": "1.033", + "latency": "34.559", + "cost_per_test": "0.004317", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.033, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-reasoning/4f7e599d-c5c0-42c7-a8b5-ae7d84a17590.json b/data/vals_ai/xai/grok-4-fast-reasoning/4f7e599d-c5c0-42c7-a8b5-ae7d84a17590.json new file mode 100644 index 0000000000000000000000000000000000000000..e87d9ec610e53ecdce572bfa5ca6d5ea1aeca0b7 --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-reasoning/4f7e599d-c5c0-42c7-a8b5-ae7d84a17590.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/xai_grok-4-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-reasoning", + "id": "xai/grok-4-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.137, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.137", + "raw_stderr": "0.754", + "latency": "6.12", + "cost_per_test": "0.000925", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.754, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 53.143, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "53.143", + "raw_stderr": "2.667", + "latency": "119.467", + "cost_per_test": "0.009458", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.667, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 85.64, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "85.64", + "raw_stderr": "1.792", + "latency": "27.834", + "cost_per_test": "0.003139", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.792, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 78.973, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "78.973", + "raw_stderr": "1.1", + "latency": "51.14", + "cost_per_test": "0.004507", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.1, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-reasoning/5c0cd6a8-c561-4e05-bffb-d35ee36e255c.json b/data/vals_ai/xai/grok-4-fast-reasoning/5c0cd6a8-c561-4e05-bffb-d35ee36e255c.json new file mode 100644 index 0000000000000000000000000000000000000000..d72a1ebf1e070d9a5e8c74b3e4fc717e9662a74d --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-reasoning/5c0cd6a8-c561-4e05-bffb-d35ee36e255c.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medcode/xai_grok-4-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedCode", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medcode", + "benchmark_name": "MedCode", + "benchmark_updated": "2026-07-08", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-reasoning", + "id": "xai/grok-4-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medcode:overall:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.medcode.overall", + "source_data": { + "dataset_name": "MedCode - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medcode", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedCode (Overall).", + "metric_id": "vals_ai.medcode.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode" + } + }, + "score_details": { + "score": 37.385, + "details": { + "benchmark_slug": "medcode", + "benchmark_name": "MedCode", + "benchmark_updated": "2026-07-08", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "37.385", + "raw_stderr": "1.941", + "latency": "17.515", + "cost_per_test": "0.002143", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.941, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-reasoning/60542dbc-c6a3-4bc5-a26b-e29472e54160.json b/data/vals_ai/xai/grok-4-fast-reasoning/60542dbc-c6a3-4bc5-a26b-e29472e54160.json new file mode 100644 index 0000000000000000000000000000000000000000..e0448c762979aa61fbc17729ef65efb78371f596 --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-reasoning/60542dbc-c6a3-4bc5-a26b-e29472e54160.json @@ -0,0 +1,505 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medqa/xai_grok-4-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedQA", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "dataset_type": "public", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-reasoning", + "id": "xai/grok-4-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medqa:asian:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.medqa.asian", + "source_data": { + "dataset_name": "MedQA - Asian", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "asian", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Asian).", + "metric_id": "vals_ai.medqa.asian.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 91.8, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "asian", + "task_name": "Asian", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "91.8", + "raw_stderr": "0.614", + "latency": "5.623", + "cost_per_test": "0.000755", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.614, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:black:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.medqa.black", + "source_data": { + "dataset_name": "MedQA - Black", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "black", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Black).", + "metric_id": "vals_ai.medqa.black.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 91.9, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "black", + "task_name": "Black", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "91.9", + "raw_stderr": "0.61", + "latency": "5.464", + "cost_per_test": "0.000762", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.61, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:hispanic:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.medqa.hispanic", + "source_data": { + "dataset_name": "MedQA - Hispanic", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "hispanic", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Hispanic).", + "metric_id": "vals_ai.medqa.hispanic.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 91.95, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "hispanic", + "task_name": "Hispanic", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "91.95", + "raw_stderr": "0.608", + "latency": "5.329", + "cost_per_test": "0.000746", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.608, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:indigenous:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.medqa.indigenous", + "source_data": { + "dataset_name": "MedQA - Indigenous", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "indigenous", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Indigenous).", + "metric_id": "vals_ai.medqa.indigenous.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 92.0, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "indigenous", + "task_name": "Indigenous", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "92", + "raw_stderr": "0.607", + "latency": "5.407", + "cost_per_test": "0.000752", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.607, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:overall:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.medqa.overall", + "source_data": { + "dataset_name": "MedQA - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Overall).", + "metric_id": "vals_ai.medqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 92.067, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "92.067", + "raw_stderr": "0.247", + "latency": "5.343", + "cost_per_test": "0.000744", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.247, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:unbiased:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.medqa.unbiased", + "source_data": { + "dataset_name": "MedQA - Unbiased", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "unbiased", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Unbiased).", + "metric_id": "vals_ai.medqa.unbiased.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 92.6, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "unbiased", + "task_name": "Unbiased", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "92.6", + "raw_stderr": "0.585", + "latency": "5.017", + "cost_per_test": "0.000714", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.585, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:white:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.medqa.white", + "source_data": { + "dataset_name": "MedQA - White", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "white", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (White).", + "metric_id": "vals_ai.medqa.white.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 92.15, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "white", + "task_name": "White", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "92.15", + "raw_stderr": "0.601", + "latency": "5.22", + "cost_per_test": "0.000735", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.601, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-reasoning/61154c90-9ab7-4ca7-8a27-22bb64cd61ce.json b/data/vals_ai/xai/grok-4-fast-reasoning/61154c90-9ab7-4ca7-8a27-22bb64cd61ce.json new file mode 100644 index 0000000000000000000000000000000000000000..b3a3ea52057510c2940c6b5f6352d1f76eb765a4 --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-reasoning/61154c90-9ab7-4ca7-8a27-22bb64cd61ce.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/xai_grok-4-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-reasoning", + "id": "xai/grok-4-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 81.632, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "81.632", + "raw_stderr": "2.137", + "latency": "12.423", + "cost_per_test": "0.002535", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.137, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-reasoning/62642b33-913c-4e9f-ad2e-529556c86d90.json b/data/vals_ai/xai/grok-4-fast-reasoning/62642b33-913c-4e9f-ad2e-529556c86d90.json new file mode 100644 index 0000000000000000000000000000000000000000..9b9c5b0b65c95e356c1c1afd7ec0e5b99b57dbdb --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-reasoning/62642b33-913c-4e9f-ad2e-529556c86d90.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/xai_grok-4-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-reasoning", + "id": "xai/grok-4-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 9.333, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "9.333", + "raw_stderr": "4.305", + "latency": "1370.235", + "cost_per_test": "1.353118", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.305, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 13.667, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "13.667", + "raw_stderr": "7.647", + "latency": "1446.674", + "cost_per_test": "0.595122", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.647, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 11.5, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "11.5", + "raw_stderr": "4.388", + "latency": "1408.454", + "cost_per_test": "0.97412", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.388, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-reasoning/6d7fb51d-65e4-43a7-a251-ab19d0ec3b41.json b/data/vals_ai/xai/grok-4-fast-reasoning/6d7fb51d-65e4-43a7-a251-ab19d0ec3b41.json new file mode 100644 index 0000000000000000000000000000000000000000..572b3ca8f0a474872dd63b3863f2999d02f1825c --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-reasoning/6d7fb51d-65e4-43a7-a251-ab19d0ec3b41.json @@ -0,0 +1,371 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/xai_grok-4-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-reasoning", + "id": "xai/grok-4-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 7.143, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "7.143", + "raw_stderr": "4.022", + "latency": "154.031", + "cost_per_test": "0.063596", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.022, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 36.782, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "36.782", + "raw_stderr": "2.991", + "latency": "139.014", + "cost_per_test": "0.057395", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.991, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 65.979, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "65.979", + "raw_stderr": "3.41", + "latency": "102.033", + "cost_per_test": "0.039325", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.41, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "119.67", + "cost_per_test": "0.035087", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 45.4, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "45.4", + "raw_stderr": "2.229", + "latency": "125.811", + "cost_per_test": "0.050771", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.229, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-reasoning/8b3e5ea5-73c9-4406-bb42-d9e769750152.json b/data/vals_ai/xai/grok-4-fast-reasoning/8b3e5ea5-73c9-4406-bb42-d9e769750152.json new file mode 100644 index 0000000000000000000000000000000000000000..f64280d184a74564d99f25e2d87d2a32e339fec4 --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-reasoning/8b3e5ea5-73c9-4406-bb42-d9e769750152.json @@ -0,0 +1,840 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mgsm/xai_grok-4-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MGSM", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "dataset_type": "public", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-reasoning", + "id": "xai/grok-4-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mgsm:mgsm_bn:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_bn", + "source_data": { + "dataset_name": "MGSM - Bengali", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_bn", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Bengali).", + "metric_id": "vals_ai.mgsm.mgsm_bn.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 86.4, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_bn", + "task_name": "Bengali", + "dataset_type": "public", + "industry": "math", + "raw_score": "86.4", + "raw_stderr": "2.168", + "latency": "5.596", + "cost_per_test": "0.000906", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.168, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_de:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_de", + "source_data": { + "dataset_name": "MGSM - German", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_de", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (German).", + "metric_id": "vals_ai.mgsm.mgsm_de.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 92.0, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_de", + "task_name": "German", + "dataset_type": "public", + "industry": "math", + "raw_score": "92", + "raw_stderr": "1.716", + "latency": "3.262", + "cost_per_test": "0.000556", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.716, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_en:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_en", + "source_data": { + "dataset_name": "MGSM - English", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_en", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (English).", + "metric_id": "vals_ai.mgsm.mgsm_en.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 95.6, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_en", + "task_name": "English", + "dataset_type": "public", + "industry": "math", + "raw_score": "95.6", + "raw_stderr": "1.297", + "latency": "1.957", + "cost_per_test": "0.000377", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.297, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_es:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_es", + "source_data": { + "dataset_name": "MGSM - Spanish", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_es", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Spanish).", + "metric_id": "vals_ai.mgsm.mgsm_es.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 92.0, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_es", + "task_name": "Spanish", + "dataset_type": "public", + "industry": "math", + "raw_score": "92", + "raw_stderr": "1.716", + "latency": "2.974", + "cost_per_test": "0.000482", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.716, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_fr:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_fr", + "source_data": { + "dataset_name": "MGSM - French", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_fr", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (French).", + "metric_id": "vals_ai.mgsm.mgsm_fr.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 83.6, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_fr", + "task_name": "French", + "dataset_type": "public", + "industry": "math", + "raw_score": "83.6", + "raw_stderr": "2.342", + "latency": "3.377", + "cost_per_test": "0.000546", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.342, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_ja:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_ja", + "source_data": { + "dataset_name": "MGSM - Japanese", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_ja", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Japanese).", + "metric_id": "vals_ai.mgsm.mgsm_ja.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 89.2, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_ja", + "task_name": "Japanese", + "dataset_type": "public", + "industry": "math", + "raw_score": "89.2", + "raw_stderr": "1.963", + "latency": "3.355", + "cost_per_test": "0.000519", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.963, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_ru:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_ru", + "source_data": { + "dataset_name": "MGSM - Russian", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_ru", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Russian).", + "metric_id": "vals_ai.mgsm.mgsm_ru.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 95.2, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_ru", + "task_name": "Russian", + "dataset_type": "public", + "industry": "math", + "raw_score": "95.2", + "raw_stderr": "1.352", + "latency": "2.841", + "cost_per_test": "0.00049", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.352, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_sw:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_sw", + "source_data": { + "dataset_name": "MGSM - Swahili", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_sw", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Swahili).", + "metric_id": "vals_ai.mgsm.mgsm_sw.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 90.8, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_sw", + "task_name": "Swahili", + "dataset_type": "public", + "industry": "math", + "raw_score": "90.8", + "raw_stderr": "1.828", + "latency": "5.003", + "cost_per_test": "0.000727", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.828, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_te:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_te", + "source_data": { + "dataset_name": "MGSM - Telugu", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_te", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Telugu).", + "metric_id": "vals_ai.mgsm.mgsm_te.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 88.8, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_te", + "task_name": "Telugu", + "dataset_type": "public", + "industry": "math", + "raw_score": "88.8", + "raw_stderr": "1.995", + "latency": "6.529", + "cost_per_test": "0.001155", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.995, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_th:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_th", + "source_data": { + "dataset_name": "MGSM - Thai", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_th", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Thai).", + "metric_id": "vals_ai.mgsm.mgsm_th.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 94.0, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_th", + "task_name": "Thai", + "dataset_type": "public", + "industry": "math", + "raw_score": "94", + "raw_stderr": "1.502", + "latency": "4.971", + "cost_per_test": "0.000859", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.502, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:mgsm_zh:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.mgsm.mgsm_zh", + "source_data": { + "dataset_name": "MGSM - Chinese", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "mgsm_zh", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Chinese).", + "metric_id": "vals_ai.mgsm.mgsm_zh.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 92.0, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "mgsm_zh", + "task_name": "Chinese", + "dataset_type": "public", + "industry": "math", + "raw_score": "92", + "raw_stderr": "1.716", + "latency": "2.517", + "cost_per_test": "0.000412", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.716, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mgsm:overall:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.mgsm.overall", + "source_data": { + "dataset_name": "MGSM - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mgsm" + ], + "additional_details": { + "benchmark_slug": "mgsm", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Overall).", + "metric_id": "vals_ai.mgsm.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm" + } + }, + "score_details": { + "score": 90.873, + "details": { + "benchmark_slug": "mgsm", + "benchmark_name": "MGSM", + "benchmark_updated": "2026-01-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "math", + "raw_score": "90.873", + "raw_stderr": "1.781", + "latency": "3.853", + "cost_per_test": "0.000639", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.781, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-reasoning/945237de-c3b7-4ece-ada7-db39b307ce37.json b/data/vals_ai/xai/grok-4-fast-reasoning/945237de-c3b7-4ece-ada7-db39b307ce37.json new file mode 100644 index 0000000000000000000000000000000000000000..b93ed7360fb839f595bb41727ea14d5f1e92dbb4 --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-reasoning/945237de-c3b7-4ece-ada7-db39b307ce37.json @@ -0,0 +1,432 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/xai_grok-4-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-reasoning", + "id": "xai/grok-4-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 86.798, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.798", + "raw_stderr": "0.797", + "latency": "2.854", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.797, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.873, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.873", + "raw_stderr": "0.188", + "latency": "4.848", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.188, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.657, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.657", + "raw_stderr": "0.593", + "latency": "4.673", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.593, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.601, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.601", + "raw_stderr": "0.453", + "latency": "3.803", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.453, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.265, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.265", + "raw_stderr": "0.837", + "latency": "2.779", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.837, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 66.411, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "66.411", + "raw_stderr": "1.844", + "latency": "3.861", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.844, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-reasoning/b679dcd9-a98e-4906-9f94-1a24f397928b.json b/data/vals_ai/xai/grok-4-fast-reasoning/b679dcd9-a98e-4906-9f94-1a24f397928b.json new file mode 100644 index 0000000000000000000000000000000000000000..6637b07d0c20e551fe46c416507153cd591e3ae4 --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-reasoning/b679dcd9-a98e-4906-9f94-1a24f397928b.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/xai_grok-4-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-reasoning", + "id": "xai/grok-4-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 84.343, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.343", + "raw_stderr": "2.583", + "latency": "22.395", + "cost_per_test": "0.003746", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.583, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 86.364, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.364", + "raw_stderr": "2.476", + "latency": "20.422", + "cost_per_test": "0.002352", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.476, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 85.354, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.354", + "raw_stderr": "1.789", + "latency": "21.408", + "cost_per_test": "0.003049", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.789, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-reasoning/c49a05bc-214e-440c-97e6-e1a0288ca2da.json b/data/vals_ai/xai/grok-4-fast-reasoning/c49a05bc-214e-440c-97e6-e1a0288ca2da.json new file mode 100644 index 0000000000000000000000000000000000000000..08a9d2da6b2bb324574f2a427a7a4eb9cb0f4a25 --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-reasoning/c49a05bc-214e-440c-97e6-e1a0288ca2da.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2/xai_grok-4-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.0", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2", + "benchmark_name": "Terminal-Bench 2.0", + "benchmark_updated": "2026-06-04", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-reasoning", + "id": "xai/grok-4-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2:easy:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.terminal-bench-2.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.0 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.0 (Easy).", + "metric_id": "vals_ai.terminal-bench-2.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "score_details": { + "score": 100.0, + "details": { + "benchmark_slug": "terminal-bench-2", + "benchmark_name": "Terminal-Bench 2.0", + "benchmark_updated": "2026-06-04", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "100", + "raw_stderr": "0", + "latency": "9595.719", + "cost_per_test": "0.7634", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2:hard:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.terminal-bench-2.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.0 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.0 (Hard).", + "metric_id": "vals_ai.terminal-bench-2.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "score_details": { + "score": 13.333, + "details": { + "benchmark_slug": "terminal-bench-2", + "benchmark_name": "Terminal-Bench 2.0", + "benchmark_updated": "2026-06-04", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "13.333", + "raw_stderr": "6.312", + "latency": "1279.429", + "cost_per_test": "0.101787", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.312, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2:medium:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.terminal-bench-2.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.0 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.0 (Medium).", + "metric_id": "vals_ai.terminal-bench-2.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "score_details": { + "score": 32.727, + "details": { + "benchmark_slug": "terminal-bench-2", + "benchmark_name": "Terminal-Bench 2.0", + "benchmark_updated": "2026-06-04", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "32.727", + "raw_stderr": "6.385", + "latency": "697.87", + "cost_per_test": "0.05552", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.385, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2:overall:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.terminal-bench-2.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.0 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.0 (Overall).", + "metric_id": "vals_ai.terminal-bench-2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "score_details": { + "score": 29.213, + "details": { + "benchmark_slug": "terminal-bench-2", + "benchmark_name": "Terminal-Bench 2.0", + "benchmark_updated": "2026-06-04", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "29.213", + "raw_stderr": "4.848", + "latency": "431.268", + "cost_per_test": "0.03431", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.848, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-reasoning/c7899eec-0fe7-41c6-a028-e4decd89e9ee.json b/data/vals_ai/xai/grok-4-fast-reasoning/c7899eec-0fe7-41c6-a028-e4decd89e9ee.json new file mode 100644 index 0000000000000000000000000000000000000000..1eb68dab8d4fe901a9255754b9eb93a6608de7fa --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-reasoning/c7899eec-0fe7-41c6-a028-e4decd89e9ee.json @@ -0,0 +1,356 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/xai_grok-4-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-reasoning", + "id": "xai/grok-4-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 23.81, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "23.81", + "raw_stderr": "5.345", + "latency": "39.547", + "cost_per_test": "0.008286", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.345, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 17.064, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "17.064", + "raw_stderr": "4.821", + "latency": "32.868", + "cost_per_test": "0.010474", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.821, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 24.691, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "24.691", + "raw_stderr": "5.822", + "latency": "42.558", + "cost_per_test": "0.043415", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.822, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 48.409, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "48.409", + "raw_stderr": "5.613", + "latency": "22.788", + "cost_per_test": "0.015989", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.613, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 29.761, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "29.761", + "raw_stderr": "3.043", + "latency": "31.734", + "cost_per_test": "0.011583", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.043, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-reasoning/cf904b72-22b0-4f10-a4b3-319602a9b8da.json b/data/vals_ai/xai/grok-4-fast-reasoning/cf904b72-22b0-4f10-a4b3-319602a9b8da.json new file mode 100644 index 0000000000000000000000000000000000000000..1c72247973943a91bbc030e9c28b2d7053c99ab6 --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-reasoning/cf904b72-22b0-4f10-a4b3-319602a9b8da.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/xai_grok-4-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-reasoning", + "id": "xai/grok-4-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 67.249, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.249", + "raw_stderr": "1.602", + "latency": "8.663", + "cost_per_test": "0.001343", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.602, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 62.354, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.354", + "raw_stderr": "1.654", + "latency": "16.413", + "cost_per_test": "0.076169", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.654, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 66.9, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.9", + "raw_stderr": "0.925", + "latency": "11.839", + "cost_per_test": "0.040289", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.925, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 71.096, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.096", + "raw_stderr": "1.548", + "latency": "10.442", + "cost_per_test": "0.043356", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.548, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-reasoning/d86a93cc-304f-470d-bed1-dc2806947349.json b/data/vals_ai/xai/grok-4-fast-reasoning/d86a93cc-304f-470d-bed1-dc2806947349.json new file mode 100644 index 0000000000000000000000000000000000000000..dcb14668f1c35946d13e90095c893d9ceb716998 --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-reasoning/d86a93cc-304f-470d-bed1-dc2806947349.json @@ -0,0 +1,103 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/xai_grok-4-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-reasoning", + "id": "xai/grok-4-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 72.775, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.775", + "raw_stderr": "1.07", + "latency": "19.482", + "cost_per_test": "0.00288", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.07, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-reasoning/f4d12a07-d22c-45ec-9585-575f9886198c.json b/data/vals_ai/xai/grok-4-fast-reasoning/f4d12a07-d22c-45ec-9585-575f9886198c.json new file mode 100644 index 0000000000000000000000000000000000000000..061348d63fdae881f1238dcef9e1787cbabcc14c --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-reasoning/f4d12a07-d22c-45ec-9585-575f9886198c.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/case_law_v2/xai_grok-4-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CaseLaw v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "case_law_v2", + "benchmark_name": "CaseLaw v2", + "benchmark_updated": "2026-05-04", + "dataset_type": "private", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/case_law_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-reasoning", + "id": "xai/grok-4-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "case_law_v2:overall:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.case_law_v2.overall", + "source_data": { + "dataset_name": "CaseLaw v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "case_law_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/case_law_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CaseLaw v2 (Overall).", + "metric_id": "vals_ai.case_law_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/case_law_v2" + } + }, + "score_details": { + "score": 65.702, + "details": { + "benchmark_slug": "case_law_v2", + "benchmark_name": "CaseLaw v2", + "benchmark_updated": "2026-05-04", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "legal", + "raw_score": "65.702", + "raw_stderr": "0.429", + "latency": "23.709", + "cost_per_test": "0.007292", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.429, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4-fast-reasoning/fd243316-2ad7-49fa-bb63-f25cbc4d2758.json b/data/vals_ai/xai/grok-4-fast-reasoning/fd243316-2ad7-49fa-bb63-f25cbc4d2758.json new file mode 100644 index 0000000000000000000000000000000000000000..83846205dfe87c40fc5ea622c48e310b80a448cc --- /dev/null +++ b/data/vals_ai/xai/grok-4-fast-reasoning/fd243316-2ad7-49fa-bb63-f25cbc4d2758.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/xai_grok-4-fast-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4-fast-reasoning", + "id": "xai/grok-4-fast-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4-fast-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 55.087, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "55.087", + "raw_stderr": "1.402", + "latency": "20.319", + "cost_per_test": "0.005216", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.402, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 42.09, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "42.09", + "raw_stderr": "0.95", + "latency": "14.475", + "cost_per_test": "0.004554", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.95, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:grok/grok-4-fast-reasoning:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 29.094, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "29.094", + "raw_stderr": "1.281", + "latency": "8.631", + "cost_per_test": "0.003893", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.281, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.20-0309-reasoning/0687d87d-5fb6-4920-b789-5f8a848ee06c.json b/data/vals_ai/xai/grok-4.20-0309-reasoning/0687d87d-5fb6-4920-b789-5f8a848ee06c.json new file mode 100644 index 0000000000000000000000000000000000000000..5975d2c62a409f174121173cbe8cbf5e1102d7ae --- /dev/null +++ b/data/vals_ai/xai/grok-4.20-0309-reasoning/0687d87d-5fb6-4920-b789-5f8a848ee06c.json @@ -0,0 +1,371 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/xai_grok-4.20-0309-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.20-0309-reasoning", + "id": "xai/grok-4.20-0309-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.20-0309-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 21.429, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "21.429", + "raw_stderr": "6.408", + "latency": "215.953", + "cost_per_test": "0.878904", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.408, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 69.732, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "69.732", + "raw_stderr": "2.849", + "latency": "143.561", + "cost_per_test": "0.553798", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.849, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 87.113, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "87.113", + "raw_stderr": "2.412", + "latency": "104.34", + "cost_per_test": "0.380521", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.412, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "199.309", + "cost_per_test": "0.898914", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 72.2, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "72.2", + "raw_stderr": "2.006", + "latency": "134.759", + "cost_per_test": "0.515946", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.006, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.20-0309-reasoning/1bae6870-7f23-4a0c-849b-bed51cc8546c.json b/data/vals_ai/xai/grok-4.20-0309-reasoning/1bae6870-7f23-4a0c-849b-bed51cc8546c.json new file mode 100644 index 0000000000000000000000000000000000000000..328078750c3092dcd0f42a0451ab9dd60b6c5f5e --- /dev/null +++ b/data/vals_ai/xai/grok-4.20-0309-reasoning/1bae6870-7f23-4a0c-849b-bed51cc8546c.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/xai_grok-4.20-0309-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.20-0309-reasoning", + "id": "xai/grok-4.20-0309-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.20-0309-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 14.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "14", + "raw_stderr": "3.47", + "latency": "494.946", + "cost_per_test": "0.31815", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.47, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.20-0309-reasoning/27b55e23-ce3c-4abc-a133-3bac765de05a.json b/data/vals_ai/xai/grok-4.20-0309-reasoning/27b55e23-ce3c-4abc-a133-3bac765de05a.json new file mode 100644 index 0000000000000000000000000000000000000000..93d01696732a11c5ed5eeea9f9bd40d19709e350 --- /dev/null +++ b/data/vals_ai/xai/grok-4.20-0309-reasoning/27b55e23-ce3c-4abc-a133-3bac765de05a.json @@ -0,0 +1,536 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_multimodal_index/xai_grok-4.20-0309-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Multimodal Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.20-0309-reasoning", + "id": "xai/grok-4.20-0309-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.20-0309-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_multimodal_index:corp_fin_v2:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.vals_multimodal_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Multimodal Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (CorpFin v2).", + "metric_id": "vals_ai.vals_multimodal_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 65.851, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "65.851", + "raw_stderr": "1.619", + "latency": "9.418", + "cost_per_test": "0.130871", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.619, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:finance_agent:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.vals_multimodal_index.finance_agent", + "source_data": { + "dataset_name": "Vals Multimodal Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_multimodal_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 28.573, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "28.573", + "raw_stderr": "0.598", + "latency": "351.797", + "cost_per_test": "0.804911", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.598, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:mortgage_tax:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.vals_multimodal_index.mortgage_tax", + "source_data": { + "dataset_name": "Vals Multimodal Index - Mortgage Tax", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "mortgage_tax", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Mortgage Tax).", + "metric_id": "vals_ai.vals_multimodal_index.mortgage_tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 45.35, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "mortgage_tax", + "task_name": "Mortgage Tax", + "dataset_type": "private", + "industry": "index", + "raw_score": "45.35", + "raw_stderr": "0.988", + "latency": "11.124", + "cost_per_test": "0.024144", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.988, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:overall:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.vals_multimodal_index.overall", + "source_data": { + "dataset_name": "Vals Multimodal Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Overall).", + "metric_id": "vals_ai.vals_multimodal_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 39.059, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "39.059", + "raw_stderr": "0.701", + "latency": "171.06", + "cost_per_test": "0.398468", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.701, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:sage:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.vals_multimodal_index.sage", + "source_data": { + "dataset_name": "Vals Multimodal Index - SAGE", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "sage", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SAGE).", + "metric_id": "vals_ai.vals_multimodal_index.sage.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 38.242, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "sage", + "task_name": "SAGE", + "dataset_type": "private", + "industry": "index", + "raw_score": "38.242", + "raw_stderr": "3.425", + "latency": "66.358", + "cost_per_test": "0.063231", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.425, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:swebench:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.vals_multimodal_index.swebench", + "source_data": { + "dataset_name": "Vals Multimodal Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SWE-bench).", + "metric_id": "vals_ai.vals_multimodal_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "66.667", + "raw_stderr": "4.668", + "latency": "151.343", + "cost_per_test": "0.586871" + }, + "uncertainty": { + "standard_error": { + "value": 4.668, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:terminal_bench_2_1:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.vals_multimodal_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Multimodal Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_multimodal_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 44.195, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "44.195", + "raw_stderr": "0.991", + "latency": "305.726", + "cost_per_test": "0.407362", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.991, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:vibe_code_bench:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.vals_multimodal_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Multimodal Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_multimodal_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 1.515, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "1.515", + "raw_stderr": "1.515", + "latency": "301.651", + "cost_per_test": "0.771885", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.515, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.20-0309-reasoning/29957439-c29e-4bb7-8167-ba2919f4f588.json b/data/vals_ai/xai/grok-4.20-0309-reasoning/29957439-c29e-4bb7-8167-ba2919f4f588.json new file mode 100644 index 0000000000000000000000000000000000000000..7d2e9983af0840aba8e70dce97cff0c4a0b8323a --- /dev/null +++ b/data/vals_ai/xai/grok-4.20-0309-reasoning/29957439-c29e-4bb7-8167-ba2919f4f588.json @@ -0,0 +1,505 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medqa/xai_grok-4.20-0309-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedQA", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "dataset_type": "public", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.20-0309-reasoning", + "id": "xai/grok-4.20-0309-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.20-0309-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medqa:asian:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.medqa.asian", + "source_data": { + "dataset_name": "MedQA - Asian", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "asian", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Asian).", + "metric_id": "vals_ai.medqa.asian.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 94.3, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "asian", + "task_name": "Asian", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "94.3", + "raw_stderr": "0.518", + "latency": "19.327", + "cost_per_test": "0.009332", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.518, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:black:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.medqa.black", + "source_data": { + "dataset_name": "MedQA - Black", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "black", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Black).", + "metric_id": "vals_ai.medqa.black.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 94.5, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "black", + "task_name": "Black", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "94.5", + "raw_stderr": "0.51", + "latency": "8.869", + "cost_per_test": "0.009265", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.51, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:hispanic:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.medqa.hispanic", + "source_data": { + "dataset_name": "MedQA - Hispanic", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "hispanic", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Hispanic).", + "metric_id": "vals_ai.medqa.hispanic.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 94.1, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "hispanic", + "task_name": "Hispanic", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "94.1", + "raw_stderr": "0.527", + "latency": "8.629", + "cost_per_test": "0.009188", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.527, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:indigenous:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.medqa.indigenous", + "source_data": { + "dataset_name": "MedQA - Indigenous", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "indigenous", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Indigenous).", + "metric_id": "vals_ai.medqa.indigenous.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 94.65, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "indigenous", + "task_name": "Indigenous", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "94.65", + "raw_stderr": "0.503", + "latency": "8.736", + "cost_per_test": "0.009334", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.503, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:overall:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.medqa.overall", + "source_data": { + "dataset_name": "MedQA - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Overall).", + "metric_id": "vals_ai.medqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 94.55, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "94.55", + "raw_stderr": "0.207", + "latency": "12.225", + "cost_per_test": "0.009106", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.207, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:unbiased:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.medqa.unbiased", + "source_data": { + "dataset_name": "MedQA - Unbiased", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "unbiased", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (Unbiased).", + "metric_id": "vals_ai.medqa.unbiased.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 95.25, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "unbiased", + "task_name": "Unbiased", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "95.25", + "raw_stderr": "0.476", + "latency": "19.348", + "cost_per_test": "0.008533", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.476, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "medqa:white:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.medqa.white", + "source_data": { + "dataset_name": "MedQA - White", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/medqa" + ], + "additional_details": { + "benchmark_slug": "medqa", + "task_key": "white", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedQA (White).", + "metric_id": "vals_ai.medqa.white.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medqa" + } + }, + "score_details": { + "score": 94.5, + "details": { + "benchmark_slug": "medqa", + "benchmark_name": "MedQA", + "benchmark_updated": "2026-04-16", + "task_key": "white", + "task_name": "White", + "dataset_type": "public", + "industry": "healthcare", + "raw_score": "94.5", + "raw_stderr": "0.51", + "latency": "8.442", + "cost_per_test": "0.008982", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.51, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.20-0309-reasoning/2aee142e-758a-4e8f-ab59-6a75e8f6b8cf.json b/data/vals_ai/xai/grok-4.20-0309-reasoning/2aee142e-758a-4e8f-ab59-6a75e8f6b8cf.json new file mode 100644 index 0000000000000000000000000000000000000000..9b47e914a3257305352099e30f06e822933a3c1e --- /dev/null +++ b/data/vals_ai/xai/grok-4.20-0309-reasoning/2aee142e-758a-4e8f-ab59-6a75e8f6b8cf.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/xai_grok-4.20-0309-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.20-0309-reasoning", + "id": "xai/grok-4.20-0309-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.20-0309-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 4.063, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "4.063", + "raw_stderr": "2.058", + "latency": "301.651", + "cost_per_test": "0.771885", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.058, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.20-0309-reasoning/3f5294bb-dead-49df-9293-dc2986050a81.json b/data/vals_ai/xai/grok-4.20-0309-reasoning/3f5294bb-dead-49df-9293-dc2986050a81.json new file mode 100644 index 0000000000000000000000000000000000000000..414c7ec63e4a614d46bff7f2ddb5bde38c283fd3 --- /dev/null +++ b/data/vals_ai/xai/grok-4.20-0309-reasoning/3f5294bb-dead-49df-9293-dc2986050a81.json @@ -0,0 +1,408 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/xai_grok-4.20-0309-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.20-0309-reasoning", + "id": "xai/grok-4.20-0309-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.20-0309-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 65.851, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "65.851", + "raw_stderr": "1.619", + "latency": "9.418", + "cost_per_test": "0.130871", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.619, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 28.573, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "28.573", + "raw_stderr": "0.598", + "latency": "351.797", + "cost_per_test": "0.804911", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.598, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:overall:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 39.496, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "39.496", + "raw_stderr": "0.772", + "latency": "223.987", + "cost_per_test": "0.54038", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.772, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "66.667", + "raw_stderr": "4.668", + "latency": "151.343", + "cost_per_test": "0.586871" + }, + "uncertainty": { + "standard_error": { + "value": 4.668, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 44.195, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "44.195", + "raw_stderr": "0.991", + "latency": "305.726", + "cost_per_test": "0.407362", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.991, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 1.515, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "1.515", + "raw_stderr": "1.515", + "latency": "301.651", + "cost_per_test": "0.771885", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.515, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.20-0309-reasoning/476142bd-4ad7-480e-b5f0-47a16166dd47.json b/data/vals_ai/xai/grok-4.20-0309-reasoning/476142bd-4ad7-480e-b5f0-47a16166dd47.json new file mode 100644 index 0000000000000000000000000000000000000000..de82e5380839b932adc16df3c513ba60adab2567 --- /dev/null +++ b/data/vals_ai/xai/grok-4.20-0309-reasoning/476142bd-4ad7-480e-b5f0-47a16166dd47.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/xai_grok-4.20-0309-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.20-0309-reasoning", + "id": "xai/grok-4.20-0309-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.20-0309-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 63.054, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.054", + "raw_stderr": "1.648", + "latency": "3.81", + "cost_per_test": "0.008149", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.648, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 62.121, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.121", + "raw_stderr": "1.656", + "latency": "15.783", + "cost_per_test": "0.351139", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.656, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 63.675, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.675", + "raw_stderr": "0.947", + "latency": "9.67", + "cost_per_test": "0.163386", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.947, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 65.851, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.851", + "raw_stderr": "1.619", + "latency": "9.418", + "cost_per_test": "0.130871", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.619, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.20-0309-reasoning/63c3f756-23bc-4ff2-9d9e-0218482e9f04.json b/data/vals_ai/xai/grok-4.20-0309-reasoning/63c3f756-23bc-4ff2-9d9e-0218482e9f04.json new file mode 100644 index 0000000000000000000000000000000000000000..a382bb8792ad0118ac22a3fb30824ad64b70590d --- /dev/null +++ b/data/vals_ai/xai/grok-4.20-0309-reasoning/63c3f756-23bc-4ff2-9d9e-0218482e9f04.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/xai_grok-4.20-0309-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.20-0309-reasoning", + "id": "xai/grok-4.20-0309-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.20-0309-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 99.068, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "99.068", + "raw_stderr": "0.535", + "latency": "22.478", + "cost_per_test": "0.008132", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 66.0, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66", + "raw_stderr": "2.532", + "latency": "149.086", + "cost_per_test": "0.113578", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.532, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 87.728, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "87.728", + "raw_stderr": "1.677", + "latency": "56.706", + "cost_per_test": "0.035761", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.677, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 84.265, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "84.265", + "raw_stderr": "1.028", + "latency": "76.09", + "cost_per_test": "0.05249", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.028, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.20-0309-reasoning/6e4e207f-a71f-4075-9073-96f038350440.json b/data/vals_ai/xai/grok-4.20-0309-reasoning/6e4e207f-a71f-4075-9073-96f038350440.json new file mode 100644 index 0000000000000000000000000000000000000000..e41ba6aceacf0c82fd1ed86239599cf27720cbf1 --- /dev/null +++ b/data/vals_ai/xai/grok-4.20-0309-reasoning/6e4e207f-a71f-4075-9073-96f038350440.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medcode/xai_grok-4.20-0309-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedCode", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medcode", + "benchmark_name": "MedCode", + "benchmark_updated": "2026-07-08", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.20-0309-reasoning", + "id": "xai/grok-4.20-0309-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.20-0309-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medcode:overall:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.medcode.overall", + "source_data": { + "dataset_name": "MedCode - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medcode", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedCode (Overall).", + "metric_id": "vals_ai.medcode.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode" + } + }, + "score_details": { + "score": 32.156, + "details": { + "benchmark_slug": "medcode", + "benchmark_name": "MedCode", + "benchmark_updated": "2026-07-08", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "32.156", + "raw_stderr": "2.124", + "latency": "16.546", + "cost_per_test": "0.036189", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.124, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.20-0309-reasoning/71f3d1c1-f1f9-4186-baa7-6c8956847dec.json b/data/vals_ai/xai/grok-4.20-0309-reasoning/71f3d1c1-f1f9-4186-baa7-6c8956847dec.json new file mode 100644 index 0000000000000000000000000000000000000000..57a48036cfec12fcf25cb13ea78b2392e1bb8d6a --- /dev/null +++ b/data/vals_ai/xai/grok-4.20-0309-reasoning/71f3d1c1-f1f9-4186-baa7-6c8956847dec.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/case_law_v2/xai_grok-4.20-0309-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CaseLaw v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "case_law_v2", + "benchmark_name": "CaseLaw v2", + "benchmark_updated": "2026-05-04", + "dataset_type": "private", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/case_law_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.20-0309-reasoning", + "id": "xai/grok-4.20-0309-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.20-0309-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "case_law_v2:overall:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.case_law_v2.overall", + "source_data": { + "dataset_name": "CaseLaw v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "case_law_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/case_law_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CaseLaw v2 (Overall).", + "metric_id": "vals_ai.case_law_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/case_law_v2" + } + }, + "score_details": { + "score": 54.448, + "details": { + "benchmark_slug": "case_law_v2", + "benchmark_name": "CaseLaw v2", + "benchmark_updated": "2026-05-04", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "legal", + "raw_score": "54.448", + "raw_stderr": "0.284", + "latency": "10.663", + "cost_per_test": "0.068241", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.284, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.20-0309-reasoning/76978063-a4f9-465a-863a-4f0053b7e314.json b/data/vals_ai/xai/grok-4.20-0309-reasoning/76978063-a4f9-465a-863a-4f0053b7e314.json new file mode 100644 index 0000000000000000000000000000000000000000..98e9b756e7fc7f292314936a38accb492ea9359d --- /dev/null +++ b/data/vals_ai/xai/grok-4.20-0309-reasoning/76978063-a4f9-465a-863a-4f0053b7e314.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/xai_grok-4.20-0309-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.20-0309-reasoning", + "id": "xai/grok-4.20-0309-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.20-0309-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 83.333, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "83.333", + "raw_stderr": "8.333", + "latency": "6802.405", + "cost_per_test": "9.063814", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 8.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 16.667, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "16.667", + "raw_stderr": "1.925", + "latency": "906.987", + "cost_per_test": "1.208509", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.925, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 56.364, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "56.364", + "raw_stderr": "1.05", + "latency": "494.72", + "cost_per_test": "0.659186", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.05, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 44.195, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "44.195", + "raw_stderr": "0.991", + "latency": "305.726", + "cost_per_test": "0.407362", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.991, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.20-0309-reasoning/7ecdf4f6-7f8d-41c9-ab3a-c5efb29cb9fd.json b/data/vals_ai/xai/grok-4.20-0309-reasoning/7ecdf4f6-7f8d-41c9-ab3a-c5efb29cb9fd.json new file mode 100644 index 0000000000000000000000000000000000000000..f09b79a56cd90cc448033c984a16925312e79374 --- /dev/null +++ b/data/vals_ai/xai/grok-4.20-0309-reasoning/7ecdf4f6-7f8d-41c9-ab3a-c5efb29cb9fd.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/xai_grok-4.20-0309-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.20-0309-reasoning", + "id": "xai/grok-4.20-0309-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.20-0309-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 63.369, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.369", + "raw_stderr": "1.378", + "latency": "13.394", + "cost_per_test": "0.017152", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.378, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 74.121, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "74.121", + "raw_stderr": "0.859", + "latency": "13.394", + "cost_per_test": "0.017152", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.859, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 84.873, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "84.873", + "raw_stderr": "1.025", + "latency": "13.394", + "cost_per_test": "0.017152", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.025, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.20-0309-reasoning/8063ab1b-4246-44a7-815a-20fca699b0d4.json b/data/vals_ai/xai/grok-4.20-0309-reasoning/8063ab1b-4246-44a7-815a-20fca699b0d4.json new file mode 100644 index 0000000000000000000000000000000000000000..ba59508454ceaa43c8fd04c4cce298b7bbd8fd34 --- /dev/null +++ b/data/vals_ai/xai/grok-4.20-0309-reasoning/8063ab1b-4246-44a7-815a-20fca699b0d4.json @@ -0,0 +1,356 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/xai_grok-4.20-0309-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.20-0309-reasoning", + "id": "xai/grok-4.20-0309-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.20-0309-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 34.921, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "34.921", + "raw_stderr": "5.962", + "latency": "65.101", + "cost_per_test": "0.061403", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.962, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 40.476, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "40.476", + "raw_stderr": "6.345", + "latency": "84.742", + "cost_per_test": "0.044921", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.345, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 40.123, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "40.123", + "raw_stderr": "6.472", + "latency": "119.528", + "cost_per_test": "0.130055", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.472, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 39.329, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "39.329", + "raw_stderr": "5.459", + "latency": "49.23", + "cost_per_test": "0.083369", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.459, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 38.242, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "38.242", + "raw_stderr": "3.425", + "latency": "66.358", + "cost_per_test": "0.063231", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.425, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.20-0309-reasoning/ad5e6a0f-c13c-4d2c-8ff1-20eac69eb5d8.json b/data/vals_ai/xai/grok-4.20-0309-reasoning/ad5e6a0f-c13c-4d2c-8ff1-20eac69eb5d8.json new file mode 100644 index 0000000000000000000000000000000000000000..6db038a09fdbb3c1503bd5a0c2a12520818bc079 --- /dev/null +++ b/data/vals_ai/xai/grok-4.20-0309-reasoning/ad5e6a0f-c13c-4d2c-8ff1-20eac69eb5d8.json @@ -0,0 +1,103 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/xai_grok-4.20-0309-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.20-0309-reasoning", + "id": "xai/grok-4.20-0309-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.20-0309-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 83.468, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.468", + "raw_stderr": "0.893", + "latency": "44.461", + "cost_per_test": "0.026669", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.893, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.20-0309-reasoning/b642afb1-c539-427b-a256-5a8486c865af.json b/data/vals_ai/xai/grok-4.20-0309-reasoning/b642afb1-c539-427b-a256-5a8486c865af.json new file mode 100644 index 0000000000000000000000000000000000000000..b0b4324907f728f35ebd221b6ab308cefa51e1f7 --- /dev/null +++ b/data/vals_ai/xai/grok-4.20-0309-reasoning/b642afb1-c539-427b-a256-5a8486c865af.json @@ -0,0 +1,740 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/xai_grok-4.20-0309-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.20-0309-reasoning", + "id": "xai/grok-4.20-0309-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.20-0309-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 17.428, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "17.428", + "raw_stderr": "2.564", + "latency": "504.099", + "cost_per_test": "0.95371", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.564, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 17.492, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "17.492", + "raw_stderr": "0.099", + "latency": "351.413", + "cost_per_test": "0.791809", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.099, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 30.188, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "30.188", + "raw_stderr": "0.858", + "latency": "389.636", + "cost_per_test": "0.862158", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.858, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 39.954, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "39.954", + "raw_stderr": "0.851", + "latency": "347.25", + "cost_per_test": "0.833034", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.851, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 34.433, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "34.433", + "raw_stderr": "1.683", + "latency": "166.506", + "cost_per_test": "0.297602", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.683, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 2.447, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "2.447", + "raw_stderr": "0.959", + "latency": "387.791", + "cost_per_test": "0.905306", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.959, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 45.874, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "45.874", + "raw_stderr": "0.564", + "latency": "388.485", + "cost_per_test": "0.67815", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.564, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 37.241, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "37.241", + "raw_stderr": "3.71", + "latency": "341.249", + "cost_per_test": "0.866793", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.71, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 33.329, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "33.329", + "raw_stderr": "1.501", + "latency": "209.24", + "cost_per_test": "0.446844", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.501, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:overall:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 28.492, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "28.492", + "raw_stderr": "0.315", + "latency": "351.413", + "cost_per_test": "0.791809", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.315, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 15.536, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "15.536", + "raw_stderr": "1.652", + "latency": "448.504", + "cost_per_test": "1.29945", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.652, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.20-0309-reasoning/b8fc73e6-718c-4721-824c-f93b6250e942.json b/data/vals_ai/xai/grok-4.20-0309-reasoning/b8fc73e6-718c-4721-824c-f93b6250e942.json new file mode 100644 index 0000000000000000000000000000000000000000..b9fee637b7670f00b0c9c0887fd64f9f6ab41851 --- /dev/null +++ b/data/vals_ai/xai/grok-4.20-0309-reasoning/b8fc73e6-718c-4721-824c-f93b6250e942.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/xai_grok-4.20-0309-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.20-0309-reasoning", + "id": "xai/grok-4.20-0309-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.20-0309-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 50.079, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "50.079", + "raw_stderr": "1.41", + "latency": "13.1", + "cost_per_test": "0.026984", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.41, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 45.35, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "45.35", + "raw_stderr": "0.988", + "latency": "11.124", + "cost_per_test": "0.024144", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.988, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 40.62, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "40.62", + "raw_stderr": "1.385", + "latency": "9.149", + "cost_per_test": "0.021303", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.385, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.20-0309-reasoning/d6371dfe-8265-4de9-ae11-6cf925ab3fd8.json b/data/vals_ai/xai/grok-4.20-0309-reasoning/d6371dfe-8265-4de9-ae11-6cf925ab3fd8.json new file mode 100644 index 0000000000000000000000000000000000000000..a7e01fc61e63d9ce0b9cd65ba8f343e419cc30ef --- /dev/null +++ b/data/vals_ai/xai/grok-4.20-0309-reasoning/d6371dfe-8265-4de9-ae11-6cf925ab3fd8.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2/xai_grok-4.20-0309-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.0", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2", + "benchmark_name": "Terminal-Bench 2.0", + "benchmark_updated": "2026-06-04", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.20-0309-reasoning", + "id": "xai/grok-4.20-0309-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.20-0309-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2:easy:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.terminal-bench-2.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.0 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.0 (Easy).", + "metric_id": "vals_ai.terminal-bench-2.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "score_details": { + "score": 100.0, + "details": { + "benchmark_slug": "terminal-bench-2", + "benchmark_name": "Terminal-Bench 2.0", + "benchmark_updated": "2026-06-04", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "100", + "raw_stderr": "0", + "latency": "3591.339", + "cost_per_test": "3.428921", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2:hard:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.terminal-bench-2.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.0 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.0 (Hard).", + "metric_id": "vals_ai.terminal-bench-2.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "score_details": { + "score": 16.667, + "details": { + "benchmark_slug": "terminal-bench-2", + "benchmark_name": "Terminal-Bench 2.0", + "benchmark_updated": "2026-06-04", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "16.667", + "raw_stderr": "6.92", + "latency": "478.845", + "cost_per_test": "0.457189", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.92, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2:medium:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.terminal-bench-2.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.0 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.0 (Medium).", + "metric_id": "vals_ai.terminal-bench-2.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "score_details": { + "score": 49.091, + "details": { + "benchmark_slug": "terminal-bench-2", + "benchmark_name": "Terminal-Bench 2.0", + "benchmark_updated": "2026-06-04", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "49.091", + "raw_stderr": "6.803", + "latency": "261.188", + "cost_per_test": "0.249376", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.803, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2:overall:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.terminal-bench-2.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.0 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.0 (Overall).", + "metric_id": "vals_ai.terminal-bench-2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "score_details": { + "score": 40.449, + "details": { + "benchmark_slug": "terminal-bench-2", + "benchmark_name": "Terminal-Bench 2.0", + "benchmark_updated": "2026-06-04", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "40.449", + "raw_stderr": "5.232", + "latency": "161.408", + "cost_per_test": "0.154109", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.232, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.20-0309-reasoning/d8db4a3a-aea0-4112-b78e-8d3841ea9ac0.json b/data/vals_ai/xai/grok-4.20-0309-reasoning/d8db4a3a-aea0-4112-b78e-8d3841ea9ac0.json new file mode 100644 index 0000000000000000000000000000000000000000..222a1672b9dec8ed589e306710cbcce5e6e7a062 --- /dev/null +++ b/data/vals_ai/xai/grok-4.20-0309-reasoning/d8db4a3a-aea0-4112-b78e-8d3841ea9ac0.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/xai_grok-4.20-0309-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.20-0309-reasoning", + "id": "xai/grok-4.20-0309-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.20-0309-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 63.412, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "63.412", + "raw_stderr": "2.095", + "latency": "18.596", + "cost_per_test": "0.031303", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.095, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.20-0309-reasoning/de7cf54f-67d1-4f66-b8a1-f097b3213acd.json b/data/vals_ai/xai/grok-4.20-0309-reasoning/de7cf54f-67d1-4f66-b8a1-f097b3213acd.json new file mode 100644 index 0000000000000000000000000000000000000000..31550d7551805ede62ac62a694d01eb36fac1f13 --- /dev/null +++ b/data/vals_ai/xai/grok-4.20-0309-reasoning/de7cf54f-67d1-4f66-b8a1-f097b3213acd.json @@ -0,0 +1,438 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/xai_grok-4.20-0309-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.20-0309-reasoning", + "id": "xai/grok-4.20-0309-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.20-0309-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 85.107, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.107", + "raw_stderr": "0.839", + "latency": "7.074", + "cost_per_test": "0.005368", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.839, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.019, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.019", + "raw_stderr": "0.268", + "latency": "7.293", + "cost_per_test": "0.005972", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.268, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 70.469, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "70.469", + "raw_stderr": "0.872", + "latency": "7.387", + "cost_per_test": "0.006328", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.872, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.738, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.738", + "raw_stderr": "0.48", + "latency": "7.574", + "cost_per_test": "0.006065", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.48, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.909, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.909", + "raw_stderr": "1.114", + "latency": "7.072", + "cost_per_test": "0.005725", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.114, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.184, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.184", + "raw_stderr": "1.726", + "latency": "9.042", + "cost_per_test": "0.006934", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.726, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.20-0309-reasoning/f1946081-3cfe-4efa-a7c0-1ddbfc2444ef.json b/data/vals_ai/xai/grok-4.20-0309-reasoning/f1946081-3cfe-4efa-a7c0-1ddbfc2444ef.json new file mode 100644 index 0000000000000000000000000000000000000000..29ef3a8ee514e967da03ed11b015d4fd27fb0e1d --- /dev/null +++ b/data/vals_ai/xai/grok-4.20-0309-reasoning/f1946081-3cfe-4efa-a7c0-1ddbfc2444ef.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/xai_grok-4.20-0309-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.20-0309-reasoning", + "id": "xai/grok-4.20-0309-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.20-0309-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 87.374, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.374", + "raw_stderr": "2.36", + "latency": "60.125", + "cost_per_test": "0.035734", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.36, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 89.899, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.899", + "raw_stderr": "2.142", + "latency": "58.717", + "cost_per_test": "0.033188", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.142, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 88.636, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.636", + "raw_stderr": "1.594", + "latency": "59.421", + "cost_per_test": "0.034461", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.594, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.20-0309-reasoning/f6f5be4d-a09b-4cad-a674-471a41ef8f51.json b/data/vals_ai/xai/grok-4.20-0309-reasoning/f6f5be4d-a09b-4cad-a674-471a41ef8f51.json new file mode 100644 index 0000000000000000000000000000000000000000..f6fc2f6d310bbf1001d715b5458c5c802c99fa93 --- /dev/null +++ b/data/vals_ai/xai/grok-4.20-0309-reasoning/f6f5be4d-a09b-4cad-a674-471a41ef8f51.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/xai_grok-4.20-0309-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.20-0309-reasoning", + "id": "xai/grok-4.20-0309-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.20-0309-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.142, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.142", + "raw_stderr": "0.877", + "latency": "8.357", + "cost_per_test": "0.009299", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.877, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.381, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.381", + "raw_stderr": "0.999", + "latency": "8.986", + "cost_per_test": "0.011664", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.999, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.724, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.724", + "raw_stderr": "0.862", + "latency": "11.56", + "cost_per_test": "0.016085", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.862, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.561, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.561", + "raw_stderr": "1.63", + "latency": "10.573", + "cost_per_test": "0.012189", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.63, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.692, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.692", + "raw_stderr": "1.047", + "latency": "6.504", + "cost_per_test": "0.008769", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.047, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.179, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.179", + "raw_stderr": "1.202", + "latency": "32.25", + "cost_per_test": "0.032227", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.202, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.929, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.929", + "raw_stderr": "1.374", + "latency": "8.982", + "cost_per_test": "0.010252", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.374, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.428, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.428", + "raw_stderr": "2.142", + "latency": "7.596", + "cost_per_test": "0.010326", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.142, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 72.934, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.934", + "raw_stderr": "1.339", + "latency": "13.977", + "cost_per_test": "0.016795", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.339, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 95.115, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "95.115", + "raw_stderr": "0.586", + "latency": "22.816", + "cost_per_test": "0.013096", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.586, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.307, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.307", + "raw_stderr": "1.197", + "latency": "7.093", + "cost_per_test": "0.008299", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.197, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.254, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.254", + "raw_stderr": "0.34", + "latency": "11.654", + "cost_per_test": "0.012836", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.34, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.367, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.367", + "raw_stderr": "1.667", + "latency": "6.419", + "cost_per_test": "0.008539", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.667, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.455, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.455", + "raw_stderr": "0.776", + "latency": "11.019", + "cost_per_test": "0.014481", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.776, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.338, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.338", + "raw_stderr": "1.252", + "latency": "7.022", + "cost_per_test": "0.007685", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.252, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.20-0309-reasoning/fce8d414-06cc-46a4-90b2-8a4ea9a343a6.json b/data/vals_ai/xai/grok-4.20-0309-reasoning/fce8d414-06cc-46a4-90b2-8a4ea9a343a6.json new file mode 100644 index 0000000000000000000000000000000000000000..0dd3f84437e7e667d3d13be7041551f645d8b294 --- /dev/null +++ b/data/vals_ai/xai/grok-4.20-0309-reasoning/fce8d414-06cc-46a4-90b2-8a4ea9a343a6.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/xai_grok-4.20-0309-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.20-0309-reasoning", + "id": "xai/grok-4.20-0309-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.20-0309-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 22.5, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "22.5", + "raw_stderr": "12.057", + "latency": "622.262", + "cost_per_test": "0.654763", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 12.057, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 37.833, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "37.833", + "raw_stderr": "8.886", + "latency": "1171.599", + "cost_per_test": "0.764235", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 8.886, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 30.166, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "30.166", + "raw_stderr": "7.489", + "latency": "896.93", + "cost_per_test": "0.709499", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "2000000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.489, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 2000000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.20-0309-reasoning/fea4d0d2-5c86-48e4-a290-791f933df611.json b/data/vals_ai/xai/grok-4.20-0309-reasoning/fea4d0d2-5c86-48e4-a290-791f933df611.json new file mode 100644 index 0000000000000000000000000000000000000000..688fe60359b6c93e19791b06fac4d693a7019f73 --- /dev/null +++ b/data/vals_ai/xai/grok-4.20-0309-reasoning/fea4d0d2-5c86-48e4-a290-791f933df611.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/aime/xai_grok-4.20-0309-reasoning/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - AIME", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "dataset_type": "public", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.20-0309-reasoning", + "id": "xai/grok-4.20-0309-reasoning", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.20-0309-reasoning", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "aime:aime_2024:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.aime.aime_2024", + "source_data": { + "dataset_name": "AIME - AIME 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/aime" + ], + "additional_details": { + "benchmark_slug": "aime", + "task_key": "aime_2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for AIME (AIME 2024).", + "metric_id": "vals_ai.aime.aime_2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "score_details": { + "score": 98.75, + "details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "task_key": "aime_2024", + "task_name": "AIME 2024", + "dataset_type": "public", + "industry": "math", + "raw_score": "98.75", + "raw_stderr": "0.61", + "latency": "97.839", + "cost_per_test": "0.051914", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.61, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "aime:aime_2025:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.aime.aime_2025", + "source_data": { + "dataset_name": "AIME - AIME 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/aime" + ], + "additional_details": { + "benchmark_slug": "aime", + "task_key": "aime_2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for AIME (AIME 2025).", + "metric_id": "vals_ai.aime.aime_2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "score_details": { + "score": 94.167, + "details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "task_key": "aime_2025", + "task_name": "AIME 2025", + "dataset_type": "public", + "industry": "math", + "raw_score": "94.167", + "raw_stderr": "0.833", + "latency": "115.759", + "cost_per_test": "0.06688", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.833, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "aime:overall:grok/grok-4.20-0309-reasoning:score", + "evaluation_name": "vals_ai.aime.overall", + "source_data": { + "dataset_name": "AIME - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/aime" + ], + "additional_details": { + "benchmark_slug": "aime", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for AIME (Overall).", + "metric_id": "vals_ai.aime.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/aime" + } + }, + "score_details": { + "score": 96.458, + "details": { + "benchmark_slug": "aime", + "benchmark_name": "AIME", + "benchmark_updated": "2026-04-16", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "math", + "raw_score": "96.458", + "raw_stderr": "0.516", + "latency": "106.799", + "cost_per_test": "0.059397", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.516, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.3/039aac6d-5793-4643-8af6-8bcad2195527.json b/data/vals_ai/xai/grok-4.3/039aac6d-5793-4643-8af6-8bcad2195527.json new file mode 100644 index 0000000000000000000000000000000000000000..4ae8c4168a681d1343062988ed85c3e8ba833484 --- /dev/null +++ b/data/vals_ai/xai/grok-4.3/039aac6d-5793-4643-8af6-8bcad2195527.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/xai_grok-4.3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.3", + "id": "xai/grok-4.3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:grok/grok-4.3:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.447, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.447", + "raw_stderr": "0.689", + "latency": "1725.751", + "cost_per_test": "0.003271", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.689, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:grok/grok-4.3:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 66.0, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66", + "raw_stderr": "2.532", + "latency": "6285.07", + "cost_per_test": "0.04454", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.532, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:grok/grok-4.3:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 89.034, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "89.034", + "raw_stderr": "1.597", + "latency": "1736.065", + "cost_per_test": "0.013581", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.597, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:grok/grok-4.3:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 84.494, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "84.494", + "raw_stderr": "1.024", + "latency": "3248.962", + "cost_per_test": "0.020464", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.024, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.3/11f05db0-9bfd-4497-bff4-0d96ecc46e11.json b/data/vals_ai/xai/grok-4.3/11f05db0-9bfd-4497-bff4-0d96ecc46e11.json new file mode 100644 index 0000000000000000000000000000000000000000..0f40ed051c1674759383e4d9f31d49a2b3a178cf --- /dev/null +++ b/data/vals_ai/xai/grok-4.3/11f05db0-9bfd-4497-bff4-0d96ecc46e11.json @@ -0,0 +1,696 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/emb/xai_grok-4.3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - EMB", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.3", + "id": "xai/grok-4.3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "emb:comps:grok/grok-4.3:score", + "evaluation_name": "vals_ai.emb.comps", + "source_data": { + "dataset_name": "EMB - Comps", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "comps", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Comps).", + "metric_id": "vals_ai.emb.comps.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 16.558, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "comps", + "task_name": "Comps", + "dataset_type": "private", + "industry": "finance", + "raw_score": "16.558", + "raw_stderr": "4.949", + "latency": "246.455", + "cost_per_test": "0.816859", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.949, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:dataroom:grok/grok-4.3:score", + "evaluation_name": "vals_ai.emb.dataroom", + "source_data": { + "dataset_name": "EMB - Dataroom", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dataroom", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Dataroom).", + "metric_id": "vals_ai.emb.dataroom.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 29.176, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dataroom", + "task_name": "Dataroom", + "dataset_type": "private", + "industry": "finance", + "raw_score": "29.176", + "raw_stderr": "7.144", + "latency": "292.666", + "cost_per_test": "0.528428", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.144, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:dcf:grok/grok-4.3:score", + "evaluation_name": "vals_ai.emb.dcf", + "source_data": { + "dataset_name": "EMB - DCF", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dcf", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (DCF).", + "metric_id": "vals_ai.emb.dcf.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 11.421, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dcf", + "task_name": "DCF", + "dataset_type": "private", + "industry": "finance", + "raw_score": "11.421", + "raw_stderr": "4.91", + "latency": "277.15", + "cost_per_test": "0.566648", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.91, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:lbo:grok/grok-4.3:score", + "evaluation_name": "vals_ai.emb.lbo", + "source_data": { + "dataset_name": "EMB - LBO", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "lbo", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (LBO).", + "metric_id": "vals_ai.emb.lbo.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 9.367, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "lbo", + "task_name": "LBO", + "dataset_type": "private", + "industry": "finance", + "raw_score": "9.367", + "raw_stderr": "3.177", + "latency": "300.226", + "cost_per_test": "0.595911", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.177, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:ma:grok/grok-4.3:score", + "evaluation_name": "vals_ai.emb.ma", + "source_data": { + "dataset_name": "EMB - M&A", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "ma", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (M&A).", + "metric_id": "vals_ai.emb.ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 27.454, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "ma", + "task_name": "M&A", + "dataset_type": "private", + "industry": "finance", + "raw_score": "27.454", + "raw_stderr": "6.432", + "latency": "262.581", + "cost_per_test": "0.534988", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.432, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:misc:grok/grok-4.3:score", + "evaluation_name": "vals_ai.emb.misc", + "source_data": { + "dataset_name": "EMB - Misc", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "misc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Misc).", + "metric_id": "vals_ai.emb.misc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 18.493, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "misc", + "task_name": "Misc", + "dataset_type": "private", + "industry": "finance", + "raw_score": "18.493", + "raw_stderr": "5.185", + "latency": "254.744", + "cost_per_test": "0.815867", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.185, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:operating:grok/grok-4.3:score", + "evaluation_name": "vals_ai.emb.operating", + "source_data": { + "dataset_name": "EMB - Operating", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "operating", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Operating).", + "metric_id": "vals_ai.emb.operating.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 16.29, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "operating", + "task_name": "Operating", + "dataset_type": "private", + "industry": "finance", + "raw_score": "16.29", + "raw_stderr": "5.497", + "latency": "265.023", + "cost_per_test": "0.591564", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.497, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:overall:grok/grok-4.3:score", + "evaluation_name": "vals_ai.emb.overall", + "source_data": { + "dataset_name": "EMB - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Overall).", + "metric_id": "vals_ai.emb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 18.394, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "18.394", + "raw_stderr": "1.667", + "latency": "271.547", + "cost_per_test": "0.634124", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.667, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:scratch:grok/grok-4.3:score", + "evaluation_name": "vals_ai.emb.scratch", + "source_data": { + "dataset_name": "EMB - Scratch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "scratch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Scratch).", + "metric_id": "vals_ai.emb.scratch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 29.429, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "scratch", + "task_name": "Scratch", + "dataset_type": "private", + "industry": "finance", + "raw_score": "29.429", + "raw_stderr": "2.696", + "latency": "281.483", + "cost_per_test": "0.554337", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.696, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:template:grok/grok-4.3:score", + "evaluation_name": "vals_ai.emb.template", + "source_data": { + "dataset_name": "EMB - Template", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "template", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Template).", + "metric_id": "vals_ai.emb.template.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 6.847, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "template", + "task_name": "Template", + "dataset_type": "private", + "industry": "finance", + "raw_score": "6.847", + "raw_stderr": "2.313", + "latency": "261.611", + "cost_per_test": "0.713911", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.313, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.3/12c531aa-45e5-45ca-86e8-606759730728.json b/data/vals_ai/xai/grok-4.3/12c531aa-45e5-45ca-86e8-606759730728.json new file mode 100644 index 0000000000000000000000000000000000000000..0b2c4ed47fe4953232ed8394ed13153ea5fd2fdd --- /dev/null +++ b/data/vals_ai/xai/grok-4.3/12c531aa-45e5-45ca-86e8-606759730728.json @@ -0,0 +1,195 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/cyber/xai_grok-4.3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CyberBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "beta", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.3", + "id": "xai/grok-4.3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "cyber:overall:grok/grok-4.3:score", + "evaluation_name": "vals_ai.cyber.overall", + "source_data": { + "dataset_name": "CyberBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (Overall).", + "metric_id": "vals_ai.cyber.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 47.457, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "beta", + "raw_score": "47.457", + "raw_stderr": "6.114", + "latency": "1883.666", + "cost_per_test": "2.556509", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.114, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "cyber:patch:grok/grok-4.3:score", + "evaluation_name": "vals_ai.cyber.patch", + "source_data": { + "dataset_name": "CyberBench - Patch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "patch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (Patch).", + "metric_id": "vals_ai.cyber.patch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 64.407, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "patch", + "task_name": "Patch", + "dataset_type": "private", + "industry": "beta", + "raw_score": "64.407", + "raw_stderr": "6.233", + "latency": "2124.568", + "cost_per_test": "2.725555", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.233, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "cyber:poc:grok/grok-4.3:score", + "evaluation_name": "vals_ai.cyber.poc", + "source_data": { + "dataset_name": "CyberBench - PoC", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "poc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (PoC).", + "metric_id": "vals_ai.cyber.poc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 30.508, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "poc", + "task_name": "PoC", + "dataset_type": "private", + "industry": "beta", + "raw_score": "30.508", + "raw_stderr": "5.994", + "latency": "1642.764", + "cost_per_test": "2.387463", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.994, + "method": "vals_reported" + } + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.3/13f99ce8-fba2-45c2-9da9-5a213fd20d65.json b/data/vals_ai/xai/grok-4.3/13f99ce8-fba2-45c2-9da9-5a213fd20d65.json new file mode 100644 index 0000000000000000000000000000000000000000..8c6b688fa375d1dff1519c4193e1f0b7c683bbfc --- /dev/null +++ b/data/vals_ai/xai/grok-4.3/13f99ce8-fba2-45c2-9da9-5a213fd20d65.json @@ -0,0 +1,426 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/xai_grok-4.3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.3", + "id": "xai/grok-4.3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:grok/grok-4.3:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 89.912, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.912", + "raw_stderr": "0.746", + "latency": "316.505", + "cost_per_test": "0.002733", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.746, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:grok/grok-4.3:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.805, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.805", + "raw_stderr": "0.203", + "latency": "338.953", + "cost_per_test": "0.003208", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.203, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:grok/grok-4.3:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 89.182, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.182", + "raw_stderr": "0.57", + "latency": "275.969", + "cost_per_test": "0.003143", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.57, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:grok/grok-4.3:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 84.458, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.458", + "raw_stderr": "0.424", + "latency": "348.006", + "cost_per_test": "0.003055", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.424, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:grok/grok-4.3:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.606, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.606", + "raw_stderr": "0.877", + "latency": "347.169", + "cost_per_test": "0.002656", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.877, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:grok/grok-4.3:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.785, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.785", + "raw_stderr": "1.674", + "latency": "461.433", + "cost_per_test": "0.003536", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.674, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.3/1b93c4a2-fbe8-466f-a239-d0c803c43c0a.json b/data/vals_ai/xai/grok-4.3/1b93c4a2-fbe8-466f-a239-d0c803c43c0a.json new file mode 100644 index 0000000000000000000000000000000000000000..5e4210850ca7155fc16e423b09791712d65b6a17 --- /dev/null +++ b/data/vals_ai/xai/grok-4.3/1b93c4a2-fbe8-466f-a239-d0c803c43c0a.json @@ -0,0 +1,296 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2/xai_grok-4.3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.0", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2", + "benchmark_name": "Terminal-Bench 2.0", + "benchmark_updated": "2026-06-04", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.3", + "id": "xai/grok-4.3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2:easy:grok/grok-4.3:score", + "evaluation_name": "vals_ai.terminal-bench-2.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.0 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.0 (Easy).", + "metric_id": "vals_ai.terminal-bench-2.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "terminal-bench-2", + "benchmark_name": "Terminal-Bench 2.0", + "benchmark_updated": "2026-06-04", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66.667", + "raw_stderr": "23.57", + "latency": "43494.837", + "cost_per_test": "24.821464", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 23.57, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2:hard:grok/grok-4.3:score", + "evaluation_name": "vals_ai.terminal-bench-2.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.0 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.0 (Hard).", + "metric_id": "vals_ai.terminal-bench-2.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "score_details": { + "score": 24.444, + "details": { + "benchmark_slug": "terminal-bench-2", + "benchmark_name": "Terminal-Bench 2.0", + "benchmark_updated": "2026-06-04", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "24.444", + "raw_stderr": "5.966", + "latency": "5799.312", + "cost_per_test": "3.309529", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.966, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2:medium:grok/grok-4.3:score", + "evaluation_name": "vals_ai.terminal-bench-2.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.0 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.0 (Medium).", + "metric_id": "vals_ai.terminal-bench-2.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "score_details": { + "score": 52.121, + "details": { + "benchmark_slug": "terminal-bench-2", + "benchmark_name": "Terminal-Bench 2.0", + "benchmark_updated": "2026-06-04", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "52.121", + "raw_stderr": "5.722", + "latency": "3163.261", + "cost_per_test": "1.805197", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.722, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2:overall:grok/grok-4.3:score", + "evaluation_name": "vals_ai.terminal-bench-2.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.0 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.0 (Overall).", + "metric_id": "vals_ai.terminal-bench-2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2" + } + }, + "score_details": { + "score": 43.446, + "details": { + "benchmark_slug": "terminal-bench-2", + "benchmark_name": "Terminal-Bench 2.0", + "benchmark_updated": "2026-06-04", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "43.446", + "raw_stderr": "4.406", + "latency": "1954.824", + "cost_per_test": "1.115571", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.406, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.3/1bb2bfca-2e62-473a-8abc-f779e1dd68c1.json b/data/vals_ai/xai/grok-4.3/1bb2bfca-2e62-473a-8abc-f779e1dd68c1.json new file mode 100644 index 0000000000000000000000000000000000000000..1264a9ac749ca886ec6c09967e87d4d2bddb0530 --- /dev/null +++ b/data/vals_ai/xai/grok-4.3/1bb2bfca-2e62-473a-8abc-f779e1dd68c1.json @@ -0,0 +1,3486 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/hlab/xai_grok-4.3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Harvey's Legal Agent Benchmark", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.3", + "id": "xai/grok-4.3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "hlab:antitrust_competition:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "143.378", + "cost_per_test": "0.348081", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:arbitration_international_dispute_resolution:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "199.312", + "cost_per_test": "0.354648", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:banking_finance:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "110.467", + "cost_per_test": "0.223979", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:bankruptcy_restructuring:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "165.818", + "cost_per_test": "0.462489", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:capital_markets:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "125.773", + "cost_per_test": "0.232911", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:corporate_governance:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "174.347", + "cost_per_test": "0.394478", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:corporate_ma:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "239.67", + "cost_per_test": "0.475678", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Criteria Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Criteria Pass Rate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 64.843, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate", + "task_name": "Criteria Pass Rate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "64.843", + "raw_stderr": "0.574", + "latency": "168.727", + "cost_per_test": "0.389144", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.574, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__antitrust_competition:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 61.594, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "61.594", + "raw_stderr": "2.933", + "latency": "143.378", + "cost_per_test": "0.348081", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.933, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__arbitration_international_dispute_resolution:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 68.1, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "68.1", + "raw_stderr": "2.795", + "latency": "199.312", + "cost_per_test": "0.354648", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.795, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__banking_finance:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 67.286, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "67.286", + "raw_stderr": "2.866", + "latency": "110.467", + "cost_per_test": "0.223979", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.866, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__bankruptcy_restructuring:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 52.941, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "52.941", + "raw_stderr": "2.941", + "latency": "165.818", + "cost_per_test": "0.462489", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.941, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__capital_markets:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 74.902, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "74.902", + "raw_stderr": "2.721", + "latency": "125.773", + "cost_per_test": "0.232911", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.721, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_governance:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 75.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "75", + "raw_stderr": "2.538", + "latency": "174.347", + "cost_per_test": "0.394478", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.538, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_ma:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 70.192, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "70.192", + "raw_stderr": "2.594", + "latency": "239.67", + "cost_per_test": "0.475678", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.594, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__data_privacy_cybersecurity:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 73.852, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "73.852", + "raw_stderr": "2.617", + "latency": "159.252", + "cost_per_test": "0.356562", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.617, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__emerging_companies_venture_capital:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 58.703, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "58.703", + "raw_stderr": "2.881", + "latency": "160.625", + "cost_per_test": "0.33705", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.881, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__employment_labor:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 55.431, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "55.431", + "raw_stderr": "3.048", + "latency": "121.972", + "cost_per_test": "0.276528", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.048, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__energy_natural_resources:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 69.637, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "69.637", + "raw_stderr": "2.646", + "latency": "204.764", + "cost_per_test": "0.350581", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.646, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__environmental_esg:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 62.007, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "62.007", + "raw_stderr": "2.911", + "latency": "136.843", + "cost_per_test": "0.279814", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.911, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__funds_asset_management:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 47.813, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "47.813", + "raw_stderr": "2.701", + "latency": "213.791", + "cost_per_test": "0.705274", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.701, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__healthcare_life_sciences:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 72.607, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "72.607", + "raw_stderr": "2.566", + "latency": "172.176", + "cost_per_test": "0.43244", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.566, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__immigration:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 66.182, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "66.182", + "raw_stderr": "2.858", + "latency": "148.654", + "cost_per_test": "0.35425", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.858, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__insurance:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 47.985, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "47.985", + "raw_stderr": "3.029", + "latency": "145.798", + "cost_per_test": "0.345132", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.029, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__intellectual_property:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 73.852, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "73.852", + "raw_stderr": "2.617", + "latency": "136.238", + "cost_per_test": "0.330997", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.617, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__international_trade_sanctions:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 67.064, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "67.064", + "raw_stderr": "2.967", + "latency": "117.001", + "cost_per_test": "0.296944", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.967, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__litigation_dispute_resolution:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 67.416, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "67.416", + "raw_stderr": "2.874", + "latency": "174.521", + "cost_per_test": "0.413946", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.874, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__real_estate:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 70.262, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "70.262", + "raw_stderr": "2.472", + "latency": "217.603", + "cost_per_test": "0.312805", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.472, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__structured_finance_securitization:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 61.905, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "61.905", + "raw_stderr": "2.741", + "latency": "185.05", + "cost_per_test": "0.388122", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.741, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__tax:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 63.211, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "63.211", + "raw_stderr": "2.793", + "latency": "179.412", + "cost_per_test": "0.380436", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.793, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__trusts_estates_private_client:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 69.847, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "69.847", + "raw_stderr": "2.841", + "latency": "276.676", + "cost_per_test": "0.974394", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.841, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__white_collar_defense_investigations:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 60.943, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "60.943", + "raw_stderr": "2.836", + "latency": "140.318", + "cost_per_test": "0.311913", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.836, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:data_privacy_cybersecurity:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "159.252", + "cost_per_test": "0.356562", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:emerging_companies_venture_capital:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "160.625", + "cost_per_test": "0.33705", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:employment_labor:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "121.972", + "cost_per_test": "0.276528", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:energy_natural_resources:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "204.764", + "cost_per_test": "0.350581", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:environmental_esg:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "136.843", + "cost_per_test": "0.279814", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:funds_asset_management:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "213.791", + "cost_per_test": "0.705274", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:healthcare_life_sciences:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "172.176", + "cost_per_test": "0.43244", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:immigration:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "148.654", + "cost_per_test": "0.35425", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:insurance:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "145.798", + "cost_per_test": "0.345132", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:intellectual_property:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "136.238", + "cost_per_test": "0.330997", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:international_trade_sanctions:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "117.001", + "cost_per_test": "0.296944", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:litigation_dispute_resolution:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "174.521", + "cost_per_test": "0.413946", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:overall:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.overall", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Overall).", + "metric_id": "vals_ai.hlab.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.417, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0.417", + "raw_stderr": "0", + "latency": "168.727", + "cost_per_test": "0.389144", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:real_estate:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "217.603", + "cost_per_test": "0.312805", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:structured_finance_securitization:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "185.05", + "cost_per_test": "0.388122", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:tax:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "179.412", + "cost_per_test": "0.380436", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:trusts_estates_private_client:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "276.676", + "cost_per_test": "0.974394", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:white_collar_defense_investigations:grok/grok-4.3:score", + "evaluation_name": "vals_ai.hlab.white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "140.318", + "cost_per_test": "0.311913", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.3/21bbf526-0f92-4a14-a4ef-5a6022a0c942.json b/data/vals_ai/xai/grok-4.3/21bbf526-0f92-4a14-a4ef-5a6022a0c942.json new file mode 100644 index 0000000000000000000000000000000000000000..1031f6607b4f8894835f616d088e90f91aa8ef01 --- /dev/null +++ b/data/vals_ai/xai/grok-4.3/21bbf526-0f92-4a14-a4ef-5a6022a0c942.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/xai_grok-4.3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.3", + "id": "xai/grok-4.3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:grok/grok-4.3:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 75.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "75", + "raw_stderr": "0", + "latency": "10473.371", + "cost_per_test": "21.85401", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:grok/grok-4.3:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 14.444, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "14.444", + "raw_stderr": "2.222", + "latency": "1396.45", + "cost_per_test": "2.913868", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.222, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:grok/grok-4.3:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 54.545, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "54.545", + "raw_stderr": "3.785", + "latency": "761.7", + "cost_per_test": "1.589383", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.785, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:grok/grok-4.3:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 41.948, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "41.948", + "raw_stderr": "2.622", + "latency": "470.713", + "cost_per_test": "0.982203", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.622, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.3/2274c751-e72f-4311-825f-7ebb4fd0340e.json b/data/vals_ai/xai/grok-4.3/2274c751-e72f-4311-825f-7ebb4fd0340e.json new file mode 100644 index 0000000000000000000000000000000000000000..385794d066d931358cfe7ad292b28390311cec61 --- /dev/null +++ b/data/vals_ai/xai/grok-4.3/2274c751-e72f-4311-825f-7ebb4fd0340e.json @@ -0,0 +1,98 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/case_law_v2/xai_grok-4.3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CaseLaw v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "case_law_v2", + "benchmark_name": "CaseLaw v2", + "benchmark_updated": "2026-05-04", + "dataset_type": "private", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/case_law_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.3", + "id": "xai/grok-4.3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "case_law_v2:overall:grok/grok-4.3:score", + "evaluation_name": "vals_ai.case_law_v2.overall", + "source_data": { + "dataset_name": "CaseLaw v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "case_law_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/case_law_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CaseLaw v2 (Overall).", + "metric_id": "vals_ai.case_law_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/case_law_v2" + } + }, + "score_details": { + "score": 79.314, + "details": { + "benchmark_slug": "case_law_v2", + "benchmark_name": "CaseLaw v2", + "benchmark_updated": "2026-05-04", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "legal", + "raw_score": "79.314", + "raw_stderr": "1.368", + "latency": "42.813", + "cost_per_test": "0.040307", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.368, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.3/329024d2-c80b-4888-a8f0-02bc3228bfc2.json b/data/vals_ai/xai/grok-4.3/329024d2-c80b-4888-a8f0-02bc3228bfc2.json new file mode 100644 index 0000000000000000000000000000000000000000..261c968f8c06f9f0e585d387a7d90234e1980632 --- /dev/null +++ b/data/vals_ai/xai/grok-4.3/329024d2-c80b-4888-a8f0-02bc3228bfc2.json @@ -0,0 +1,284 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/xai_grok-4.3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.3", + "id": "xai/grok-4.3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:grok/grok-4.3:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 66.783, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.783", + "raw_stderr": "1.608", + "latency": "45.867", + "cost_per_test": "0.004871", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.608, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:grok/grok-4.3:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 68.765, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.765", + "raw_stderr": "1.582", + "latency": "38.739", + "cost_per_test": "0.316319", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.582, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:grok/grok-4.3:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 68.532, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.532", + "raw_stderr": "0.915", + "latency": "36.444", + "cost_per_test": "0.134161", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.915, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:grok/grok-4.3:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 70.047, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "70.047", + "raw_stderr": "1.564", + "latency": "24.725", + "cost_per_test": "0.081293", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.564, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.3/3afa9bff-0e48-4d67-8a58-fd57bdd945cd.json b/data/vals_ai/xai/grok-4.3/3afa9bff-0e48-4d67-8a58-fd57bdd945cd.json new file mode 100644 index 0000000000000000000000000000000000000000..9d41bbd2ca6015ea30fe71a317f987c48124dffd --- /dev/null +++ b/data/vals_ai/xai/grok-4.3/3afa9bff-0e48-4d67-8a58-fd57bdd945cd.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/xai_grok-4.3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.3", + "id": "xai/grok-4.3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:grok/grok-4.3:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 19.403, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "19.403", + "raw_stderr": "4.243", + "latency": "589.407", + "cost_per_test": "1.28472", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.243, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.3/421e7a72-9057-4c49-94d5-cbda9d0f814e.json b/data/vals_ai/xai/grok-4.3/421e7a72-9057-4c49-94d5-cbda9d0f814e.json new file mode 100644 index 0000000000000000000000000000000000000000..e71e971309611a23ccc46390fa6140a1dfcd4d0c --- /dev/null +++ b/data/vals_ai/xai/grok-4.3/421e7a72-9057-4c49-94d5-cbda9d0f814e.json @@ -0,0 +1,222 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/xai_grok-4.3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.3", + "id": "xai/grok-4.3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:grok/grok-4.3:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 60.589, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.589", + "raw_stderr": "1.397", + "latency": "391.972", + "cost_per_test": "0.013656", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.397, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:grok/grok-4.3:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 70.81, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "70.81", + "raw_stderr": "0.896", + "latency": "391.972", + "cost_per_test": "0.013656", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.896, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:grok/grok-4.3:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 81.03, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "81.03", + "raw_stderr": "1.121", + "latency": "391.972", + "cost_per_test": "0.013656", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.121, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.3/4f3136ca-d8ff-44c1-8ba4-6863b32e6bf5.json b/data/vals_ai/xai/grok-4.3/4f3136ca-d8ff-44c1-8ba4-6863b32e6bf5.json new file mode 100644 index 0000000000000000000000000000000000000000..18454f91ca9a79618bd147f3a7000f1b55461f51 --- /dev/null +++ b/data/vals_ai/xai/grok-4.3/4f3136ca-d8ff-44c1-8ba4-6863b32e6bf5.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/xai_grok-4.3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.3", + "id": "xai/grok-4.3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:grok/grok-4.3:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 74.399, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "74.399", + "raw_stderr": "2.019", + "latency": "100.457", + "cost_per_test": "0.015293", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.019, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.3/4f88f9b8-2630-4951-ba3d-87a26cfeb8f0.json b/data/vals_ai/xai/grok-4.3/4f88f9b8-2630-4951-ba3d-87a26cfeb8f0.json new file mode 100644 index 0000000000000000000000000000000000000000..ad15cb01e434603f5e8bbfd12b613278fc6da719 --- /dev/null +++ b/data/vals_ai/xai/grok-4.3/4f88f9b8-2630-4951-ba3d-87a26cfeb8f0.json @@ -0,0 +1,231 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/xai_grok-4.3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.3", + "id": "xai/grok-4.3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:grok/grok-4.3:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 92.929, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.929", + "raw_stderr": "1.822", + "latency": "396.598", + "cost_per_test": "0.024086", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.822, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:grok/grok-4.3:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 89.899, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.899", + "raw_stderr": "2.142", + "latency": "1082.674", + "cost_per_test": "0.026645", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.142, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:grok/grok-4.3:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 91.414, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.414", + "raw_stderr": "1.406", + "latency": "739.636", + "cost_per_test": "0.025366", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.406, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.3/4fd7e36e-2bf3-4ef1-90e8-5f89b27b4a0d.json b/data/vals_ai/xai/grok-4.3/4fd7e36e-2bf3-4ef1-90e8-5f89b27b4a0d.json new file mode 100644 index 0000000000000000000000000000000000000000..26eb61a558b49d172da0978a3155744858b7d966 --- /dev/null +++ b/data/vals_ai/xai/grok-4.3/4fd7e36e-2bf3-4ef1-90e8-5f89b27b4a0d.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/xai_grok-4.3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.3", + "id": "xai/grok-4.3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:grok/grok-4.3:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 13.0, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "13", + "raw_stderr": "6.272", + "latency": "555.026", + "cost_per_test": "0.32626", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.272, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:grok/grok-4.3:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 17.667, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "17.667", + "raw_stderr": "10.959", + "latency": "702.606", + "cost_per_test": "0.377249", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 10.959, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "ioi:overall:grok/grok-4.3:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 15.334, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "15.334", + "raw_stderr": "6.313", + "latency": "628.816", + "cost_per_test": "0.351754", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.313, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.3/5c3f2ead-3dcf-47fb-8f7b-097616f61b61.json b/data/vals_ai/xai/grok-4.3/5c3f2ead-3dcf-47fb-8f7b-097616f61b61.json new file mode 100644 index 0000000000000000000000000000000000000000..f05997afdf3bd1ca740e2a6673ca69a975a01c7c --- /dev/null +++ b/data/vals_ai/xai/grok-4.3/5c3f2ead-3dcf-47fb-8f7b-097616f61b61.json @@ -0,0 +1,1011 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/xai_grok-4.3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.3", + "id": "xai/grok-4.3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:grok/grok-4.3:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.305, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.305", + "raw_stderr": "0.933", + "latency": "314.075", + "cost_per_test": "0.00588", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.933, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:grok/grok-4.3:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.902, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.902", + "raw_stderr": "0.914", + "latency": "417.861", + "cost_per_test": "0.007727", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.914, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:grok/grok-4.3:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.608, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.608", + "raw_stderr": "0.824", + "latency": "307.571", + "cost_per_test": "0.012042", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.824, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:grok/grok-4.3:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.024, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.024", + "raw_stderr": "1.544", + "latency": "505.088", + "cost_per_test": "0.00878", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.544, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:grok/grok-4.3:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.166, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.166", + "raw_stderr": "1.025", + "latency": "413.591", + "cost_per_test": "0.005946", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.025, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:grok/grok-4.3:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.211, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.211", + "raw_stderr": "1.171", + "latency": "426.7", + "cost_per_test": "0.024835", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.171, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:grok/grok-4.3:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.641, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.641", + "raw_stderr": "1.324", + "latency": "402.568", + "cost_per_test": "0.00841", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.324, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:grok/grok-4.3:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.953, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.953", + "raw_stderr": "2.124", + "latency": "3277.635", + "cost_per_test": "0.007354", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.124, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:grok/grok-4.3:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.387, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.387", + "raw_stderr": "1.315", + "latency": "345.632", + "cost_per_test": "0.013321", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.315, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:grok/grok-4.3:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.538, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.538", + "raw_stderr": "0.739", + "latency": "332.566", + "cost_per_test": "0.008528", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.739, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:grok/grok-4.3:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.199, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.199", + "raw_stderr": "1.2", + "latency": "336.051", + "cost_per_test": "0.00688", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.2, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:grok/grok-4.3:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.838, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.838", + "raw_stderr": "0.33", + "latency": "637.088", + "cost_per_test": "0.009489", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.33, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:grok/grok-4.3:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.774, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.774", + "raw_stderr": "1.517", + "latency": "1170.986", + "cost_per_test": "0.006207", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.517, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:grok/grok-4.3:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.302, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.302", + "raw_stderr": "0.74", + "latency": "284.034", + "cost_per_test": "0.011467", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.74, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:grok/grok-4.3:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.719, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.719", + "raw_stderr": "1.162", + "latency": "384.867", + "cost_per_test": "0.005471", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.162, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.3/5c45c810-4050-490d-b744-71f5b8c945b7.json b/data/vals_ai/xai/grok-4.3/5c45c810-4050-490d-b744-71f5b8c945b7.json new file mode 100644 index 0000000000000000000000000000000000000000..a9ef8ab59d83706917a36c9f7727fe90b3596c76 --- /dev/null +++ b/data/vals_ai/xai/grok-4.3/5c45c810-4050-490d-b744-71f5b8c945b7.json @@ -0,0 +1,101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/xai_grok-4.3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.3", + "id": "xai/grok-4.3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:grok/grok-4.3:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 83.064, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.064", + "raw_stderr": "0.902", + "latency": "150.948", + "cost_per_test": "0.02182", + "temperature": "0.7", + "top_p": "0.95", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.902, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.3/69848c7b-17e6-42a5-9b17-dae83e640dd5.json b/data/vals_ai/xai/grok-4.3/69848c7b-17e6-42a5-9b17-dae83e640dd5.json new file mode 100644 index 0000000000000000000000000000000000000000..12715034a915dff2c4dfde1b08fd08365a50097b --- /dev/null +++ b/data/vals_ai/xai/grok-4.3/69848c7b-17e6-42a5-9b17-dae83e640dd5.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medcode/xai_grok-4.3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedCode", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medcode", + "benchmark_name": "MedCode", + "benchmark_updated": "2026-07-08", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.3", + "id": "xai/grok-4.3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medcode:overall:grok/grok-4.3:score", + "evaluation_name": "vals_ai.medcode.overall", + "source_data": { + "dataset_name": "MedCode - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medcode", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedCode (Overall).", + "metric_id": "vals_ai.medcode.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode" + } + }, + "score_details": { + "score": 38.068, + "details": { + "benchmark_slug": "medcode", + "benchmark_name": "MedCode", + "benchmark_updated": "2026-07-08", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "38.068", + "raw_stderr": "2.081", + "latency": "43.924", + "cost_per_test": "0.022202", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.081, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.3/6a907200-342e-483f-9999-10f6313cb139.json b/data/vals_ai/xai/grok-4.3/6a907200-342e-483f-9999-10f6313cb139.json new file mode 100644 index 0000000000000000000000000000000000000000..0a0e300b0a19952e2e757af723c79f8e9e9f8a9b --- /dev/null +++ b/data/vals_ai/xai/grok-4.3/6a907200-342e-483f-9999-10f6313cb139.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/programbench/xai_grok-4.3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProgramBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.3", + "id": "xai/grok-4.3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "programbench:almost:grok/grok-4.3:score", + "evaluation_name": "vals_ai.programbench.almost", + "source_data": { + "dataset_name": "ProgramBench - Almost Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "almost", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Almost Resolved).", + "metric_id": "vals_ai.programbench.almost.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "almost", + "task_name": "Almost Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "236.134", + "cost_per_test": "0.355823", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "programbench:overall:grok/grok-4.3:score", + "evaluation_name": "vals_ai.programbench.overall", + "source_data": { + "dataset_name": "ProgramBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Overall).", + "metric_id": "vals_ai.programbench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "236.134", + "cost_per_test": "0.355823", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "programbench:partial:grok/grok-4.3:score", + "evaluation_name": "vals_ai.programbench.partial", + "source_data": { + "dataset_name": "ProgramBench - Raw Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "partial", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Raw Pass Rate).", + "metric_id": "vals_ai.programbench.partial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 22.105, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "partial", + "task_name": "Raw Pass Rate", + "dataset_type": "public", + "industry": "coding", + "raw_score": "22.105", + "raw_stderr": "1.522", + "latency": "236.134", + "cost_per_test": "0.355823", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.522, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "programbench:strict:grok/grok-4.3:score", + "evaluation_name": "vals_ai.programbench.strict", + "source_data": { + "dataset_name": "ProgramBench - Fully Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "strict", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Fully Resolved).", + "metric_id": "vals_ai.programbench.strict.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "strict", + "task_name": "Fully Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "236.134", + "cost_per_test": "0.355823", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.3/6c4a86e9-8982-4550-9d84-0aa68e45d8b5.json b/data/vals_ai/xai/grok-4.3/6c4a86e9-8982-4550-9d84-0aa68e45d8b5.json new file mode 100644 index 0000000000000000000000000000000000000000..452935afceae1894f5cc10129554c92bcec9a093 --- /dev/null +++ b/data/vals_ai/xai/grok-4.3/6c4a86e9-8982-4550-9d84-0aa68e45d8b5.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/skillsbench/xai_grok-4.3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SkillsBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "skillsbench", + "benchmark_name": "SkillsBench", + "benchmark_updated": "2026-06-24", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/skillsbench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.3", + "id": "xai/grok-4.3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "skillsbench:no_skills:grok/grok-4.3:score", + "evaluation_name": "vals_ai.skillsbench.no_skills", + "source_data": { + "dataset_name": "SkillsBench - No Skills", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/skillsbench" + ], + "additional_details": { + "benchmark_slug": "skillsbench", + "task_key": "no_skills", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/skillsbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SkillsBench (No Skills).", + "metric_id": "vals_ai.skillsbench.no_skills.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/skillsbench" + } + }, + "score_details": { + "score": 19.305, + "details": { + "benchmark_slug": "skillsbench", + "benchmark_name": "SkillsBench", + "benchmark_updated": "2026-06-24", + "task_key": "no_skills", + "task_name": "No Skills", + "dataset_type": "public", + "industry": "coding", + "raw_score": "19.305", + "raw_stderr": "3.515", + "latency": "278.486", + "cost_per_test": "0.361569", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.515, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "skillsbench:overall:grok/grok-4.3:score", + "evaluation_name": "vals_ai.skillsbench.overall", + "source_data": { + "dataset_name": "SkillsBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/skillsbench" + ], + "additional_details": { + "benchmark_slug": "skillsbench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/skillsbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SkillsBench (Overall).", + "metric_id": "vals_ai.skillsbench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/skillsbench" + } + }, + "score_details": { + "score": 40.638, + "details": { + "benchmark_slug": "skillsbench", + "benchmark_name": "SkillsBench", + "benchmark_updated": "2026-06-24", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "40.638", + "raw_stderr": "4.615", + "latency": "277.173", + "cost_per_test": "0.419712", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.615, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "skillsbench:with_skills:grok/grok-4.3:score", + "evaluation_name": "vals_ai.skillsbench.with_skills", + "source_data": { + "dataset_name": "SkillsBench - With Skills", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/skillsbench" + ], + "additional_details": { + "benchmark_slug": "skillsbench", + "task_key": "with_skills", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/skillsbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SkillsBench (With Skills).", + "metric_id": "vals_ai.skillsbench.with_skills.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/skillsbench" + } + }, + "score_details": { + "score": 40.638, + "details": { + "benchmark_slug": "skillsbench", + "benchmark_name": "SkillsBench", + "benchmark_updated": "2026-06-24", + "task_key": "with_skills", + "task_name": "With Skills", + "dataset_type": "public", + "industry": "coding", + "raw_score": "40.638", + "raw_stderr": "4.615", + "latency": "277.173", + "cost_per_test": "0.419712", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.615, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.3/6d0f732d-fee7-4705-953f-785e56e911a7.json b/data/vals_ai/xai/grok-4.3/6d0f732d-fee7-4705-953f-785e56e911a7.json new file mode 100644 index 0000000000000000000000000000000000000000..378395d8afc5fc901d69c4b1b0b4bace52bea32f --- /dev/null +++ b/data/vals_ai/xai/grok-4.3/6d0f732d-fee7-4705-953f-785e56e911a7.json @@ -0,0 +1,381 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/xai_grok-4.3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.3", + "id": "xai/grok-4.3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:grok/grok-4.3:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 38.095, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "38.095", + "raw_stderr": "7.584", + "latency": "156.072", + "cost_per_test": "0.372986", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.584, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:grok/grok-4.3:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 68.199, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "68.199", + "raw_stderr": "2.888", + "latency": "129.452", + "cost_per_test": "0.24886", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.888, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:grok/grok-4.3:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 83.505, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "83.505", + "raw_stderr": "2.671", + "latency": "92.515", + "cost_per_test": "0.152287", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.671, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:grok/grok-4.3:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "197.838", + "cost_per_test": "0.377856", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:overall:grok/grok-4.3:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 71.4, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "71.4", + "raw_stderr": "2.023", + "latency": "117.767", + "cost_per_test": "0.22259", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.023, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.3/70ab085f-1f94-482e-b6b8-ec1b3d911cf3.json b/data/vals_ai/xai/grok-4.3/70ab085f-1f94-482e-b6b8-ec1b3d911cf3.json new file mode 100644 index 0000000000000000000000000000000000000000..b4f44b5bc6a04770a262da93b57c32e5419bcc21 --- /dev/null +++ b/data/vals_ai/xai/grok-4.3/70ab085f-1f94-482e-b6b8-ec1b3d911cf3.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/public-benefits-bench/xai_grok-4.3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Public Benefits Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "social mobility", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.3", + "id": "xai/grok-4.3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "public-benefits-bench:overall:grok/grok-4.3:score", + "evaluation_name": "vals_ai.public-benefits-bench.overall", + "source_data": { + "dataset_name": "Public Benefits Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Public Benefits Bench v1.1 (Overall).", + "metric_id": "vals_ai.public-benefits-bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "score_details": { + "score": 51.691, + "details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "social mobility", + "raw_score": "51.691", + "raw_stderr": "1.3", + "latency": "344.193", + "cost_per_test": "0.264337", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.3, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.3/86227df5-b7a2-41dd-b1e7-796c47db2bd6.json b/data/vals_ai/xai/grok-4.3/86227df5-b7a2-41dd-b1e7-796c47db2bd6.json new file mode 100644 index 0000000000000000000000000000000000000000..a9c9f0154b4d347f204ceccb388ec69c3c33cc37 --- /dev/null +++ b/data/vals_ai/xai/grok-4.3/86227df5-b7a2-41dd-b1e7-796c47db2bd6.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/code-migration/xai_grok-4.3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Code Migration", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.3", + "id": "xai/grok-4.3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "code-migration:cli:grok/grok-4.3:score", + "evaluation_name": "vals_ai.code-migration.cli", + "source_data": { + "dataset_name": "Code Migration - CLI", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cli", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (CLI).", + "metric_id": "vals_ai.code-migration.cli.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 9.06, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cli", + "task_name": "CLI", + "dataset_type": "private", + "industry": "coding", + "raw_score": "9.06", + "raw_stderr": "1.593", + "latency": "172.747", + "cost_per_test": "0.360596", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.593, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "code-migration:cobol:grok/grok-4.3:score", + "evaluation_name": "vals_ai.code-migration.cobol", + "source_data": { + "dataset_name": "Code Migration - COBOL", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cobol", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (COBOL).", + "metric_id": "vals_ai.code-migration.cobol.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cobol", + "task_name": "COBOL", + "dataset_type": "private", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "203.671", + "cost_per_test": "0.585897", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "code-migration:code_quality:grok/grok-4.3:score", + "evaluation_name": "vals_ai.code-migration.code_quality", + "source_data": { + "dataset_name": "Code Migration - Code Quality", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "code_quality", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Code Quality).", + "metric_id": "vals_ai.code-migration.code_quality.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 3.098, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "code_quality", + "task_name": "Code Quality", + "dataset_type": "private", + "industry": "coding", + "raw_score": "3.098", + "raw_stderr": "0.131", + "latency": "180.478", + "cost_per_test": "0.416921", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.131, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "code-migration:overall:grok/grok-4.3:score", + "evaluation_name": "vals_ai.code-migration.overall", + "source_data": { + "dataset_name": "Code Migration - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Overall).", + "metric_id": "vals_ai.code-migration.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 6.795, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "6.795", + "raw_stderr": "1.195", + "latency": "180.478", + "cost_per_test": "0.416921", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.195, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.3/8e315731-3081-400c-a840-437aa795fac3.json b/data/vals_ai/xai/grok-4.3/8e315731-3081-400c-a840-437aa795fac3.json new file mode 100644 index 0000000000000000000000000000000000000000..ccbbdf389d79c689f82d959f9e4795ad916d7243 --- /dev/null +++ b/data/vals_ai/xai/grok-4.3/8e315731-3081-400c-a840-437aa795fac3.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/xai_grok-4.3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.3", + "id": "xai/grok-4.3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:grok/grok-4.3:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 37.52, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "37.52", + "raw_stderr": "1.365", + "latency": "583.078", + "cost_per_test": "0.014746", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "256000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.365, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:grok/grok-4.3:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 48.252, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "48.252", + "raw_stderr": "0.973", + "latency": "443.266", + "cost_per_test": "0.012184", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "256000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.973, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 256000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:grok/grok-4.3:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 58.983, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.983", + "raw_stderr": "1.387", + "latency": "303.455", + "cost_per_test": "0.009623", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "256000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.387, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 256000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.3/96129247-4823-4e71-ad3f-7a779aa2254d.json b/data/vals_ai/xai/grok-4.3/96129247-4823-4e71-ad3f-7a779aa2254d.json new file mode 100644 index 0000000000000000000000000000000000000000..6a4842b502c92a97d8ec59d00ee8bf628e6bdf92 --- /dev/null +++ b/data/vals_ai/xai/grok-4.3/96129247-4823-4e71-ad3f-7a779aa2254d.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/xai_grok-4.3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.3", + "id": "xai/grok-4.3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:grok/grok-4.3:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 11.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "11", + "raw_stderr": "3.129", + "latency": "1643.019", + "cost_per_test": "0.244122", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.129, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.3/a0197e74-033c-4cbd-ab76-a44d349a5b33.json b/data/vals_ai/xai/grok-4.3/a0197e74-033c-4cbd-ab76-a44d349a5b33.json new file mode 100644 index 0000000000000000000000000000000000000000..6bf31f99cc6da108be11cb496c874e862637e50c --- /dev/null +++ b/data/vals_ai/xai/grok-4.3/a0197e74-033c-4cbd-ab76-a44d349a5b33.json @@ -0,0 +1,356 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/xai_grok-4.3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.3", + "id": "xai/grok-4.3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:grok/grok-4.3:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 5.556, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "5.556", + "raw_stderr": "2.787", + "latency": "111.882", + "cost_per_test": "0.043649", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.787, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:grok/grok-4.3:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 35.318, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "35.318", + "raw_stderr": "6.199", + "latency": "73.416", + "cost_per_test": "0.033332", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.199, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:calculus:grok/grok-4.3:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 48.148, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "48.148", + "raw_stderr": "6.333", + "latency": "65.817", + "cost_per_test": "0.058546", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:grok/grok-4.3:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 18.334, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "18.334", + "raw_stderr": "4.105", + "latency": "103.302", + "cost_per_test": "0.036954", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.105, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:grok/grok-4.3:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 19.736, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "19.736", + "raw_stderr": "2.647", + "latency": "96.2", + "cost_per_test": "0.037978", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.647, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.3/b6fb6e0f-2770-466d-b498-276639d57bb4.json b/data/vals_ai/xai/grok-4.3/b6fb6e0f-2770-466d-b498-276639d57bb4.json new file mode 100644 index 0000000000000000000000000000000000000000..ccf12ff773ef64beb9d779684596912d1f927380 --- /dev/null +++ b/data/vals_ai/xai/grok-4.3/b6fb6e0f-2770-466d-b498-276639d57bb4.json @@ -0,0 +1,366 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/public-benefits-bench-v1/xai_grok-4.3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Public Benefits Bench v1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "public-benefits-bench-v1", + "benchmark_name": "Public Benefits Bench v1", + "benchmark_updated": "2026-06-09", + "dataset_type": "private", + "industry": "social mobility", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench-v1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.3", + "id": "xai/grok-4.3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "public-benefits-bench-v1:both:grok/grok-4.3:score", + "evaluation_name": "vals_ai.public-benefits-bench-v1.both", + "source_data": { + "dataset_name": "Public Benefits Bench v1 - Both", + "source_type": "other", + "additional_details": { + "benchmark_slug": "public-benefits-bench-v1", + "task_key": "both", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench-v1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Public Benefits Bench v1 (Both).", + "metric_id": "vals_ai.public-benefits-bench-v1.both.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench-v1" + } + }, + "score_details": { + "score": 50.068, + "details": { + "benchmark_slug": "public-benefits-bench-v1", + "benchmark_name": "Public Benefits Bench v1", + "benchmark_updated": "2026-06-09", + "task_key": "both", + "task_name": "Both", + "dataset_type": "private", + "industry": "social mobility", + "raw_score": "50.068", + "raw_stderr": "1.301", + "latency": "418.342", + "cost_per_test": "1.207438", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.301, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "public-benefits-bench-v1:multi_turn:grok/grok-4.3:score", + "evaluation_name": "vals_ai.public-benefits-bench-v1.multi_turn", + "source_data": { + "dataset_name": "Public Benefits Bench v1 - Multi Turn", + "source_type": "other", + "additional_details": { + "benchmark_slug": "public-benefits-bench-v1", + "task_key": "multi_turn", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench-v1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Public Benefits Bench v1 (Multi Turn).", + "metric_id": "vals_ai.public-benefits-bench-v1.multi_turn.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench-v1" + } + }, + "score_details": { + "score": 46.888, + "details": { + "benchmark_slug": "public-benefits-bench-v1", + "benchmark_name": "Public Benefits Bench v1", + "benchmark_updated": "2026-06-09", + "task_key": "multi_turn", + "task_name": "Multi Turn", + "dataset_type": "private", + "industry": "social mobility", + "raw_score": "46.888", + "raw_stderr": "1.298", + "latency": "155.65", + "cost_per_test": "0.049242", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.298, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "public-benefits-bench-v1:neither:grok/grok-4.3:score", + "evaluation_name": "vals_ai.public-benefits-bench-v1.neither", + "source_data": { + "dataset_name": "Public Benefits Bench v1 - Neither", + "source_type": "other", + "additional_details": { + "benchmark_slug": "public-benefits-bench-v1", + "task_key": "neither", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench-v1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Public Benefits Bench v1 (Neither).", + "metric_id": "vals_ai.public-benefits-bench-v1.neither.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench-v1" + } + }, + "score_details": { + "score": 37.957, + "details": { + "benchmark_slug": "public-benefits-bench-v1", + "benchmark_name": "Public Benefits Bench v1", + "benchmark_updated": "2026-06-09", + "task_key": "neither", + "task_name": "Neither", + "dataset_type": "private", + "industry": "social mobility", + "raw_score": "37.957", + "raw_stderr": "1.262", + "latency": "21.776", + "cost_per_test": "0.006594", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.262, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "public-benefits-bench-v1:overall:grok/grok-4.3:score", + "evaluation_name": "vals_ai.public-benefits-bench-v1.overall", + "source_data": { + "dataset_name": "Public Benefits Bench v1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "public-benefits-bench-v1", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench-v1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Public Benefits Bench v1 (Overall).", + "metric_id": "vals_ai.public-benefits-bench-v1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench-v1" + } + }, + "score_details": { + "score": 50.068, + "details": { + "benchmark_slug": "public-benefits-bench-v1", + "benchmark_name": "Public Benefits Bench v1", + "benchmark_updated": "2026-06-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "social mobility", + "raw_score": "50.068", + "raw_stderr": "1.301", + "latency": "418.342", + "cost_per_test": "1.207438", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.301, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "public-benefits-bench-v1:web_search:grok/grok-4.3:score", + "evaluation_name": "vals_ai.public-benefits-bench-v1.web_search", + "source_data": { + "dataset_name": "Public Benefits Bench v1 - Web Search", + "source_type": "other", + "additional_details": { + "benchmark_slug": "public-benefits-bench-v1", + "task_key": "web_search", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench-v1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Public Benefits Bench v1 (Web Search).", + "metric_id": "vals_ai.public-benefits-bench-v1.web_search.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench-v1" + } + }, + "score_details": { + "score": 44.046, + "details": { + "benchmark_slug": "public-benefits-bench-v1", + "benchmark_name": "Public Benefits Bench v1", + "benchmark_updated": "2026-06-09", + "task_key": "web_search", + "task_name": "Web Search", + "dataset_type": "private", + "industry": "social mobility", + "raw_score": "44.046", + "raw_stderr": "1.291", + "latency": "105.689", + "cost_per_test": "0.065444", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.291, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.3/bb125d27-f96c-4a10-9a5f-ad047a7054c8.json b/data/vals_ai/xai/grok-4.3/bb125d27-f96c-4a10-9a5f-ad047a7054c8.json new file mode 100644 index 0000000000000000000000000000000000000000..711dac2e5a3dcda5fcc43b5a9d9fae6b7587f3a3 --- /dev/null +++ b/data/vals_ai/xai/grok-4.3/bb125d27-f96c-4a10-9a5f-ad047a7054c8.json @@ -0,0 +1,762 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/xai_grok-4.3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.3", + "id": "xai/grok-4.3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:grok/grok-4.3:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 31.863, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "31.863", + "raw_stderr": "2.297", + "latency": "677.894", + "cost_per_test": "0.876016", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.297, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:grok/grok-4.3:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 26.373, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "26.373", + "raw_stderr": "0.703", + "latency": "615.499", + "cost_per_test": "0.848345", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.703, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:grok/grok-4.3:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 35.857, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "35.857", + "raw_stderr": "2.76", + "latency": "777.896", + "cost_per_test": "1.111988", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.76, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:grok/grok-4.3:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 47.688, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "47.688", + "raw_stderr": "3.56", + "latency": "678.916", + "cost_per_test": "0.896477", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.56, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:grok/grok-4.3:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 47.734, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "47.734", + "raw_stderr": "3.601", + "latency": "292.659", + "cost_per_test": "0.335047", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.601, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:grok/grok-4.3:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 5.914, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "5.914", + "raw_stderr": "0.201", + "latency": "657.335", + "cost_per_test": "1.036963", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.201, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:grok/grok-4.3:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 53.198, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "53.198", + "raw_stderr": "0.902", + "latency": "749.693", + "cost_per_test": "0.78556", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.902, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:grok/grok-4.3:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 51.788, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "51.788", + "raw_stderr": "0.465", + "latency": "629.819", + "cost_per_test": "0.871512", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.465, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:grok/grok-4.3:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 50.245, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "50.245", + "raw_stderr": "0.753", + "latency": "371.126", + "cost_per_test": "0.459991", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.753, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:overall:grok/grok-4.3:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 37.728, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "37.728", + "raw_stderr": "0.373", + "latency": "615.499", + "cost_per_test": "0.848345", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.373, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:grok/grok-4.3:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 15.265, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "15.265", + "raw_stderr": "1.76", + "latency": "709.986", + "cost_per_test": "1.289733", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.76, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.3/c0958bb7-db4a-47e9-8050-aff3cda4847e.json b/data/vals_ai/xai/grok-4.3/c0958bb7-db4a-47e9-8050-aff3cda4847e.json new file mode 100644 index 0000000000000000000000000000000000000000..19e35d38b1705abc626efff1663e2da26864205a --- /dev/null +++ b/data/vals_ai/xai/grok-4.3/c0958bb7-db4a-47e9-8050-aff3cda4847e.json @@ -0,0 +1,1224 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_research/xai_grok-4.3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Legal Research Bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.3", + "id": "xai/grok-4.3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_research:administrative_regulatory:grok/grok-4.3:score", + "evaluation_name": "vals_ai.legal_research.administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 25.758, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "25.758", + "raw_stderr": "5.424", + "latency": "72.945", + "cost_per_test": "0.19233", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.424, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:business_commercial:grok/grok-4.3:score", + "evaluation_name": "vals_ai.legal_research.business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 13.953, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "13.953", + "raw_stderr": "5.347", + "latency": "73.592", + "cost_per_test": "0.170453", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.347, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:civil_litigation:grok/grok-4.3:score", + "evaluation_name": "vals_ai.legal_research.civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 3.704, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "3.704", + "raw_stderr": "3.704", + "latency": "68.656", + "cost_per_test": "0.19461", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.704, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:constitutional_civil_rights:grok/grok-4.3:score", + "evaluation_name": "vals_ai.legal_research.constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 13.158, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "13.158", + "raw_stderr": "5.557", + "latency": "125.212", + "cost_per_test": "0.250093", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.557, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:criminal:grok/grok-4.3:score", + "evaluation_name": "vals_ai.legal_research.criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 14.035, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "14.035", + "raw_stderr": "4.642", + "latency": "110.22", + "cost_per_test": "0.199087", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.642, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:family:grok/grok-4.3:score", + "evaluation_name": "vals_ai.legal_research.family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "86.589", + "cost_per_test": "0.230167", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:health:grok/grok-4.3:score", + "evaluation_name": "vals_ai.legal_research.health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 40.0, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "40", + "raw_stderr": "9.097", + "latency": "79.862", + "cost_per_test": "0.233416", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 9.097, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:immigration:grok/grok-4.3:score", + "evaluation_name": "vals_ai.legal_research.immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 9.091, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "9.091", + "raw_stderr": "9.091", + "latency": "242.659", + "cost_per_test": "0.366775", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 9.091, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:overall:grok/grok-4.3:score", + "evaluation_name": "vals_ai.legal_research.overall", + "source_data": { + "dataset_name": "Legal Research Bench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Overall).", + "metric_id": "vals_ai.legal_research.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 15.385, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "legal", + "raw_score": "15.385", + "raw_stderr": "2.508", + "latency": "84.583", + "cost_per_test": "0.203143", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.508, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass:grok/grok-4.3:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass", + "source_data": { + "dataset_name": "Legal Research Bench - Weighted Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Weighted Pass).", + "metric_id": "vals_ai.legal_research.weighted_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 69.491, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass", + "task_name": "Weighted Pass", + "dataset_type": "private", + "industry": "legal", + "raw_score": "69.491", + "raw_stderr": "1.827", + "latency": "84.583", + "cost_per_test": "0.203143", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.827, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__administrative_regulatory:grok/grok-4.3:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.weighted_pass__administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 73.866, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "73.866", + "raw_stderr": "3.085", + "latency": "72.945", + "cost_per_test": "0.19233", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.085, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__business_commercial:grok/grok-4.3:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.weighted_pass__business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 75.476, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "75.476", + "raw_stderr": "3.45", + "latency": "73.592", + "cost_per_test": "0.170453", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.45, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__civil_litigation:grok/grok-4.3:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.weighted_pass__civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 68.852, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "68.852", + "raw_stderr": "4.95", + "latency": "68.656", + "cost_per_test": "0.19461", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.95, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__constitutional_civil_rights:grok/grok-4.3:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 74.824, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "74.824", + "raw_stderr": "3.27", + "latency": "125.212", + "cost_per_test": "0.250093", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.27, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__criminal:grok/grok-4.3:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.weighted_pass__criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 69.316, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "69.316", + "raw_stderr": "3.232", + "latency": "110.22", + "cost_per_test": "0.199087", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.232, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__family:grok/grok-4.3:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.weighted_pass__family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 47.346, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "47.346", + "raw_stderr": "6.258", + "latency": "86.589", + "cost_per_test": "0.230167", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.258, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__health:grok/grok-4.3:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.weighted_pass__health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 81.212, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "81.212", + "raw_stderr": "3.744", + "latency": "79.862", + "cost_per_test": "0.233416", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.744, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__immigration:grok/grok-4.3:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.weighted_pass__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 67.726, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "67.726", + "raw_stderr": "8.228", + "latency": "242.659", + "cost_per_test": "0.366775", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 8.228, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.3/d70c56cc-5558-41ca-bdb7-56088f2c7333.json b/data/vals_ai/xai/grok-4.3/d70c56cc-5558-41ca-bdb7-56088f2c7333.json new file mode 100644 index 0000000000000000000000000000000000000000..4fdbe1c68665249aea9431af0a6e763d34480e8b --- /dev/null +++ b/data/vals_ai/xai/grok-4.3/d70c56cc-5558-41ca-bdb7-56088f2c7333.json @@ -0,0 +1,552 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_multimodal_index/xai_grok-4.3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Multimodal Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.3", + "id": "xai/grok-4.3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_multimodal_index:corp_fin_v2:grok/grok-4.3:score", + "evaluation_name": "vals_ai.vals_multimodal_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Multimodal Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (CorpFin v2).", + "metric_id": "vals_ai.vals_multimodal_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 70.047, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "70.047", + "raw_stderr": "1.564", + "latency": "24.725", + "cost_per_test": "0.081293", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.564, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:finance_agent:grok/grok-4.3:score", + "evaluation_name": "vals_ai.vals_multimodal_index.finance_agent", + "source_data": { + "dataset_name": "Vals Multimodal Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_multimodal_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 37.766, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "37.766", + "raw_stderr": "0.698", + "latency": "622.363", + "cost_per_test": "0.842163", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.698, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:mortgage_tax:grok/grok-4.3:score", + "evaluation_name": "vals_ai.vals_multimodal_index.mortgage_tax", + "source_data": { + "dataset_name": "Vals Multimodal Index - Mortgage Tax", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "mortgage_tax", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Mortgage Tax).", + "metric_id": "vals_ai.vals_multimodal_index.mortgage_tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 48.252, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "mortgage_tax", + "task_name": "Mortgage Tax", + "dataset_type": "private", + "industry": "index", + "raw_score": "48.252", + "raw_stderr": "0.973", + "latency": "443.266", + "cost_per_test": "0.012184", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "256000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.973, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 256000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:overall:grok/grok-4.3:score", + "evaluation_name": "vals_ai.vals_multimodal_index.overall", + "source_data": { + "dataset_name": "Vals Multimodal Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Overall).", + "metric_id": "vals_ai.vals_multimodal_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 43.293, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "43.293", + "raw_stderr": "1.236", + "latency": "339.036", + "cost_per_test": "0.49718", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.236, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:sage:grok/grok-4.3:score", + "evaluation_name": "vals_ai.vals_multimodal_index.sage", + "source_data": { + "dataset_name": "Vals Multimodal Index - SAGE", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "sage", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SAGE).", + "metric_id": "vals_ai.vals_multimodal_index.sage.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 19.736, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "sage", + "task_name": "SAGE", + "dataset_type": "private", + "industry": "index", + "raw_score": "19.736", + "raw_stderr": "2.647", + "latency": "96.2", + "cost_per_test": "0.037978", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.647, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:swebench:grok/grok-4.3:score", + "evaluation_name": "vals_ai.vals_multimodal_index.swebench", + "source_data": { + "dataset_name": "Vals Multimodal Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SWE-bench).", + "metric_id": "vals_ai.vals_multimodal_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 70.588, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "70.588", + "raw_stderr": "4.512", + "latency": "126.579", + "cost_per_test": "0.23972" + }, + "uncertainty": { + "standard_error": { + "value": 4.512, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:terminal_bench_2_1:grok/grok-4.3:score", + "evaluation_name": "vals_ai.vals_multimodal_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Multimodal Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_multimodal_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 41.948, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "41.948", + "raw_stderr": "2.622", + "latency": "470.713", + "cost_per_test": "0.982203", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.622, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:vibe_code_bench:grok/grok-4.3:score", + "evaluation_name": "vals_ai.vals_multimodal_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Multimodal Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_multimodal_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 15.475, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "15.475", + "raw_stderr": "5.576", + "latency": "589.407", + "cost_per_test": "1.28472", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.576, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.3/fff065df-7bb9-4d87-9b76-959f7f2cf300.json b/data/vals_ai/xai/grok-4.3/fff065df-7bb9-4d87-9b76-959f7f2cf300.json new file mode 100644 index 0000000000000000000000000000000000000000..b90dc3ce3ccde872b9d3e2433c25ce84473484d0 --- /dev/null +++ b/data/vals_ai/xai/grok-4.3/fff065df-7bb9-4d87-9b76-959f7f2cf300.json @@ -0,0 +1,418 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/xai_grok-4.3/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.3", + "id": "xai/grok-4.3", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.3", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:grok/grok-4.3:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 70.047, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "70.047", + "raw_stderr": "1.564", + "latency": "24.725", + "cost_per_test": "0.081293", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.564, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:grok/grok-4.3:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 37.766, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "37.766", + "raw_stderr": "0.698", + "latency": "622.363", + "cost_per_test": "0.842163", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.698, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:overall:grok/grok-4.3:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 46.48, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "46.48", + "raw_stderr": "1.364", + "latency": "366.757", + "cost_per_test": "0.68602", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.364, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:grok/grok-4.3:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 70.588, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "70.588", + "raw_stderr": "4.512", + "latency": "126.579", + "cost_per_test": "0.23972" + }, + "uncertainty": { + "standard_error": { + "value": 4.512, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:grok/grok-4.3:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 41.948, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "41.948", + "raw_stderr": "2.622", + "latency": "470.713", + "cost_per_test": "0.982203", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.622, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:grok/grok-4.3:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 15.475, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "15.475", + "raw_stderr": "5.576", + "latency": "589.407", + "cost_per_test": "1.28472", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.576, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.5/09dc1efd-0bb8-41da-adf6-33ee96ad2001.json b/data/vals_ai/xai/grok-4.5/09dc1efd-0bb8-41da-adf6-33ee96ad2001.json new file mode 100644 index 0000000000000000000000000000000000000000..9b9e1d849ec07def7858208854532ef6ccedd475 --- /dev/null +++ b/data/vals_ai/xai/grok-4.5/09dc1efd-0bb8-41da-adf6-33ee96ad2001.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/xai_grok-4.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.5", + "id": "xai/grok-4.5", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.5", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:grok/grok-4.5:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 99.068, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "99.068", + "raw_stderr": "0.535", + "latency": "449.171", + "cost_per_test": "0.005859", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:hard:grok/grok-4.5:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 73.429, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "73.429", + "raw_stderr": "2.361", + "latency": "831.288", + "cost_per_test": "0.100417", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.361, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:medium:grok/grok-4.5:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 89.556, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "89.556", + "raw_stderr": "1.563", + "latency": "487.608", + "cost_per_test": "0.029428", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.563, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "lcb:overall:grok/grok-4.5:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 87.351, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "87.351", + "raw_stderr": "0.961", + "latency": "589.356", + "cost_per_test": "0.045235", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.961, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.5/141dd237-432a-4acc-8540-9666c9807b45.json b/data/vals_ai/xai/grok-4.5/141dd237-432a-4acc-8540-9666c9807b45.json new file mode 100644 index 0000000000000000000000000000000000000000..8421db36c19e626b6f28b63da257d0a23f18fe40 --- /dev/null +++ b/data/vals_ai/xai/grok-4.5/141dd237-432a-4acc-8540-9666c9807b45.json @@ -0,0 +1,312 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/xai_grok-4.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.5", + "id": "xai/grok-4.5", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.5", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:grok/grok-4.5:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 91.667, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "91.667", + "raw_stderr": "8.333", + "latency": "13893.236", + "cost_per_test": "6.317739", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 8.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:grok/grok-4.5:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 56.667, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "56.667", + "raw_stderr": "1.925", + "latency": "1852.431", + "cost_per_test": "0.842365", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.925, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:grok/grok-4.5:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 72.121, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "72.121", + "raw_stderr": "5.781", + "latency": "1010.417", + "cost_per_test": "0.459472", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.781, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:grok/grok-4.5:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 67.79, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "67.79", + "raw_stderr": "4.416", + "latency": "624.415", + "cost_per_test": "0.283943", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.416, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.5/1b3809a6-c8b0-416c-859d-b6ee67ede5c1.json b/data/vals_ai/xai/grok-4.5/1b3809a6-c8b0-416c-859d-b6ee67ede5c1.json new file mode 100644 index 0000000000000000000000000000000000000000..8b28a6cc9452a371688d346cb7d83a4bc2ec0f76 --- /dev/null +++ b/data/vals_ai/xai/grok-4.5/1b3809a6-c8b0-416c-859d-b6ee67ede5c1.json @@ -0,0 +1,104 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/xai_grok-4.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.5", + "id": "xai/grok-4.5", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.5", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:grok/grok-4.5:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 86.884, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "86.884", + "raw_stderr": "1.944", + "latency": "704.513", + "cost_per_test": "0.033208", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.944, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.5/1b9688d9-1bd5-4059-b690-03e684a8048e.json b/data/vals_ai/xai/grok-4.5/1b9688d9-1bd5-4059-b690-03e684a8048e.json new file mode 100644 index 0000000000000000000000000000000000000000..37a514c1ad61a456e9558c35e993ab4896a8da61 --- /dev/null +++ b/data/vals_ai/xai/grok-4.5/1b9688d9-1bd5-4059-b690-03e684a8048e.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/xai_grok-4.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.5", + "id": "xai/grok-4.5", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.5", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:grok/grok-4.5:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 66.317, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.317", + "raw_stderr": "1.614", + "latency": "154.026", + "cost_per_test": "0.007543", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.614, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:grok/grok-4.5:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 67.483, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.483", + "raw_stderr": "1.599", + "latency": "109.542", + "cost_per_test": "0.524857", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.599, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:grok/grok-4.5:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 67.405, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.405", + "raw_stderr": "0.924", + "latency": "91.881", + "cost_per_test": "0.221291", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.924, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:grok/grok-4.5:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 68.415, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.415", + "raw_stderr": "1.587", + "latency": "12.076", + "cost_per_test": "0.131473", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.587, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.5/2fa47425-a631-4f13-bb4c-e2cc6f2e4d6c.json b/data/vals_ai/xai/grok-4.5/2fa47425-a631-4f13-bb4c-e2cc6f2e4d6c.json new file mode 100644 index 0000000000000000000000000000000000000000..dc802205ec963d524602522e1ede54a6ee4a2c92 --- /dev/null +++ b/data/vals_ai/xai/grok-4.5/2fa47425-a631-4f13-bb4c-e2cc6f2e4d6c.json @@ -0,0 +1,762 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/xai_grok-4.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.5", + "id": "xai/grok-4.5", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.5", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:grok/grok-4.5:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 38.568, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "38.568", + "raw_stderr": "1.902", + "latency": "234.84", + "cost_per_test": "0.943055", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.902, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:grok/grok-4.5:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 36.743, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "36.743", + "raw_stderr": "0.382", + "latency": "289.014", + "cost_per_test": "1.156495", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.382, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:grok/grok-4.5:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 41.768, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "41.768", + "raw_stderr": "1.428", + "latency": "377.919", + "cost_per_test": "1.583537", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.428, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:grok/grok-4.5:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 56.412, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.412", + "raw_stderr": "4.296", + "latency": "236.687", + "cost_per_test": "1.05146", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.296, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:grok/grok-4.5:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 70.848, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "70.848", + "raw_stderr": "1.094", + "latency": "178.595", + "cost_per_test": "0.373969", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.094, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:grok/grok-4.5:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 15.619, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "15.619", + "raw_stderr": "1.775", + "latency": "471.26", + "cost_per_test": "1.54595", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.775, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:grok/grok-4.5:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 71.132, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.132", + "raw_stderr": "2.382", + "latency": "245.415", + "cost_per_test": "1.166846", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.382, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:grok/grok-4.5:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 63.58, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.58", + "raw_stderr": "1.515", + "latency": "213.377", + "cost_per_test": "1.012928", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.515, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:grok/grok-4.5:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 57.839, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.839", + "raw_stderr": "0.742", + "latency": "190.111", + "cost_per_test": "0.709931", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.742, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:overall:grok/grok-4.5:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 48.349, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "48.349", + "raw_stderr": "0.458", + "latency": "289.014", + "cost_per_test": "1.156495", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.458, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:grok/grok-4.5:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 19.372, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "19.372", + "raw_stderr": "0.428", + "latency": "496.806", + "cost_per_test": "2.127011", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.428, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.5/4e873f1b-bf2c-4eee-a743-5d8419e73786.json b/data/vals_ai/xai/grok-4.5/4e873f1b-bf2c-4eee-a743-5d8419e73786.json new file mode 100644 index 0000000000000000000000000000000000000000..044bfd9c8cabcc9ca7fa5da34bea7d005f28aee5 --- /dev/null +++ b/data/vals_ai/xai/grok-4.5/4e873f1b-bf2c-4eee-a743-5d8419e73786.json @@ -0,0 +1,105 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/xai_grok-4.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.5", + "id": "xai/grok-4.5", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.5", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:grok/grok-4.5:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 61.792, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "61.792", + "raw_stderr": "1.168", + "latency": "714.648", + "cost_per_test": "0.030117", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.168, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.5/51eda23b-3321-4e7e-9e72-1ccf37873c47.json b/data/vals_ai/xai/grok-4.5/51eda23b-3321-4e7e-9e72-1ccf37873c47.json new file mode 100644 index 0000000000000000000000000000000000000000..2a8efaf4c2de26162cf2db8bc67459efc3a6551a --- /dev/null +++ b/data/vals_ai/xai/grok-4.5/51eda23b-3321-4e7e-9e72-1ccf37873c47.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/xai_grok-4.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.5", + "id": "xai/grok-4.5", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.5", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:grok/grok-4.5:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 62.551, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.551", + "raw_stderr": "1.384", + "latency": "195.458", + "cost_per_test": "0.019211", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.384, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:grok/grok-4.5:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 71.668, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.668", + "raw_stderr": "0.892", + "latency": "195.458", + "cost_per_test": "0.019211", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.892, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:grok/grok-4.5:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 80.785, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "80.785", + "raw_stderr": "1.127", + "latency": "195.458", + "cost_per_test": "0.019211", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.127, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.5/66b10b42-3abb-4f9e-bb94-cf9bd72b9deb.json b/data/vals_ai/xai/grok-4.5/66b10b42-3abb-4f9e-bb94-cf9bd72b9deb.json new file mode 100644 index 0000000000000000000000000000000000000000..530dbf223224799997540de5190bfd70344d11ae --- /dev/null +++ b/data/vals_ai/xai/grok-4.5/66b10b42-3abb-4f9e-bb94-cf9bd72b9deb.json @@ -0,0 +1,550 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_multimodal_index/xai_grok-4.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Multimodal Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.5", + "id": "xai/grok-4.5", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.5", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_multimodal_index:corp_fin_v2:grok/grok-4.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Multimodal Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (CorpFin v2).", + "metric_id": "vals_ai.vals_multimodal_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 68.415, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "68.415", + "raw_stderr": "1.587", + "latency": "12.076", + "cost_per_test": "0.131473", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.587, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:finance_agent:grok/grok-4.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.finance_agent", + "source_data": { + "dataset_name": "Vals Multimodal Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_multimodal_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 48.751, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "48.751", + "raw_stderr": "0.1", + "latency": "264.863", + "cost_per_test": "1.119629", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.1, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:mortgage_tax:grok/grok-4.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.mortgage_tax", + "source_data": { + "dataset_name": "Vals Multimodal Index - Mortgage Tax", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "mortgage_tax", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Mortgage Tax).", + "metric_id": "vals_ai.vals_multimodal_index.mortgage_tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 61.804, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "mortgage_tax", + "task_name": "Mortgage Tax", + "dataset_type": "private", + "industry": "index", + "raw_score": "61.804", + "raw_stderr": "0.968", + "latency": "117.461", + "cost_per_test": "0.014998", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.968, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:overall:grok/grok-4.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.overall", + "source_data": { + "dataset_name": "Vals Multimodal Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Overall).", + "metric_id": "vals_ai.vals_multimodal_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 63.42, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "63.42", + "raw_stderr": "1.292", + "latency": "318.501", + "cost_per_test": "0.910544", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.292, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:sage:grok/grok-4.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.sage", + "source_data": { + "dataset_name": "Vals Multimodal Index - SAGE", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "sage", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SAGE).", + "metric_id": "vals_ai.vals_multimodal_index.sage.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 34.967, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "sage", + "task_name": "SAGE", + "dataset_type": "private", + "industry": "index", + "raw_score": "34.967", + "raw_stderr": "3.316", + "latency": "169.791", + "cost_per_test": "0.115513", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.316, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:swebench:grok/grok-4.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.swebench", + "source_data": { + "dataset_name": "Vals Multimodal Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SWE-bench).", + "metric_id": "vals_ai.vals_multimodal_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 86.275, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "86.275", + "raw_stderr": "3.407", + "latency": "210.846", + "cost_per_test": "0.606056" + }, + "uncertainty": { + "standard_error": { + "value": 3.407, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:terminal_bench_2_1:grok/grok-4.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Multimodal Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_multimodal_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 67.79, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "67.79", + "raw_stderr": "4.416", + "latency": "624.415", + "cost_per_test": "0.283943", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.416, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:vibe_code_bench:grok/grok-4.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Multimodal Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_multimodal_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 72.754, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "72.754", + "raw_stderr": "5.807", + "latency": "830.057", + "cost_per_test": "4.102194", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.807, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.5/6b11eb9b-9055-47e8-8c42-c1fee177dc77.json b/data/vals_ai/xai/grok-4.5/6b11eb9b-9055-47e8-8c42-c1fee177dc77.json new file mode 100644 index 0000000000000000000000000000000000000000..f9732af48e18c8058e4cf43421f22cea9597ffa3 --- /dev/null +++ b/data/vals_ai/xai/grok-4.5/6b11eb9b-9055-47e8-8c42-c1fee177dc77.json @@ -0,0 +1,104 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medcode/xai_grok-4.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedCode", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medcode", + "benchmark_name": "MedCode", + "benchmark_updated": "2026-07-08", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.5", + "id": "xai/grok-4.5", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.5", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medcode:overall:grok/grok-4.5:score", + "evaluation_name": "vals_ai.medcode.overall", + "source_data": { + "dataset_name": "MedCode - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medcode", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedCode (Overall).", + "metric_id": "vals_ai.medcode.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medcode" + } + }, + "score_details": { + "score": 43.291, + "details": { + "benchmark_slug": "medcode", + "benchmark_name": "MedCode", + "benchmark_updated": "2026-07-08", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "43.291", + "raw_stderr": "2.313", + "latency": "59.637", + "cost_per_test": "0.048461", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.313, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.5/88a67a41-d1a3-4939-8ba2-22c5c3e45164.json b/data/vals_ai/xai/grok-4.5/88a67a41-d1a3-4939-8ba2-22c5c3e45164.json new file mode 100644 index 0000000000000000000000000000000000000000..47680ca8039f7e542a77abc4aebb2236108d98bf --- /dev/null +++ b/data/vals_ai/xai/grok-4.5/88a67a41-d1a3-4939-8ba2-22c5c3e45164.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/xai_grok-4.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.5", + "id": "xai/grok-4.5", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.5", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:grok/grok-4.5:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 30.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "30", + "raw_stderr": "4.606", + "latency": "460.33", + "cost_per_test": "0.484548", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.606, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.5/97ed6163-3cf1-4e6c-a6b4-15b7afe5052b.json b/data/vals_ai/xai/grok-4.5/97ed6163-3cf1-4e6c-a6b4-15b7afe5052b.json new file mode 100644 index 0000000000000000000000000000000000000000..71344e429e8557aea3d01437d8bdb0e980ef12d6 --- /dev/null +++ b/data/vals_ai/xai/grok-4.5/97ed6163-3cf1-4e6c-a6b4-15b7afe5052b.json @@ -0,0 +1,418 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/xai_grok-4.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.5", + "id": "xai/grok-4.5", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.5", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:grok/grok-4.5:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 68.415, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "68.415", + "raw_stderr": "1.587", + "latency": "12.076", + "cost_per_test": "0.131473", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.587, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:grok/grok-4.5:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 48.751, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "48.751", + "raw_stderr": "0.1", + "latency": "264.863", + "cost_per_test": "1.119629", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.1, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:overall:grok/grok-4.5:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 65.299, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "65.299", + "raw_stderr": "1.406", + "latency": "388.451", + "cost_per_test": "1.248659", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.406, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:grok/grok-4.5:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 86.275, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "86.275", + "raw_stderr": "3.407", + "latency": "210.846", + "cost_per_test": "0.606056" + }, + "uncertainty": { + "standard_error": { + "value": 3.407, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:grok/grok-4.5:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 67.79, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "67.79", + "raw_stderr": "4.416", + "latency": "624.415", + "cost_per_test": "0.283943", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.416, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:grok/grok-4.5:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 72.754, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "72.754", + "raw_stderr": "5.807", + "latency": "830.057", + "cost_per_test": "4.102194", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.807, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.5/984b131b-aae2-401a-96ea-f1ab28867d2a.json b/data/vals_ai/xai/grok-4.5/984b131b-aae2-401a-96ea-f1ab28867d2a.json new file mode 100644 index 0000000000000000000000000000000000000000..52758d015a8b9392a0084805b49bcd5e135d8e51 --- /dev/null +++ b/data/vals_ai/xai/grok-4.5/984b131b-aae2-401a-96ea-f1ab28867d2a.json @@ -0,0 +1,381 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/xai_grok-4.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.5", + "id": "xai/grok-4.5", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.5", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:grok/grok-4.5:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 73.81, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "73.81", + "raw_stderr": "6.867", + "latency": "327.841", + "cost_per_test": "1.093841", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.867, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:grok/grok-4.5:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 83.525, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "83.525", + "raw_stderr": "2.301", + "latency": "227.657", + "cost_per_test": "0.604684", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.301, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:grok/grok-4.5:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 93.814, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "93.814", + "raw_stderr": "1.734", + "latency": "131.31", + "cost_per_test": "0.329545", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.734, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:grok/grok-4.5:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66.667", + "raw_stderr": "33.333", + "latency": "384.216", + "cost_per_test": "0.846307", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "swebench:overall:grok/grok-4.5:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 86.6, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "86.6", + "raw_stderr": "1.525", + "latency": "199.573", + "cost_per_test": "0.54034", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.525, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.5/9f2c3ba3-3e16-4a6b-af1b-74db41c4cb12.json b/data/vals_ai/xai/grok-4.5/9f2c3ba3-3e16-4a6b-af1b-74db41c4cb12.json new file mode 100644 index 0000000000000000000000000000000000000000..ec57cc2dec9af0affb5269ffaace80a65ecd3790 --- /dev/null +++ b/data/vals_ai/xai/grok-4.5/9f2c3ba3-3e16-4a6b-af1b-74db41c4cb12.json @@ -0,0 +1,308 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/programbench/xai_grok-4.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProgramBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.5", + "id": "xai/grok-4.5", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.5", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "programbench:almost:grok/grok-4.5:score", + "evaluation_name": "vals_ai.programbench.almost", + "source_data": { + "dataset_name": "ProgramBench - Almost Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "almost", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Almost Resolved).", + "metric_id": "vals_ai.programbench.almost.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 4.5, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "almost", + "task_name": "Almost Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "4.5", + "raw_stderr": "1.47", + "latency": "1139.043", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.47, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "programbench:overall:grok/grok-4.5:score", + "evaluation_name": "vals_ai.programbench.overall", + "source_data": { + "dataset_name": "ProgramBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Overall).", + "metric_id": "vals_ai.programbench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1139.043", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "programbench:partial:grok/grok-4.5:score", + "evaluation_name": "vals_ai.programbench.partial", + "source_data": { + "dataset_name": "ProgramBench - Raw Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "partial", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Raw Pass Rate).", + "metric_id": "vals_ai.programbench.partial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 65.246, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "partial", + "task_name": "Raw Pass Rate", + "dataset_type": "public", + "industry": "coding", + "raw_score": "65.246", + "raw_stderr": "1.81", + "latency": "1139.043", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.81, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "programbench:strict:grok/grok-4.5:score", + "evaluation_name": "vals_ai.programbench.strict", + "source_data": { + "dataset_name": "ProgramBench - Fully Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "strict", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Fully Resolved).", + "metric_id": "vals_ai.programbench.strict.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "strict", + "task_name": "Fully Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1139.043", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.5/a17ac4d1-c53c-4761-b70a-3bc01c93f4af.json b/data/vals_ai/xai/grok-4.5/a17ac4d1-c53c-4761-b70a-3bc01c93f4af.json new file mode 100644 index 0000000000000000000000000000000000000000..0c7260580f393b696d39a3bcebf9f3ec64009ec7 --- /dev/null +++ b/data/vals_ai/xai/grok-4.5/a17ac4d1-c53c-4761-b70a-3bc01c93f4af.json @@ -0,0 +1,376 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/xai_grok-4.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.5", + "id": "xai/grok-4.5", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.5", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:grok/grok-4.5:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 25.397, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "25.397", + "raw_stderr": "5.477", + "latency": "121.888", + "cost_per_test": "0.153336", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.477, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:automata:grok/grok-4.5:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 32.54, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "32.54", + "raw_stderr": "6.132", + "latency": "280.804", + "cost_per_test": "0.084413", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.132, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:calculus:grok/grok-4.5:score", + "evaluation_name": "vals_ai.sage.calculus", + "source_data": { + "dataset_name": "SAGE - Calculus", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "calculus", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Calculus).", + "metric_id": "vals_ai.sage.calculus.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 59.568, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "calculus", + "task_name": "Calculus", + "dataset_type": "private", + "industry": "education", + "raw_score": "59.568", + "raw_stderr": "6.008", + "latency": "696.783", + "cost_per_test": "0.101612", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.008, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:grok/grok-4.5:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 46.964, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "46.964", + "raw_stderr": "5.598", + "latency": "106.68", + "cost_per_test": "0.108791", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.598, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "sage:overall:grok/grok-4.5:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 34.967, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "34.967", + "raw_stderr": "3.316", + "latency": "169.791", + "cost_per_test": "0.115513", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.316, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.5/ab625300-d877-4b72-bcfe-468503da1406.json b/data/vals_ai/xai/grok-4.5/ab625300-d877-4b72-bcfe-468503da1406.json new file mode 100644 index 0000000000000000000000000000000000000000..5e6f825cd24721ff53253cbfa2299b3345c22b14 --- /dev/null +++ b/data/vals_ai/xai/grok-4.5/ab625300-d877-4b72-bcfe-468503da1406.json @@ -0,0 +1,450 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/xai_grok-4.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.5", + "id": "xai/grok-4.5", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.5", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:grok/grok-4.5:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 89.507, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.507", + "raw_stderr": "0.671", + "latency": "59.01", + "cost_per_test": "0.004226", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.671, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:grok/grok-4.5:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.22, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.22", + "raw_stderr": "0.191", + "latency": "65.63", + "cost_per_test": "0.004689", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.191, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:grok/grok-4.5:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 90.129, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.129", + "raw_stderr": "0.573", + "latency": "85.022", + "cost_per_test": "0.004414", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.573, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:grok/grok-4.5:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 85.965, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.965", + "raw_stderr": "0.413", + "latency": "67.875", + "cost_per_test": "0.004535", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.413, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:grok/grok-4.5:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.802, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.802", + "raw_stderr": "0.808", + "latency": "61.924", + "cost_per_test": "0.003947", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.808, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:grok/grok-4.5:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 84.168, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.168", + "raw_stderr": "1.674", + "latency": "67.788", + "cost_per_test": "0.005398", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.674, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.5/af220001-a394-4f65-a77d-3eaf85956fe4.json b/data/vals_ai/xai/grok-4.5/af220001-a394-4f65-a77d-3eaf85956fe4.json new file mode 100644 index 0000000000000000000000000000000000000000..c20ca5ad1d50770153fe99b38399af9e4709ef47 --- /dev/null +++ b/data/vals_ai/xai/grok-4.5/af220001-a394-4f65-a77d-3eaf85956fe4.json @@ -0,0 +1,234 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/xai_grok-4.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.5", + "id": "xai/grok-4.5", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.5", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:grok/grok-4.5:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 62.798, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.798", + "raw_stderr": "1.363", + "latency": "117.059", + "cost_per_test": "0.016464", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.363, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:grok/grok-4.5:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 61.804, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.804", + "raw_stderr": "0.968", + "latency": "117.461", + "cost_per_test": "0.014998", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.968, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:grok/grok-4.5:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 60.811, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.811", + "raw_stderr": "1.376", + "latency": "117.863", + "cost_per_test": "0.013533", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.376, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.5/bccece32-1e39-4514-a05d-caf337aea4c7.json b/data/vals_ai/xai/grok-4.5/bccece32-1e39-4514-a05d-caf337aea4c7.json new file mode 100644 index 0000000000000000000000000000000000000000..2b7fe3937d9ce9ef0cb7138dc1a320ccb64d0d72 --- /dev/null +++ b/data/vals_ai/xai/grok-4.5/bccece32-1e39-4514-a05d-caf337aea4c7.json @@ -0,0 +1,1071 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/xai_grok-4.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.5", + "id": "xai/grok-4.5", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.5", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:grok/grok-4.5:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.282, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.282", + "raw_stderr": "0.867", + "latency": "377.984", + "cost_per_test": "0.009436", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.867, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:grok/grok-4.5:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.283, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.283", + "raw_stderr": "0.891", + "latency": "336.555", + "cost_per_test": "0.012806", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.891, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:grok/grok-4.5:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.608, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.608", + "raw_stderr": "0.824", + "latency": "273.466", + "cost_per_test": "0.018195", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.824, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:grok/grok-4.5:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.976, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.976", + "raw_stderr": "1.415", + "latency": "495.996", + "cost_per_test": "0.013831", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.415, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:grok/grok-4.5:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.351, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.351", + "raw_stderr": "0.968", + "latency": "325.874", + "cost_per_test": "0.009717", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.968, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:grok/grok-4.5:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.203, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.203", + "raw_stderr": "1.105", + "latency": "454.689", + "cost_per_test": "0.040013", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.105, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:grok/grok-4.5:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.985, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.985", + "raw_stderr": "1.282", + "latency": "315.299", + "cost_per_test": "0.011257", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.282, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:grok/grok-4.5:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.365, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.365", + "raw_stderr": "1.995", + "latency": "427.645", + "cost_per_test": "0.011201", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.995, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:grok/grok-4.5:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.016, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.016", + "raw_stderr": "1.175", + "latency": "394.222", + "cost_per_test": "0.023265", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.175, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:grok/grok-4.5:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 95.929, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "95.929", + "raw_stderr": "0.551", + "latency": "340.418", + "cost_per_test": "0.014781", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.551, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:grok/grok-4.5:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.229, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.229", + "raw_stderr": "1.11", + "latency": "310.979", + "cost_per_test": "0.010007", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.11, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:grok/grok-4.5:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.216, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.216", + "raw_stderr": "0.31", + "latency": "357.751", + "cost_per_test": "0.014974", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.31, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:grok/grok-4.5:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.575, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.575", + "raw_stderr": "1.477", + "latency": "356.138", + "cost_per_test": "0.009439", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.477, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:grok/grok-4.5:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.994, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.994", + "raw_stderr": "0.753", + "latency": "257.307", + "cost_per_test": "0.017827", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.753, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:grok/grok-4.5:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.226, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.226", + "raw_stderr": "1.051", + "latency": "341.948", + "cost_per_test": "0.00786", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.051, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.5/bd451225-f204-4006-9b29-76a4e7ac9858.json b/data/vals_ai/xai/grok-4.5/bd451225-f204-4006-9b29-76a4e7ac9858.json new file mode 100644 index 0000000000000000000000000000000000000000..5bce25f2373d2e1d6e3502c550ad9b0d86645553 --- /dev/null +++ b/data/vals_ai/xai/grok-4.5/bd451225-f204-4006-9b29-76a4e7ac9858.json @@ -0,0 +1,696 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/emb/xai_grok-4.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - EMB", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.5", + "id": "xai/grok-4.5", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.5", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "emb:comps:grok/grok-4.5:score", + "evaluation_name": "vals_ai.emb.comps", + "source_data": { + "dataset_name": "EMB - Comps", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "comps", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Comps).", + "metric_id": "vals_ai.emb.comps.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 39.489, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "comps", + "task_name": "Comps", + "dataset_type": "private", + "industry": "finance", + "raw_score": "39.489", + "raw_stderr": "9.454", + "latency": "1594.847", + "cost_per_test": "3.025508", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 9.454, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:dataroom:grok/grok-4.5:score", + "evaluation_name": "vals_ai.emb.dataroom", + "source_data": { + "dataset_name": "EMB - Dataroom", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dataroom", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Dataroom).", + "metric_id": "vals_ai.emb.dataroom.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 64.953, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dataroom", + "task_name": "Dataroom", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.953", + "raw_stderr": "10.409", + "latency": "1003.567", + "cost_per_test": "1.119264", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 10.409, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:dcf:grok/grok-4.5:score", + "evaluation_name": "vals_ai.emb.dcf", + "source_data": { + "dataset_name": "EMB - DCF", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dcf", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (DCF).", + "metric_id": "vals_ai.emb.dcf.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 41.455, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dcf", + "task_name": "DCF", + "dataset_type": "private", + "industry": "finance", + "raw_score": "41.455", + "raw_stderr": "8.461", + "latency": "831.683", + "cost_per_test": "1.193483", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 8.461, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:lbo:grok/grok-4.5:score", + "evaluation_name": "vals_ai.emb.lbo", + "source_data": { + "dataset_name": "EMB - LBO", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "lbo", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (LBO).", + "metric_id": "vals_ai.emb.lbo.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 48.684, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "lbo", + "task_name": "LBO", + "dataset_type": "private", + "industry": "finance", + "raw_score": "48.684", + "raw_stderr": "6.2", + "latency": "1004.88", + "cost_per_test": "1.925702", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.2, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:ma:grok/grok-4.5:score", + "evaluation_name": "vals_ai.emb.ma", + "source_data": { + "dataset_name": "EMB - M&A", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "ma", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (M&A).", + "metric_id": "vals_ai.emb.ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 64.62, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "ma", + "task_name": "M&A", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.62", + "raw_stderr": "9.802", + "latency": "976.599", + "cost_per_test": "1.016629", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 9.802, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:misc:grok/grok-4.5:score", + "evaluation_name": "vals_ai.emb.misc", + "source_data": { + "dataset_name": "EMB - Misc", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "misc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Misc).", + "metric_id": "vals_ai.emb.misc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 49.24, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "misc", + "task_name": "Misc", + "dataset_type": "private", + "industry": "finance", + "raw_score": "49.24", + "raw_stderr": "9.277", + "latency": "1024.593", + "cost_per_test": "1.245392", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 9.277, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:operating:grok/grok-4.5:score", + "evaluation_name": "vals_ai.emb.operating", + "source_data": { + "dataset_name": "EMB - Operating", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "operating", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Operating).", + "metric_id": "vals_ai.emb.operating.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 61.935, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "operating", + "task_name": "Operating", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.935", + "raw_stderr": "7.615", + "latency": "1233.928", + "cost_per_test": "1.114161", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.615, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:overall:grok/grok-4.5:score", + "evaluation_name": "vals_ai.emb.overall", + "source_data": { + "dataset_name": "EMB - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Overall).", + "metric_id": "vals_ai.emb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 52.911, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.911", + "raw_stderr": "3.111", + "latency": "1107.838", + "cost_per_test": "1.55492", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.111, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:scratch:grok/grok-4.5:score", + "evaluation_name": "vals_ai.emb.scratch", + "source_data": { + "dataset_name": "EMB - Scratch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "scratch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Scratch).", + "metric_id": "vals_ai.emb.scratch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 67.327, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "scratch", + "task_name": "Scratch", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.327", + "raw_stderr": "2.684", + "latency": "934.097", + "cost_per_test": "1.504318", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.684, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "emb:template:grok/grok-4.5:score", + "evaluation_name": "vals_ai.emb.template", + "source_data": { + "dataset_name": "EMB - Template", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "template", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Template).", + "metric_id": "vals_ai.emb.template.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 38.301, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "template", + "task_name": "Template", + "dataset_type": "private", + "industry": "finance", + "raw_score": "38.301", + "raw_stderr": "5.486", + "latency": "1281.579", + "cost_per_test": "1.605523", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.486, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.5/d16293bf-3e9a-47b9-9463-201af997a686.json b/data/vals_ai/xai/grok-4.5/d16293bf-3e9a-47b9-9463-201af997a686.json new file mode 100644 index 0000000000000000000000000000000000000000..dda96f292a671f6d6f982f765b0bd231a0bb59a7 --- /dev/null +++ b/data/vals_ai/xai/grok-4.5/d16293bf-3e9a-47b9-9463-201af997a686.json @@ -0,0 +1,1224 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_research/xai_grok-4.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Legal Research Bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.5", + "id": "xai/grok-4.5", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.5", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_research:administrative_regulatory:grok/grok-4.5:score", + "evaluation_name": "vals_ai.legal_research.administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 51.515, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "51.515", + "raw_stderr": "6.199", + "latency": "1001.46", + "cost_per_test": "1.36196", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.199, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:business_commercial:grok/grok-4.5:score", + "evaluation_name": "vals_ai.legal_research.business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 30.233, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "30.233", + "raw_stderr": "7.087", + "latency": "913.577", + "cost_per_test": "0.893714", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.087, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:civil_litigation:grok/grok-4.5:score", + "evaluation_name": "vals_ai.legal_research.civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "33.333", + "raw_stderr": "9.245", + "latency": "1225.988", + "cost_per_test": "1.215179", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 9.245, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:constitutional_civil_rights:grok/grok-4.5:score", + "evaluation_name": "vals_ai.legal_research.constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 31.579, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "31.579", + "raw_stderr": "7.642", + "latency": "907.475", + "cost_per_test": "0.836603", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.642, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:criminal:grok/grok-4.5:score", + "evaluation_name": "vals_ai.legal_research.criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 36.842, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "36.842", + "raw_stderr": "6.446", + "latency": "874.891", + "cost_per_test": "0.822469", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.446, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:family:grok/grok-4.5:score", + "evaluation_name": "vals_ai.legal_research.family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 9.091, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "9.091", + "raw_stderr": "6.273", + "latency": "839.919", + "cost_per_test": "0.991333", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 6.273, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:health:grok/grok-4.5:score", + "evaluation_name": "vals_ai.legal_research.health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "66.667", + "raw_stderr": "8.754", + "latency": "1266.05", + "cost_per_test": "2.204201", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 8.754, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:immigration:grok/grok-4.5:score", + "evaluation_name": "vals_ai.legal_research.immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 36.364, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "36.364", + "raw_stderr": "15.212", + "latency": "771.306", + "cost_per_test": "1.548677", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 15.212, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:overall:grok/grok-4.5:score", + "evaluation_name": "vals_ai.legal_research.overall", + "source_data": { + "dataset_name": "Legal Research Bench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Overall).", + "metric_id": "vals_ai.legal_research.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 37.981, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "legal", + "raw_score": "37.981", + "raw_stderr": "3.373", + "latency": "953.767", + "cost_per_test": "1.086814", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.373, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass:grok/grok-4.5:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass", + "source_data": { + "dataset_name": "Legal Research Bench - Weighted Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Weighted Pass).", + "metric_id": "vals_ai.legal_research.weighted_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 84.423, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass", + "task_name": "Weighted Pass", + "dataset_type": "private", + "industry": "legal", + "raw_score": "84.423", + "raw_stderr": "1.429", + "latency": "953.767", + "cost_per_test": "1.086814", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.429, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__administrative_regulatory:grok/grok-4.5:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.weighted_pass__administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 88.75, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "88.75", + "raw_stderr": "2.226", + "latency": "1001.46", + "cost_per_test": "1.36196", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.226, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__business_commercial:grok/grok-4.5:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.weighted_pass__business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 86.805, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "86.805", + "raw_stderr": "2.755", + "latency": "913.577", + "cost_per_test": "0.893714", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.755, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__civil_litigation:grok/grok-4.5:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.weighted_pass__civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 88.748, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "88.748", + "raw_stderr": "2.752", + "latency": "1225.988", + "cost_per_test": "1.215179", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.752, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__constitutional_civil_rights:grok/grok-4.5:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 85.897, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "85.897", + "raw_stderr": "2.679", + "latency": "907.475", + "cost_per_test": "0.836603", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.679, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__criminal:grok/grok-4.5:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.weighted_pass__criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 82.619, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "82.619", + "raw_stderr": "2.863", + "latency": "874.891", + "cost_per_test": "0.822469", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.863, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__family:grok/grok-4.5:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.weighted_pass__family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 70.56, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "70.56", + "raw_stderr": "5.33", + "latency": "839.919", + "cost_per_test": "0.991333", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 5.33, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__health:grok/grok-4.5:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.weighted_pass__health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 93.249, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "93.249", + "raw_stderr": "2.415", + "latency": "1266.05", + "cost_per_test": "2.204201", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.415, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__immigration:grok/grok-4.5:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.weighted_pass__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 82.311, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "82.311", + "raw_stderr": "7.186", + "latency": "771.306", + "cost_per_test": "1.548677", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 7.186, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.5/d18b9398-1322-481f-8e27-6387fa35760d.json b/data/vals_ai/xai/grok-4.5/d18b9398-1322-481f-8e27-6387fa35760d.json new file mode 100644 index 0000000000000000000000000000000000000000..759a0282346b83a3906d3ddd14d4b4ce825b111b --- /dev/null +++ b/data/vals_ai/xai/grok-4.5/d18b9398-1322-481f-8e27-6387fa35760d.json @@ -0,0 +1,102 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/xai_grok-4.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.5", + "id": "xai/grok-4.5", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.5", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:grok/grok-4.5:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 69.002, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "69.002", + "raw_stderr": "4.48", + "latency": "830.057", + "cost_per_test": "4.102194", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.48, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.5/d5ec2bc0-7da8-4ead-a9ce-e950d046fbc5.json b/data/vals_ai/xai/grok-4.5/d5ec2bc0-7da8-4ead-a9ce-e950d046fbc5.json new file mode 100644 index 0000000000000000000000000000000000000000..1c5eb3217535a7c5b03a80c8f217beb36e16121a --- /dev/null +++ b/data/vals_ai/xai/grok-4.5/d5ec2bc0-7da8-4ead-a9ce-e950d046fbc5.json @@ -0,0 +1,243 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/xai_grok-4.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.5", + "id": "xai/grok-4.5", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.5", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:grok/grok-4.5:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 92.929, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.929", + "raw_stderr": "1.822", + "latency": "288.128", + "cost_per_test": "0.046385", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.822, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:grok/grok-4.5:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 92.929, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.929", + "raw_stderr": "1.822", + "latency": "147.589", + "cost_per_test": "0.041673", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.822, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:grok/grok-4.5:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 92.929, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.929", + "raw_stderr": "1.288", + "latency": "217.858", + "cost_per_test": "0.044029", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.288, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.5/eaa61b82-7c94-456e-b1ea-c28dc3c22b86.json b/data/vals_ai/xai/grok-4.5/eaa61b82-7c94-456e-b1ea-c28dc3c22b86.json new file mode 100644 index 0000000000000000000000000000000000000000..c6e80edcf68885bfd103a05119c885f6aeeb856c --- /dev/null +++ b/data/vals_ai/xai/grok-4.5/eaa61b82-7c94-456e-b1ea-c28dc3c22b86.json @@ -0,0 +1,3486 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/hlab/xai_grok-4.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Harvey's Legal Agent Benchmark", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.5", + "id": "xai/grok-4.5", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.5", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "hlab:antitrust_competition:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "651.432", + "cost_per_test": "2.727866", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:arbitration_international_dispute_resolution:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "502.538", + "cost_per_test": "1.759751", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:banking_finance:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "462.168", + "cost_per_test": "1.30437", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:bankruptcy_restructuring:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "570.801", + "cost_per_test": "2.356863", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:capital_markets:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 40.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "40", + "raw_stderr": "24.495", + "latency": "544.738", + "cost_per_test": "1.23096", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 24.495, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:corporate_governance:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "669.875", + "cost_per_test": "1.86509", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:corporate_ma:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "767.352", + "cost_per_test": "2.416866", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Criteria Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Criteria Pass Rate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 90.549, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate", + "task_name": "Criteria Pass Rate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.549", + "raw_stderr": "0.352", + "latency": "614.713", + "cost_per_test": "2.01955", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.352, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__antitrust_competition:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 87.319, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.319", + "raw_stderr": "2.007", + "latency": "651.432", + "cost_per_test": "2.727866", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.007, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__arbitration_international_dispute_resolution:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 92.115, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "92.115", + "raw_stderr": "1.616", + "latency": "502.538", + "cost_per_test": "1.759751", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.616, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__banking_finance:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 95.539, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "95.539", + "raw_stderr": "1.261", + "latency": "462.168", + "cost_per_test": "1.30437", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.261, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__bankruptcy_restructuring:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 88.235, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.235", + "raw_stderr": "1.899", + "latency": "570.801", + "cost_per_test": "2.356863", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.899, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__capital_markets:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 94.51, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "94.51", + "raw_stderr": "1.429", + "latency": "544.738", + "cost_per_test": "1.23096", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.429, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_governance:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 90.753, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.753", + "raw_stderr": "1.698", + "latency": "669.875", + "cost_per_test": "1.86509", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.698, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_ma:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 95.192, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "95.192", + "raw_stderr": "1.213", + "latency": "767.352", + "cost_per_test": "2.416866", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.213, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__data_privacy_cybersecurity:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 95.053, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "95.053", + "raw_stderr": "1.291", + "latency": "535.304", + "cost_per_test": "1.84691", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.291, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__emerging_companies_venture_capital:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 86.348, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.348", + "raw_stderr": "2.009", + "latency": "659.764", + "cost_per_test": "2.105221", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.009, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__employment_labor:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 85.393, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.393", + "raw_stderr": "2.165", + "latency": "565.668", + "cost_per_test": "1.185156", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.165, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__energy_natural_resources:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 95.05, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "95.05", + "raw_stderr": "1.248", + "latency": "685.122", + "cost_per_test": "3.669584", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.248, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__environmental_esg:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 91.398, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "91.398", + "raw_stderr": "1.682", + "latency": "765.468", + "cost_per_test": "2.260337", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.682, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__funds_asset_management:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 86.006, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.006", + "raw_stderr": "1.876", + "latency": "676.129", + "cost_per_test": "2.138169", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.876, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__healthcare_life_sciences:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 92.739, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "92.739", + "raw_stderr": "1.493", + "latency": "761.657", + "cost_per_test": "3.027345", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.493, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__immigration:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 89.091, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.091", + "raw_stderr": "1.883", + "latency": "539.648", + "cost_per_test": "1.422373", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.883, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__insurance:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 89.377, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.377", + "raw_stderr": "1.868", + "latency": "661.589", + "cost_per_test": "2.129824", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.868, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__intellectual_property:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 95.053, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "95.053", + "raw_stderr": "1.291", + "latency": "474.617", + "cost_per_test": "1.436571", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.291, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__international_trade_sanctions:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 89.682, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.682", + "raw_stderr": "1.92", + "latency": "534.978", + "cost_per_test": "1.312972", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.92, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__litigation_dispute_resolution:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 90.637, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.637", + "raw_stderr": "1.786", + "latency": "666.448", + "cost_per_test": "2.62856", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.786, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__real_estate:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 93.586, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "93.586", + "raw_stderr": "1.325", + "latency": "625.735", + "cost_per_test": "1.767673", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.325, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__structured_finance_securitization:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 85.079, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.079", + "raw_stderr": "2.011", + "latency": "578.882", + "cost_per_test": "1.902953", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.011, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__tax:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 84.281, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.281", + "raw_stderr": "2.108", + "latency": "531.482", + "cost_per_test": "1.829973", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.108, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__trusts_estates_private_client:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 90.84, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.84", + "raw_stderr": "1.786", + "latency": "613.216", + "cost_per_test": "1.84661", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.786, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__white_collar_defense_investigations:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 90.572, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.572", + "raw_stderr": "1.698", + "latency": "708.498", + "cost_per_test": "2.297215", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.698, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:data_privacy_cybersecurity:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "535.304", + "cost_per_test": "1.84691", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:emerging_companies_venture_capital:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "659.764", + "cost_per_test": "2.105221", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:employment_labor:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "565.668", + "cost_per_test": "1.185156", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:energy_natural_resources:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "685.122", + "cost_per_test": "3.669584", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:environmental_esg:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "765.468", + "cost_per_test": "2.260337", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:funds_asset_management:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "676.129", + "cost_per_test": "2.138169", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:healthcare_life_sciences:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 40.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "40", + "raw_stderr": "24.495", + "latency": "761.657", + "cost_per_test": "3.027345", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 24.495, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:immigration:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "539.648", + "cost_per_test": "1.422373", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:insurance:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "661.589", + "cost_per_test": "2.129824", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:intellectual_property:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "474.617", + "cost_per_test": "1.436571", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:international_trade_sanctions:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "534.978", + "cost_per_test": "1.312972", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:litigation_dispute_resolution:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "666.448", + "cost_per_test": "2.62856", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:overall:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.overall", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Overall).", + "metric_id": "vals_ai.hlab.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 12.917, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "12.917", + "raw_stderr": "2.534", + "latency": "614.713", + "cost_per_test": "2.01955", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.534, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:real_estate:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "625.735", + "cost_per_test": "1.767673", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:structured_finance_securitization:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "578.882", + "cost_per_test": "1.902953", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:tax:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "531.482", + "cost_per_test": "1.829973", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:trusts_estates_private_client:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "613.216", + "cost_per_test": "1.84661", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "hlab:white_collar_defense_investigations:grok/grok-4.5:score", + "evaluation_name": "vals_ai.hlab.white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "708.498", + "cost_per_test": "2.297215", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-4.5/f744c761-9fb2-4577-a935-c51a593374b8.json b/data/vals_ai/xai/grok-4.5/f744c761-9fb2-4577-a935-c51a593374b8.json new file mode 100644 index 0000000000000000000000000000000000000000..fc539bbcdd0670eb6c48f79408eb932e53ef21cf --- /dev/null +++ b/data/vals_ai/xai/grok-4.5/f744c761-9fb2-4577-a935-c51a593374b8.json @@ -0,0 +1,300 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/code-migration/xai_grok-4.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Code Migration", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-4.5", + "id": "xai/grok-4.5", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-4.5", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "code-migration:cli:grok/grok-4.5:score", + "evaluation_name": "vals_ai.code-migration.cli", + "source_data": { + "dataset_name": "Code Migration - CLI", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cli", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (CLI).", + "metric_id": "vals_ai.code-migration.cli.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 25.461, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cli", + "task_name": "CLI", + "dataset_type": "private", + "industry": "coding", + "raw_score": "25.461", + "raw_stderr": "2.134", + "latency": "1656.978", + "cost_per_test": "5.995314", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.134, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "code-migration:cobol:grok/grok-4.5:score", + "evaluation_name": "vals_ai.code-migration.cobol", + "source_data": { + "dataset_name": "Code Migration - COBOL", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cobol", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (COBOL).", + "metric_id": "vals_ai.code-migration.cobol.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 70.0, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cobol", + "task_name": "COBOL", + "dataset_type": "private", + "industry": "coding", + "raw_score": "70", + "raw_stderr": "15.275", + "latency": "664.305", + "cost_per_test": "1.414834", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 15.275, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "code-migration:code_quality:grok/grok-4.5:score", + "evaluation_name": "vals_ai.code-migration.code_quality", + "source_data": { + "dataset_name": "Code Migration - Code Quality", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "code_quality", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Code Quality).", + "metric_id": "vals_ai.code-migration.code_quality.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 4.078, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "code_quality", + "task_name": "Code Quality", + "dataset_type": "private", + "industry": "coding", + "raw_score": "4.078", + "raw_stderr": "0.042", + "latency": "1408.81", + "cost_per_test": "4.850194", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 0.042, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + }, + { + "evaluation_result_id": "code-migration:overall:grok/grok-4.5:score", + "evaluation_name": "vals_ai.code-migration.overall", + "source_data": { + "dataset_name": "Code Migration - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Overall).", + "metric_id": "vals_ai.code-migration.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 36.596, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "36.596", + "raw_stderr": "4.141", + "latency": "1408.81", + "cost_per_test": "4.850194", + "temperature": "0.7", + "top_p": "0.95", + "reasoning_effort": "high", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 4.141, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "high" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-build-0.1/692f7923-b809-40bf-86b2-499337932c86.json b/data/vals_ai/xai/grok-build-0.1/692f7923-b809-40bf-86b2-499337932c86.json new file mode 100644 index 0000000000000000000000000000000000000000..05441a8d36eb02cb56cff45e959d40a4cac3efa4 --- /dev/null +++ b/data/vals_ai/xai/grok-build-0.1/692f7923-b809-40bf-86b2-499337932c86.json @@ -0,0 +1,99 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/xai_grok-build-0.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-build-0.1", + "id": "xai/grok-build-0.1", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-build-0.1", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:grok/grok-build-0.1:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 13.349, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "13.349", + "raw_stderr": "3.819", + "latency": "394.638", + "temperature": "0.7", + "top_p": "0.95", + "max_output_tokens": "256000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 3.819, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 0.95, + "max_tokens": 256000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-code-fast-1/1ffed690-8158-40de-9cc7-45ae00424283.json b/data/vals_ai/xai/grok-code-fast-1/1ffed690-8158-40de-9cc7-45ae00424283.json new file mode 100644 index 0000000000000000000000000000000000000000..94c1d1f6ad13a67d96d7d86999be1d3d6eab6104 --- /dev/null +++ b/data/vals_ai/xai/grok-code-fast-1/1ffed690-8158-40de-9cc7-45ae00424283.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/xai_grok-code-fast-1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-code-fast-1", + "id": "xai/grok-code-fast-1", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-code-fast-1", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:grok/grok-code-fast-1:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 3.333, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "3.333", + "raw_stderr": "1.925", + "latency": "867.087", + "cost_per_test": "2.032726", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "40000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.925, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 40000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:grok/grok-code-fast-1:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 5.333, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "5.333", + "raw_stderr": "1.284", + "latency": "1300.636", + "cost_per_test": "1.203294", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "40000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.284, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 40000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:grok/grok-code-fast-1:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 4.333, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "4.333", + "raw_stderr": "1.157", + "latency": "1083.862", + "cost_per_test": "1.61801", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "40000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.157, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 40000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xai/grok-code-fast-1/563e7a91-7661-4552-9215-7b42b719a5e0.json b/data/vals_ai/xai/grok-code-fast-1/563e7a91-7661-4552-9215-7b42b719a5e0.json new file mode 100644 index 0000000000000000000000000000000000000000..5979f32f7df195f1807fd1fd823f39a64805c564 --- /dev/null +++ b/data/vals_ai/xai/grok-code-fast-1/563e7a91-7661-4552-9215-7b42b719a5e0.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/xai_grok-code-fast-1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "grok-code-fast-1", + "id": "xai/grok-code-fast-1", + "developer": "xai", + "additional_details": { + "vals_model_id": "grok/grok-code-fast-1", + "vals_provider": "SpaceXAI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:grok/grok-code-fast-1:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 95.342, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "95.342", + "raw_stderr": "1.174", + "latency": "5.086", + "cost_per_test": "0.00258", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "40000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.174, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 40000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:grok/grok-code-fast-1:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 26.857, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "26.857", + "raw_stderr": "2.369", + "latency": "53.943", + "cost_per_test": "0.016003", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "40000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.369, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 40000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:grok/grok-code-fast-1:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 63.708, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "63.708", + "raw_stderr": "2.457", + "latency": "18.484", + "cost_per_test": "0.007001", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "40000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 2.457, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 40000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:grok/grok-code-fast-1:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 61.969, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "61.969", + "raw_stderr": "1.203", + "latency": "25.838", + "cost_per_test": "0.008528", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "40000", + "provider": "SpaceXAI" + }, + "uncertainty": { + "standard_error": { + "value": 1.203, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 40000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2-flash/e831f2c0-789c-46ed-a8ab-87b555d68dfd.json b/data/vals_ai/xiaomi/mimo-v2-flash/e831f2c0-789c-46ed-a8ab-87b555d68dfd.json new file mode 100644 index 0000000000000000000000000000000000000000..d4e656aad40c900d61b482f02af74bd110816c21 --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2-flash/e831f2c0-789c-46ed-a8ab-87b555d68dfd.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/xiaomi_mimo-v2-flash/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2-flash", + "id": "xiaomi/mimo-v2-flash", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2-flash", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:xiaomi/mimo-v2-flash:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 59.091, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "59.091", + "raw_stderr": "3.507", + "latency": "20.883", + "cost_per_test": "0.001223", + "temperature": "0.3", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 3.507, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:xiaomi/mimo-v2-flash:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 59.596, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "59.596", + "raw_stderr": "3.553", + "latency": "20.921", + "cost_per_test": "0.0012", + "temperature": "0.3", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 3.553, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:xiaomi/mimo-v2-flash:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 59.344, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "59.344", + "raw_stderr": "2.496", + "latency": "20.902", + "cost_per_test": "0.001212", + "temperature": "0.3", + "top_p": "0.95", + "max_output_tokens": "64000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 2.496, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.3, + "top_p": 0.95, + "max_tokens": 64000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5-pro/25e07f1e-a9e4-4ad2-807f-353d240ce16e.json b/data/vals_ai/xiaomi/mimo-v2.5-pro/25e07f1e-a9e4-4ad2-807f-353d240ce16e.json new file mode 100644 index 0000000000000000000000000000000000000000..d06d2e99bb2b319e71f19b1a7a0bc520b0d4d484 --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5-pro/25e07f1e-a9e4-4ad2-807f-353d240ce16e.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/xiaomi_mimo-v2.5-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5-pro", + "id": "xiaomi/mimo-v2.5-pro", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5-pro", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.469, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.469", + "raw_stderr": "1.851", + "latency": "37.78", + "cost_per_test": "0.001291", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.851, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.987, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.987", + "raw_stderr": "1.479", + "latency": "72.573", + "cost_per_test": "0.002321", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.479, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.459, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.459", + "raw_stderr": "1.038", + "latency": "108.606", + "cost_per_test": "0.003601", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.038, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.049, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.049", + "raw_stderr": "2.268", + "latency": "58.443", + "cost_per_test": "0.001837", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 2.268, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.915, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.915", + "raw_stderr": "1.675", + "latency": "42.503", + "cost_per_test": "0.001218", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.675, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.115, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.115", + "raw_stderr": "1.375", + "latency": "217.255", + "cost_per_test": "0.00665", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.375, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.218, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.218", + "raw_stderr": "1.748", + "latency": "35.649", + "cost_per_test": "0.000991", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.748, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.64, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.64", + "raw_stderr": "2.526", + "latency": "39.88", + "cost_per_test": "0.001114", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 2.526, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 66.939, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.939", + "raw_stderr": "1.507", + "latency": "74.224", + "cost_per_test": "0.002031", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.507, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.634, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.634", + "raw_stderr": "0.86", + "latency": "71.077", + "cost_per_test": "0.002403", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.86, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.628, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.628", + "raw_stderr": "1.582", + "latency": "36.392", + "cost_per_test": "0.000878", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.582, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.595, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.595", + "raw_stderr": "0.453", + "latency": "67.757", + "cost_per_test": "0.002085", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.453, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.96, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.96", + "raw_stderr": "2.116", + "latency": "32.952", + "cost_per_test": "0.000991", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 2.116, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.224, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.224", + "raw_stderr": "1.052", + "latency": "93.164", + "cost_per_test": "0.003086", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.052, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.09, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.09", + "raw_stderr": "1.769", + "latency": "28.102", + "cost_per_test": "0.000782", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.769, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5-pro/2ce3fb9d-ee9a-4e87-a653-4004758b06ed.json b/data/vals_ai/xiaomi/mimo-v2.5-pro/2ce3fb9d-ee9a-4e87-a653-4004758b06ed.json new file mode 100644 index 0000000000000000000000000000000000000000..2cfd4f17402e178470f1d60df016a68e3b3edc70 --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5-pro/2ce3fb9d-ee9a-4e87-a653-4004758b06ed.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/xiaomi_mimo-v2.5-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5-pro", + "id": "xiaomi/mimo-v2.5-pro", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5-pro", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 75.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "75", + "raw_stderr": "25", + "latency": "19274.134", + "cost_per_test": "0.970383", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 25.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 40.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "40", + "raw_stderr": "9.097", + "latency": "2569.885", + "cost_per_test": "0.129384", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 9.097, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 65.455, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "65.455", + "raw_stderr": "6.471", + "latency": "1401.755", + "cost_per_test": "0.070573", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 6.471, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 57.303, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "57.303", + "raw_stderr": "5.273", + "latency": "866.253", + "cost_per_test": "0.043613", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 5.273, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5-pro/3c231712-735f-4e48-adc9-d4f98d2b5e59.json b/data/vals_ai/xiaomi/mimo-v2.5-pro/3c231712-735f-4e48-adc9-d4f98d2b5e59.json new file mode 100644 index 0000000000000000000000000000000000000000..481f7deb1dd691dc1b51764b048433ad1a7951ca --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5-pro/3c231712-735f-4e48-adc9-d4f98d2b5e59.json @@ -0,0 +1,438 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/xiaomi_mimo-v2.5-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5-pro", + "id": "xiaomi/mimo-v2.5-pro", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5-pro", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 88.031, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.031", + "raw_stderr": "0.842", + "latency": "21.462", + "cost_per_test": "0.000694", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.842, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.497, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.497", + "raw_stderr": "0.223", + "latency": "31.877", + "cost_per_test": "0.000979", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.223, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.173, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.173", + "raw_stderr": "0.769", + "latency": "32.35", + "cost_per_test": "0.00105", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.769, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.129, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.129", + "raw_stderr": "0.55", + "latency": "27.142", + "cost_per_test": "0.000845", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.55, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 64.964, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "64.964", + "raw_stderr": "1.677", + "latency": "26.638", + "cost_per_test": "0.000894", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.677, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 72.98, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "72.98", + "raw_stderr": "1.841", + "latency": "23.385", + "cost_per_test": "0.000606", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.841, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5-pro/4587c69f-a1dc-450c-af3b-7cb17c313e3c.json b/data/vals_ai/xiaomi/mimo-v2.5-pro/4587c69f-a1dc-450c-af3b-7cb17c313e3c.json new file mode 100644 index 0000000000000000000000000000000000000000..3af579694d6c4a6721d142c3b4b02e51c5faad4d --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5-pro/4587c69f-a1dc-450c-af3b-7cb17c313e3c.json @@ -0,0 +1,408 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/xiaomi_mimo-v2.5-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5-pro", + "id": "xiaomi/mimo-v2.5-pro", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5-pro", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 62.471, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "62.471", + "raw_stderr": "1.653", + "latency": "16.623", + "cost_per_test": "0.029339", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.653, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 41.243, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "41.243", + "raw_stderr": "1.36", + "latency": "391.709", + "cost_per_test": "0.190978", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.36, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:overall:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 50.744, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "50.744", + "raw_stderr": "1.626", + "latency": "839.412", + "cost_per_test": "0.091426", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.626, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "66.667", + "raw_stderr": "4.668", + "latency": "518.711", + "cost_per_test": "0.026452" + }, + "uncertainty": { + "standard_error": { + "value": 4.668, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 57.303, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "57.303", + "raw_stderr": "5.273", + "latency": "866.253", + "cost_per_test": "0.043613", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 5.273, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 36.321, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "36.321", + "raw_stderr": "6.374", + "latency": "2403.764", + "cost_per_test": "0.166747", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 6.374, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5-pro/59dda77e-5278-4004-a329-1bfc92e772bf.json b/data/vals_ai/xiaomi/mimo-v2.5-pro/59dda77e-5278-4004-a329-1bfc92e772bf.json new file mode 100644 index 0000000000000000000000000000000000000000..59863d3e28bfeda20b01ad4f825b854e6b8f887e --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5-pro/59dda77e-5278-4004-a329-1bfc92e772bf.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/xiaomi_mimo-v2.5-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5-pro", + "id": "xiaomi/mimo-v2.5-pro", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5-pro", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 34.113, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "34.113", + "raw_stderr": "4.529", + "latency": "2403.764", + "cost_per_test": "0.166747", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 4.529, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5-pro/6564c020-b6b4-4025-94f4-1595302aa250.json b/data/vals_ai/xiaomi/mimo-v2.5-pro/6564c020-b6b4-4025-94f4-1595302aa250.json new file mode 100644 index 0000000000000000000000000000000000000000..6f7b706984cb6a3aaf32e240b5acb8feaf772a53 --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5-pro/6564c020-b6b4-4025-94f4-1595302aa250.json @@ -0,0 +1,676 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/emb/xiaomi_mimo-v2.5-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - EMB", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5-pro", + "id": "xiaomi/mimo-v2.5-pro", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5-pro", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "emb:comps:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.emb.comps", + "source_data": { + "dataset_name": "EMB - Comps", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "comps", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Comps).", + "metric_id": "vals_ai.emb.comps.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 47.341, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "comps", + "task_name": "Comps", + "dataset_type": "private", + "industry": "finance", + "raw_score": "47.341", + "raw_stderr": "7.039", + "latency": "1597.315", + "cost_per_test": "0.399754", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 7.039, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:dataroom:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.emb.dataroom", + "source_data": { + "dataset_name": "EMB - Dataroom", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dataroom", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Dataroom).", + "metric_id": "vals_ai.emb.dataroom.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 77.965, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dataroom", + "task_name": "Dataroom", + "dataset_type": "private", + "industry": "finance", + "raw_score": "77.965", + "raw_stderr": "6.068", + "latency": "1336.466", + "cost_per_test": "0.144435", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 6.068, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:dcf:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.emb.dcf", + "source_data": { + "dataset_name": "EMB - DCF", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dcf", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (DCF).", + "metric_id": "vals_ai.emb.dcf.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 43.63, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dcf", + "task_name": "DCF", + "dataset_type": "private", + "industry": "finance", + "raw_score": "43.63", + "raw_stderr": "7.728", + "latency": "1961.142", + "cost_per_test": "0.219228", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 7.728, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:lbo:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.emb.lbo", + "source_data": { + "dataset_name": "EMB - LBO", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "lbo", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (LBO).", + "metric_id": "vals_ai.emb.lbo.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 47.739, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "lbo", + "task_name": "LBO", + "dataset_type": "private", + "industry": "finance", + "raw_score": "47.739", + "raw_stderr": "5.122", + "latency": "2198.075", + "cost_per_test": "0.234191", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 5.122, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:ma:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.emb.ma", + "source_data": { + "dataset_name": "EMB - M&A", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "ma", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (M&A).", + "metric_id": "vals_ai.emb.ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 64.098, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "ma", + "task_name": "M&A", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.098", + "raw_stderr": "7.421", + "latency": "1662.395", + "cost_per_test": "0.166174", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 7.421, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:misc:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.emb.misc", + "source_data": { + "dataset_name": "EMB - Misc", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "misc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Misc).", + "metric_id": "vals_ai.emb.misc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 50.644, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "misc", + "task_name": "Misc", + "dataset_type": "private", + "industry": "finance", + "raw_score": "50.644", + "raw_stderr": "6.183", + "latency": "1867.773", + "cost_per_test": "0.178065", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 6.183, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:operating:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.emb.operating", + "source_data": { + "dataset_name": "EMB - Operating", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "operating", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Operating).", + "metric_id": "vals_ai.emb.operating.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 55.182, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "operating", + "task_name": "Operating", + "dataset_type": "private", + "industry": "finance", + "raw_score": "55.182", + "raw_stderr": "7.443", + "latency": "1607.152", + "cost_per_test": "0.179455", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 7.443, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:overall:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.emb.overall", + "source_data": { + "dataset_name": "EMB - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Overall).", + "metric_id": "vals_ai.emb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 55.228, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "55.228", + "raw_stderr": "3.112", + "latency": "1747.98", + "cost_per_test": "0.221264", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 3.112, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:scratch:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.emb.scratch", + "source_data": { + "dataset_name": "EMB - Scratch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "scratch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Scratch).", + "metric_id": "vals_ai.emb.scratch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 57.916, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "scratch", + "task_name": "Scratch", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.916", + "raw_stderr": "2.79", + "latency": "1753.73", + "cost_per_test": "0.210941", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 2.79, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "emb:template:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.emb.template", + "source_data": { + "dataset_name": "EMB - Template", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "template", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Template).", + "metric_id": "vals_ai.emb.template.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 52.116, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "template", + "task_name": "Template", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.116", + "raw_stderr": "4.677", + "latency": "1742.23", + "cost_per_test": "0.231587", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 4.677, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5-pro/80c228c0-615d-4188-9a00-53ff9cbe6da9.json b/data/vals_ai/xiaomi/mimo-v2.5-pro/80c228c0-615d-4188-9a00-53ff9cbe6da9.json new file mode 100644 index 0000000000000000000000000000000000000000..ec1f42373c7524e4808e1afbdd81daf61b6d952f --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5-pro/80c228c0-615d-4188-9a00-53ff9cbe6da9.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/xiaomi_mimo-v2.5-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5-pro", + "id": "xiaomi/mimo-v2.5-pro", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5-pro", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.373, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.373", + "raw_stderr": "1.67", + "latency": "11.967", + "cost_per_test": "0.00107", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.67, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.422, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.422", + "raw_stderr": "1.662", + "latency": "28.107", + "cost_per_test": "0.081304", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.662, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.422, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.422", + "raw_stderr": "0.959", + "latency": "18.899", + "cost_per_test": "0.037238", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.959, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 62.471, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.471", + "raw_stderr": "1.653", + "latency": "16.623", + "cost_per_test": "0.029339", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.653, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5-pro/9a04c661-d178-4b71-9c4f-388294093a7d.json b/data/vals_ai/xiaomi/mimo-v2.5-pro/9a04c661-d178-4b71-9c4f-388294093a7d.json new file mode 100644 index 0000000000000000000000000000000000000000..94a45d4dfba714efe9afac280bfaed9c9cee96ee --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5-pro/9a04c661-d178-4b71-9c4f-388294093a7d.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/xiaomi_mimo-v2.5-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5-pro", + "id": "xiaomi/mimo-v2.5-pro", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5-pro", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 80.808, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.808", + "raw_stderr": "3.429", + "latency": "237.64", + "cost_per_test": "0.006645", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 3.429, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 84.343, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.343", + "raw_stderr": "3.55", + "latency": "265.025", + "cost_per_test": "0.006663", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 3.55, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 82.576, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.576", + "raw_stderr": "2.468", + "latency": "251.332", + "cost_per_test": "0.006654", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 2.468, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5-pro/a1d6d5df-da7e-4de1-8090-f8251f8feac0.json b/data/vals_ai/xiaomi/mimo-v2.5-pro/a1d6d5df-da7e-4de1-8090-f8251f8feac0.json new file mode 100644 index 0000000000000000000000000000000000000000..fbc3eb61c809bbe66b7eed06f127ca1f89abafd3 --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5-pro/a1d6d5df-da7e-4de1-8090-f8251f8feac0.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/xiaomi_mimo-v2.5-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5-pro", + "id": "xiaomi/mimo-v2.5-pro", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5-pro", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 24.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "24", + "raw_stderr": "4.292", + "latency": "1267.15", + "cost_per_test": "0.071554", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 4.292, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5-pro/b9828fda-6ba7-42d7-abfc-f8e178e8e3fe.json b/data/vals_ai/xiaomi/mimo-v2.5-pro/b9828fda-6ba7-42d7-abfc-f8e178e8e3fe.json new file mode 100644 index 0000000000000000000000000000000000000000..171a603e8beca1bcbb578249c0d4ba4a0890d79b --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5-pro/b9828fda-6ba7-42d7-abfc-f8e178e8e3fe.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/xiaomi_mimo-v2.5-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5-pro", + "id": "xiaomi/mimo-v2.5-pro", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5-pro", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 83.73, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "83.73", + "raw_stderr": "2.063", + "latency": "90.71", + "cost_per_test": "0.00603", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 2.063, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5-pro/bf6dfb5f-9ee2-4a5f-b8e4-79044ddbc672.json b/data/vals_ai/xiaomi/mimo-v2.5-pro/bf6dfb5f-9ee2-4a5f-b8e4-79044ddbc672.json new file mode 100644 index 0000000000000000000000000000000000000000..66e243cf182c144f220e149618f93865b482dc50 --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5-pro/bf6dfb5f-9ee2-4a5f-b8e4-79044ddbc672.json @@ -0,0 +1,371 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/xiaomi_mimo-v2.5-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5-pro", + "id": "xiaomi/mimo-v2.5-pro", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5-pro", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "7.362", + "latency": "789.528", + "cost_per_test": "0.041415", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 7.362, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 74.713, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "74.713", + "raw_stderr": "2.696", + "latency": "533.824", + "cost_per_test": "0.026309", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 2.696, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 82.474, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "82.474", + "raw_stderr": "2.737", + "latency": "309.775", + "cost_per_test": "0.014076", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 2.737, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "833.426", + "cost_per_test": "0.043185", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 74.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "74", + "raw_stderr": "1.964", + "latency": "470.814", + "cost_per_test": "0.022968", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.964, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5-pro/cf7dc32f-8b66-4841-8bea-ae5424baa02d.json b/data/vals_ai/xiaomi/mimo-v2.5-pro/cf7dc32f-8b66-4841-8bea-ae5424baa02d.json new file mode 100644 index 0000000000000000000000000000000000000000..22748ebde55fddaa0f5e41b308fb61690dce4585 --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5-pro/cf7dc32f-8b66-4841-8bea-ae5424baa02d.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/code-migration/xiaomi_mimo-v2.5-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Code Migration", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5-pro", + "id": "xiaomi/mimo-v2.5-pro", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5-pro", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "code-migration:cli:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.code-migration.cli", + "source_data": { + "dataset_name": "Code Migration - CLI", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cli", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (CLI).", + "metric_id": "vals_ai.code-migration.cli.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 13.981, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cli", + "task_name": "CLI", + "dataset_type": "private", + "industry": "coding", + "raw_score": "13.981", + "raw_stderr": "1.688", + "latency": "3349.47", + "cost_per_test": "0.268603", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.688, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "code-migration:cobol:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.code-migration.cobol", + "source_data": { + "dataset_name": "Code Migration - COBOL", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cobol", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (COBOL).", + "metric_id": "vals_ai.code-migration.cobol.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 44.286, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cobol", + "task_name": "COBOL", + "dataset_type": "private", + "industry": "coding", + "raw_score": "44.286", + "raw_stderr": "15.714", + "latency": "1683.159", + "cost_per_test": "0.147392", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 15.714, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "code-migration:code_quality:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.code-migration.code_quality", + "source_data": { + "dataset_name": "Code Migration - Code Quality", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "code_quality", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Code Quality).", + "metric_id": "vals_ai.code-migration.code_quality.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 3.382, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "code_quality", + "task_name": "Code Quality", + "dataset_type": "private", + "industry": "coding", + "raw_score": "3.382", + "raw_stderr": "0.061", + "latency": "2932.892", + "cost_per_test": "0.2383", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.061, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "code-migration:overall:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.code-migration.overall", + "source_data": { + "dataset_name": "Code Migration - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Overall).", + "metric_id": "vals_ai.code-migration.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 21.557, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "21.557", + "raw_stderr": "4.127", + "latency": "2932.892", + "cost_per_test": "0.2383", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 4.127, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5-pro/d8b4b616-614b-48c8-956b-7398ca7041b0.json b/data/vals_ai/xiaomi/mimo-v2.5-pro/d8b4b616-614b-48c8-956b-7398ca7041b0.json new file mode 100644 index 0000000000000000000000000000000000000000..ed6d5102c9dba2b1580b5a4ee1d2efd77ff67f08 --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5-pro/d8b4b616-614b-48c8-956b-7398ca7041b0.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/xiaomi_mimo-v2.5-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5-pro", + "id": "xiaomi/mimo-v2.5-pro", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5-pro", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 99.068, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "99.068", + "raw_stderr": "0.535", + "latency": "43.905", + "cost_per_test": "0.001487", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 58.571, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "58.571", + "raw_stderr": "2.633", + "latency": "562.035", + "cost_per_test": "0.01277", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 2.633, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 86.423, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "86.423", + "raw_stderr": "1.75", + "latency": "207.281", + "cost_per_test": "0.006168", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.75, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 81.354, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "81.354", + "raw_stderr": "1.069", + "latency": "271.074", + "cost_per_test": "0.006808", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.069, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5-pro/de9ee98d-d534-4dc3-a393-e2397cb0a594.json b/data/vals_ai/xiaomi/mimo-v2.5-pro/de9ee98d-d534-4dc3-a393-e2397cb0a594.json new file mode 100644 index 0000000000000000000000000000000000000000..e9241e515e14d36be954d8e0ab303566e44f21d3 --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5-pro/de9ee98d-d534-4dc3-a393-e2397cb0a594.json @@ -0,0 +1,740 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/xiaomi_mimo-v2.5-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5-pro", + "id": "xiaomi/mimo-v2.5-pro", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5-pro", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 34.386, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "34.386", + "raw_stderr": "1.658", + "latency": "489.296", + "cost_per_test": "0.206068", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.658, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 28.848, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "28.848", + "raw_stderr": "1.247", + "latency": "497.373", + "cost_per_test": "0.206445", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.247, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 40.15, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "40.15", + "raw_stderr": "1.84", + "latency": "538.572", + "cost_per_test": "0.316099", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.84, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 53.148, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "53.148", + "raw_stderr": "1.954", + "latency": "454.889", + "cost_per_test": "0.249212", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.954, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 54.614, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "54.614", + "raw_stderr": "4.256", + "latency": "381.5", + "cost_per_test": "0.066769", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 4.256, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 9.876, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "9.876", + "raw_stderr": "1.368", + "latency": "786.146", + "cost_per_test": "0.204645", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.368, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 50.768, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "50.768", + "raw_stderr": "4.219", + "latency": "426.85", + "cost_per_test": "0.210443", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 4.219, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 61.049, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.049", + "raw_stderr": "2.424", + "latency": "389.913", + "cost_per_test": "0.219268", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 2.424, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 48.971, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "48.971", + "raw_stderr": "0.391", + "latency": "328.815", + "cost_per_test": "0.094477", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.391, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:overall:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 41.502, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "41.502", + "raw_stderr": "1.193", + "latency": "497.373", + "cost_per_test": "0.206445", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.193, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 20.558, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "20.558", + "raw_stderr": "2.335", + "latency": "704.058", + "cost_per_test": "0.298475", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 2.335, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5-pro/e210df21-dba5-4557-85a0-167bec2997fc.json b/data/vals_ai/xiaomi/mimo-v2.5-pro/e210df21-dba5-4557-85a0-167bec2997fc.json new file mode 100644 index 0000000000000000000000000000000000000000..e5b5bf46c414815b94eff09eeca7b9b7b22dddb4 --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5-pro/e210df21-dba5-4557-85a0-167bec2997fc.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/xiaomi_mimo-v2.5-pro/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5-pro", + "id": "xiaomi/mimo-v2.5-pro", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5-pro", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 63.123, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.123", + "raw_stderr": "1.38", + "latency": "62.829", + "cost_per_test": "0.002547", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.38, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 73.794, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "73.794", + "raw_stderr": "0.863", + "latency": "62.829", + "cost_per_test": "0.002547", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.863, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:xiaomi/mimo-v2.5-pro:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 84.464, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "84.464", + "raw_stderr": "1.036", + "latency": "62.829", + "cost_per_test": "0.002547", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.036, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5/04827096-c43b-48dc-85a5-559840bc7b65.json b/data/vals_ai/xiaomi/mimo-v2.5/04827096-c43b-48dc-85a5-559840bc7b65.json new file mode 100644 index 0000000000000000000000000000000000000000..ce7831d2ff0d581208c382c875163d1d573675f9 --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5/04827096-c43b-48dc-85a5-559840bc7b65.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/xiaomi_mimo-v2.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5", + "id": "xiaomi/mimo-v2.5", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 16.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "16", + "raw_stderr": "3.685", + "latency": "1943.239", + "cost_per_test": "0.074885", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 3.685, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5/32e4cf16-fbac-4c3a-9809-df49672b1150.json b/data/vals_ai/xiaomi/mimo-v2.5/32e4cf16-fbac-4c3a-9809-df49672b1150.json new file mode 100644 index 0000000000000000000000000000000000000000..832c15c9504152c2b1218ffb784bc227b3311af2 --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5/32e4cf16-fbac-4c3a-9809-df49672b1150.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/xiaomi_mimo-v2.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5", + "id": "xiaomi/mimo-v2.5", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 59.441, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.441", + "raw_stderr": "1.676", + "latency": "5.643", + "cost_per_test": "0.00025", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.676, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.14, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.14", + "raw_stderr": "1.672", + "latency": "27.898", + "cost_per_test": "0.026197", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.672, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 59.907, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.907", + "raw_stderr": "0.966", + "latency": "17.625", + "cost_per_test": "0.011733", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.966, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.14, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.14", + "raw_stderr": "1.672", + "latency": "19.334", + "cost_per_test": "0.008752", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.672, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5/3dc2c779-fec7-46fd-9cd0-39bf96d302ef.json b/data/vals_ai/xiaomi/mimo-v2.5/3dc2c779-fec7-46fd-9cd0-39bf96d302ef.json new file mode 100644 index 0000000000000000000000000000000000000000..b36ce6de108ac59e9e88b95e3b195de758cc0f28 --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5/3dc2c779-fec7-46fd-9cd0-39bf96d302ef.json @@ -0,0 +1,536 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_multimodal_index/xiaomi_mimo-v2.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Multimodal Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5", + "id": "xiaomi/mimo-v2.5", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_multimodal_index:corp_fin_v2:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Multimodal Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (CorpFin v2).", + "metric_id": "vals_ai.vals_multimodal_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 60.14, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "60.14", + "raw_stderr": "1.672", + "latency": "19.334", + "cost_per_test": "0.008752", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.672, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:finance_agent:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.finance_agent", + "source_data": { + "dataset_name": "Vals Multimodal Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_multimodal_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 37.61, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "37.61", + "raw_stderr": "1.198", + "latency": "267.315", + "cost_per_test": "0.088638", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.198, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:mortgage_tax:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.mortgage_tax", + "source_data": { + "dataset_name": "Vals Multimodal Index - Mortgage Tax", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "mortgage_tax", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Mortgage Tax).", + "metric_id": "vals_ai.vals_multimodal_index.mortgage_tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 59.26, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "mortgage_tax", + "task_name": "Mortgage Tax", + "dataset_type": "private", + "industry": "index", + "raw_score": "59.26", + "raw_stderr": "0.979", + "latency": "32.939", + "cost_per_test": "0.001523", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.979, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:overall:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.overall", + "source_data": { + "dataset_name": "Vals Multimodal Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Overall).", + "metric_id": "vals_ai.vals_multimodal_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 52.767, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "52.767", + "raw_stderr": "1.419", + "latency": "430.314", + "cost_per_test": "0.029291", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.419, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:sage:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.sage", + "source_data": { + "dataset_name": "Vals Multimodal Index - SAGE", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "sage", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SAGE).", + "metric_id": "vals_ai.vals_multimodal_index.sage.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 43.265, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "sage", + "task_name": "SAGE", + "dataset_type": "private", + "industry": "index", + "raw_score": "43.265", + "raw_stderr": "3.387", + "latency": "52.66", + "cost_per_test": "0.004398", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 3.387, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:swebench:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.swebench", + "source_data": { + "dataset_name": "Vals Multimodal Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (SWE-bench).", + "metric_id": "vals_ai.vals_multimodal_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 64.706, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "64.706", + "raw_stderr": "4.732", + "latency": "284.365", + "cost_per_test": "0.011672" + }, + "uncertainty": { + "standard_error": { + "value": 4.732, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:terminal_bench_2_1:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Multimodal Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_multimodal_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 60.674, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "60.674", + "raw_stderr": "5.207", + "latency": "775.239", + "cost_per_test": "0.021658", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 5.207, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_multimodal_index:vibe_code_bench:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.vals_multimodal_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Multimodal Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_multimodal_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Multimodal Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_multimodal_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_multimodal_index" + } + }, + "score_details": { + "score": 48.148, + "details": { + "benchmark_slug": "vals_multimodal_index", + "benchmark_name": "Vals Multimodal Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "48.148", + "raw_stderr": "6.086", + "latency": "1580.348", + "cost_per_test": "0.068395", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 6.086, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5/3e75494c-5164-466b-a449-fdb2ebdbafa1.json b/data/vals_ai/xiaomi/mimo-v2.5/3e75494c-5164-466b-a449-fdb2ebdbafa1.json new file mode 100644 index 0000000000000000000000000000000000000000..6406b79a17a4385f8ea2823e956b065b7fbe3fa7 --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5/3e75494c-5164-466b-a449-fdb2ebdbafa1.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/xiaomi_mimo-v2.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5", + "id": "xiaomi/mimo-v2.5", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.447, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.447", + "raw_stderr": "0.689", + "latency": "34.354", + "cost_per_test": "0.000472", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.689, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 59.143, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "59.143", + "raw_stderr": "2.628", + "latency": "191.375", + "cost_per_test": "0.004286", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 2.628, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 86.945, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "86.945", + "raw_stderr": "1.722", + "latency": "100.091", + "cost_per_test": "0.00197", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.722, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 81.512, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "81.512", + "raw_stderr": "1.072", + "latency": "108.607", + "cost_per_test": "0.002243", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.072, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5/453290a8-c4cd-4949-95c5-a63c2d7c6aa3.json b/data/vals_ai/xiaomi/mimo-v2.5/453290a8-c4cd-4949-95c5-a63c2d7c6aa3.json new file mode 100644 index 0000000000000000000000000000000000000000..6a9995e12fc1d5113c68c398d09f6c4374458d97 --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5/453290a8-c4cd-4949-95c5-a63c2d7c6aa3.json @@ -0,0 +1,103 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmmu/xiaomi_mimo-v2.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMMU", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5", + "id": "xiaomi/mimo-v2.5", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmmu:overall:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.mmmu.overall", + "source_data": { + "dataset_name": "MMMU - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmmu" + ], + "additional_details": { + "benchmark_slug": "mmmu", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMMU (Overall).", + "metric_id": "vals_ai.mmmu.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmmu" + } + }, + "score_details": { + "score": 80.0, + "details": { + "benchmark_slug": "mmmu", + "benchmark_name": "MMMU", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80", + "raw_stderr": "0.962", + "latency": "42.47", + "cost_per_test": "0.001396", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.962, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5/49b5465f-ea28-41c8-bcb3-95493964cfaf.json b/data/vals_ai/xiaomi/mimo-v2.5/49b5465f-ea28-41c8-bcb3-95493964cfaf.json new file mode 100644 index 0000000000000000000000000000000000000000..0ce644eab516fcb1918e425269923972ba4524f5 --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5/49b5465f-ea28-41c8-bcb3-95493964cfaf.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/xiaomi_mimo-v2.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5", + "id": "xiaomi/mimo-v2.5", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.237, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.237", + "raw_stderr": "1.108", + "latency": "18.026", + "cost_per_test": "0.000507", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.108, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.227, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.227", + "raw_stderr": "1.104", + "latency": "31.297", + "cost_per_test": "0.000852", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.104, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.841, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.841", + "raw_stderr": "0.898", + "latency": "42.513", + "cost_per_test": "0.001299", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.898, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.585, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.585", + "raw_stderr": "1.683", + "latency": "45.123", + "cost_per_test": "0.000767", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.683, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.73, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.73", + "raw_stderr": "1.177", + "latency": "18.403", + "cost_per_test": "0.000494", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.177, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.011, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.011", + "raw_stderr": "1.263", + "latency": "77.878", + "cost_per_test": "0.00212", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.263, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.362, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.362", + "raw_stderr": "1.443", + "latency": "17.843", + "cost_per_test": "0.000458", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.443, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 68.241, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "68.241", + "raw_stderr": "2.39", + "latency": "16.687", + "cost_per_test": "0.000427", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 2.39, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 66.757, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "66.757", + "raw_stderr": "1.42", + "latency": "37.066", + "cost_per_test": "0.000908", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.42, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.004, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.004", + "raw_stderr": "0.653", + "latency": "27.882", + "cost_per_test": "0.000807", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.653, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.978, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.978", + "raw_stderr": "1.324", + "latency": "18.197", + "cost_per_test": "0.000405", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.324, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.931, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.931", + "raw_stderr": "0.372", + "latency": "30.248", + "cost_per_test": "0.000774", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.372, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.758, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.758", + "raw_stderr": "1.837", + "latency": "20.433", + "cost_per_test": "0.000429", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.837, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.223, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.223", + "raw_stderr": "0.835", + "latency": "36.498", + "cost_per_test": "0.001022", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.835, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.078, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.078", + "raw_stderr": "1.387", + "latency": "15.633", + "cost_per_test": "0.000348", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.387, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5/684e580a-3743-41b7-92e2-1c103111c0fe.json b/data/vals_ai/xiaomi/mimo-v2.5/684e580a-3743-41b7-92e2-1c103111c0fe.json new file mode 100644 index 0000000000000000000000000000000000000000..0de68e33a1ef5a9e73430c02597f84b156564625 --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5/684e580a-3743-41b7-92e2-1c103111c0fe.json @@ -0,0 +1,371 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/xiaomi_mimo-v2.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5", + "id": "xiaomi/mimo-v2.5", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "7.362", + "latency": "456.781", + "cost_per_test": "0.02012", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 7.362, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 70.498, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "70.498", + "raw_stderr": "2.828", + "latency": "273.979", + "cost_per_test": "0.01136", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 2.828, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 80.928, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "80.928", + "raw_stderr": "2.828", + "latency": "152.723", + "cost_per_test": "0.005928", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 2.828, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "507.374", + "cost_per_test": "0.015828", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 71.0, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "71", + "raw_stderr": "2.031", + "latency": "244.053", + "cost_per_test": "0.010032", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 2.031, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5/73dd3e08-7015-44e6-8e51-ee47aa992be2.json b/data/vals_ai/xiaomi/mimo-v2.5/73dd3e08-7015-44e6-8e51-ee47aa992be2.json new file mode 100644 index 0000000000000000000000000000000000000000..6adcdf54a63def42435474c7fb062334824f44b8 --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5/73dd3e08-7015-44e6-8e51-ee47aa992be2.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/sage/xiaomi_mimo-v2.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SAGE", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "education", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5", + "id": "xiaomi/mimo-v2.5", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "sage:PDEs:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.sage.PDEs", + "source_data": { + "dataset_name": "SAGE - Partial Differential Equations", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "PDEs", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Partial Differential Equations).", + "metric_id": "vals_ai.sage.PDEs.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 41.27, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "PDEs", + "task_name": "Partial Differential Equations", + "dataset_type": "private", + "industry": "education", + "raw_score": "41.27", + "raw_stderr": "6.185", + "latency": "64.976", + "cost_per_test": "0.003898", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 6.185, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:automata:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.sage.automata", + "source_data": { + "dataset_name": "SAGE - Automata", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "automata", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Automata).", + "metric_id": "vals_ai.sage.automata.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 29.762, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "automata", + "task_name": "Automata", + "dataset_type": "private", + "industry": "education", + "raw_score": "29.762", + "raw_stderr": "5.834", + "latency": "50.802", + "cost_per_test": "0.002438", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 5.834, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:linear_algebra:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.sage.linear_algebra", + "source_data": { + "dataset_name": "SAGE - Linear Algebra", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "linear_algebra", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Linear Algebra).", + "metric_id": "vals_ai.sage.linear_algebra.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 58.762, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "linear_algebra", + "task_name": "Linear Algebra", + "dataset_type": "private", + "industry": "education", + "raw_score": "58.762", + "raw_stderr": "5.566", + "latency": "42.202", + "cost_per_test": "0.006857", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 5.566, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "sage:overall:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.sage.overall", + "source_data": { + "dataset_name": "SAGE - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "sage", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SAGE (Overall).", + "metric_id": "vals_ai.sage.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/sage" + } + }, + "score_details": { + "score": 43.265, + "details": { + "benchmark_slug": "sage", + "benchmark_name": "SAGE", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "education", + "raw_score": "43.265", + "raw_stderr": "3.387", + "latency": "52.66", + "cost_per_test": "0.004398", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 3.387, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5/74eb2ce1-951e-4a2c-a4ee-473796bb5db1.json b/data/vals_ai/xiaomi/mimo-v2.5/74eb2ce1-951e-4a2c-a4ee-473796bb5db1.json new file mode 100644 index 0000000000000000000000000000000000000000..753d0418d5cc671498bac02cbf3f910db35d6dcd --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5/74eb2ce1-951e-4a2c-a4ee-473796bb5db1.json @@ -0,0 +1,408 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/xiaomi_mimo-v2.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5", + "id": "xiaomi/mimo-v2.5", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 60.14, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "60.14", + "raw_stderr": "1.672", + "latency": "19.334", + "cost_per_test": "0.008752", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.672, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 37.61, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "37.61", + "raw_stderr": "1.198", + "latency": "267.315", + "cost_per_test": "0.088638", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.198, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:overall:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 51.57, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "51.57", + "raw_stderr": "1.569", + "latency": "585.32", + "cost_per_test": "0.039823", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.569, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 64.706, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "64.706", + "raw_stderr": "4.732", + "latency": "284.365", + "cost_per_test": "0.011672" + }, + "uncertainty": { + "standard_error": { + "value": 4.732, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 60.674, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "60.674", + "raw_stderr": "5.207", + "latency": "775.239", + "cost_per_test": "0.021658", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 5.207, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 48.148, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "48.148", + "raw_stderr": "6.086", + "latency": "1580.348", + "cost_per_test": "0.068395", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 6.086, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5/768ed001-6c8b-4b96-ba28-8e0c35157b55.json b/data/vals_ai/xiaomi/mimo-v2.5/768ed001-6c8b-4b96-ba28-8e0c35157b55.json new file mode 100644 index 0000000000000000000000000000000000000000..dcfe4310ac76326564bb1ddb7d6533922a3c4233 --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5/768ed001-6c8b-4b96-ba28-8e0c35157b55.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/xiaomi_mimo-v2.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5", + "id": "xiaomi/mimo-v2.5", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 72.151, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "72.151", + "raw_stderr": "1.851", + "latency": "20.261", + "cost_per_test": "0.001351", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.851, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5/7f45a4a4-070c-47c0-b8d0-932739173c7e.json b/data/vals_ai/xiaomi/mimo-v2.5/7f45a4a4-070c-47c0-b8d0-932739173c7e.json new file mode 100644 index 0000000000000000000000000000000000000000..60d7c467fc585676fe357ae39d77fc2c78244dca --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5/7f45a4a4-070c-47c0-b8d0-932739173c7e.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/xiaomi_mimo-v2.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5", + "id": "xiaomi/mimo-v2.5", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 79.798, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.798", + "raw_stderr": "2.905", + "latency": "77.418", + "cost_per_test": "0.002294", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 2.905, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 83.333, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.333", + "raw_stderr": "2.905", + "latency": "101.368", + "cost_per_test": "0.002781", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 2.905, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 81.566, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.566", + "raw_stderr": "2.054", + "latency": "89.393", + "cost_per_test": "0.002538", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 2.054, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5/9d54b17b-9d40-4c08-a218-063c420dacc3.json b/data/vals_ai/xiaomi/mimo-v2.5/9d54b17b-9d40-4c08-a218-063c420dacc3.json new file mode 100644 index 0000000000000000000000000000000000000000..bc4e99900469a78c83dc3351c4f75bb4095e2faa --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5/9d54b17b-9d40-4c08-a218-063c420dacc3.json @@ -0,0 +1,438 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/xiaomi_mimo-v2.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5", + "id": "xiaomi/mimo-v2.5", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 86.386, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.386", + "raw_stderr": "0.844", + "latency": "8.689", + "cost_per_test": "0.000189", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.844, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 74.351, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "74.351", + "raw_stderr": "0.235", + "latency": "12.892", + "cost_per_test": "0.000317", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.235, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 86.337, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.337", + "raw_stderr": "0.626", + "latency": "12.928", + "cost_per_test": "0.000331", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.626, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.886, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.886", + "raw_stderr": "0.506", + "latency": "11.184", + "cost_per_test": "0.000264", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.506, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.776, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.776", + "raw_stderr": "1.043", + "latency": "10.166", + "cost_per_test": "0.000254", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.043, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 69.582, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "69.582", + "raw_stderr": "2.038", + "latency": "11.246", + "cost_per_test": "0.000227", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 2.038, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5/b1c3eb73-e752-455e-aa0e-3a06ffe44327.json b/data/vals_ai/xiaomi/mimo-v2.5/b1c3eb73-e752-455e-aa0e-3a06ffe44327.json new file mode 100644 index 0000000000000000000000000000000000000000..1145bcd3c21fdeb4b236570223ecb5feca7306ef --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5/b1c3eb73-e752-455e-aa0e-3a06ffe44327.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mortgage_tax/xiaomi_mimo-v2.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MortgageTax", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5", + "id": "xiaomi/mimo-v2.5", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mortgage_tax:numerical_extraction:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.mortgage_tax.numerical_extraction", + "source_data": { + "dataset_name": "MortgageTax - Numerical Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "numerical_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Numerical Extraction).", + "metric_id": "vals_ai.mortgage_tax.numerical_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 61.049, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "numerical_extraction", + "task_name": "Numerical Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.049", + "raw_stderr": "1.375", + "latency": "38.218", + "cost_per_test": "0.001623", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.375, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:overall:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.mortgage_tax.overall", + "source_data": { + "dataset_name": "MortgageTax - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Overall).", + "metric_id": "vals_ai.mortgage_tax.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 59.26, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.26", + "raw_stderr": "0.979", + "latency": "32.939", + "cost_per_test": "0.001523", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.979, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mortgage_tax:semantic_extraction:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.mortgage_tax.semantic_extraction", + "source_data": { + "dataset_name": "MortgageTax - Semantic Extraction", + "source_type": "other", + "additional_details": { + "benchmark_slug": "mortgage_tax", + "task_key": "semantic_extraction", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MortgageTax (Semantic Extraction).", + "metric_id": "vals_ai.mortgage_tax.semantic_extraction.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mortgage_tax" + } + }, + "score_details": { + "score": 57.472, + "details": { + "benchmark_slug": "mortgage_tax", + "benchmark_name": "MortgageTax", + "benchmark_updated": "2026-07-09", + "task_key": "semantic_extraction", + "task_name": "Semantic Extraction", + "dataset_type": "private", + "industry": "finance", + "raw_score": "57.472", + "raw_stderr": "1.394", + "latency": "27.66", + "cost_per_test": "0.001423", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.394, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5/c740744a-d3b5-496b-9ac5-66f373127ba9.json b/data/vals_ai/xiaomi/mimo-v2.5/c740744a-d3b5-496b-9ac5-66f373127ba9.json new file mode 100644 index 0000000000000000000000000000000000000000..30a2be1ce3d254a7d9fd89a68fd6c2cb1fc020f9 --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5/c740744a-d3b5-496b-9ac5-66f373127ba9.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/xiaomi_mimo-v2.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5", + "id": "xiaomi/mimo-v2.5", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 60.752, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.752", + "raw_stderr": "1.396", + "latency": "26.3", + "cost_per_test": "0.000729", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.396, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 71.832, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.832", + "raw_stderr": "0.881", + "latency": "26.3", + "cost_per_test": "0.000729", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.881, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 82.911, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "82.911", + "raw_stderr": "1.076", + "latency": "26.3", + "cost_per_test": "0.000729", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.076, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5/d82c505b-11b7-4439-831d-91c06cfe15e8.json b/data/vals_ai/xiaomi/mimo-v2.5/d82c505b-11b7-4439-831d-91c06cfe15e8.json new file mode 100644 index 0000000000000000000000000000000000000000..242c2f4eb37caf4b4cb20e968db350187c8a71e0 --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5/d82c505b-11b7-4439-831d-91c06cfe15e8.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/code-migration/xiaomi_mimo-v2.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Code Migration", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5", + "id": "xiaomi/mimo-v2.5", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "code-migration:cli:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.code-migration.cli", + "source_data": { + "dataset_name": "Code Migration - CLI", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cli", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (CLI).", + "metric_id": "vals_ai.code-migration.cli.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 9.88, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cli", + "task_name": "CLI", + "dataset_type": "private", + "industry": "coding", + "raw_score": "9.88", + "raw_stderr": "1.46", + "latency": "2643.447", + "cost_per_test": "0.11675", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.46, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "code-migration:cobol:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.code-migration.cobol", + "source_data": { + "dataset_name": "Code Migration - COBOL", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cobol", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (COBOL).", + "metric_id": "vals_ai.code-migration.cobol.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 27.273, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cobol", + "task_name": "COBOL", + "dataset_type": "private", + "industry": "coding", + "raw_score": "27.273", + "raw_stderr": "14.084", + "latency": "1168.186", + "cost_per_test": "0.060959", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 14.084, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "code-migration:code_quality:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.code-migration.code_quality", + "source_data": { + "dataset_name": "Code Migration - Code Quality", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "code_quality", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Code Quality).", + "metric_id": "vals_ai.code-migration.code_quality.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 2.788, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "code_quality", + "task_name": "Code Quality", + "dataset_type": "private", + "industry": "coding", + "raw_score": "2.788", + "raw_stderr": "0.077", + "latency": "2274.632", + "cost_per_test": "0.102802", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.077, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "code-migration:overall:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.code-migration.overall", + "source_data": { + "dataset_name": "Code Migration - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Overall).", + "metric_id": "vals_ai.code-migration.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 14.228, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "14.228", + "raw_stderr": "3.687", + "latency": "2274.632", + "cost_per_test": "0.102802", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 3.687, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5/e15d6223-3907-48dc-9b30-06758ae329c5.json b/data/vals_ai/xiaomi/mimo-v2.5/e15d6223-3907-48dc-9b30-06758ae329c5.json new file mode 100644 index 0000000000000000000000000000000000000000..53c6d01f03c42d137a15caa9b432f24acd60a059 --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5/e15d6223-3907-48dc-9b30-06758ae329c5.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/xiaomi_mimo-v2.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5", + "id": "xiaomi/mimo-v2.5", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 100.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "100", + "raw_stderr": "0", + "latency": "17249.061", + "cost_per_test": "0.481881", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 36.667, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "36.667", + "raw_stderr": "8.949", + "latency": "2299.875", + "cost_per_test": "0.064251", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 8.949, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 70.909, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "70.909", + "raw_stderr": "6.181", + "latency": "1254.477", + "cost_per_test": "0.035046", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 6.181, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 60.674, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "60.674", + "raw_stderr": "5.207", + "latency": "775.239", + "cost_per_test": "0.021658", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 5.207, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5/eefb3ff3-8f56-4c9c-9340-e68e5994d8f4.json b/data/vals_ai/xiaomi/mimo-v2.5/eefb3ff3-8f56-4c9c-9340-e68e5994d8f4.json new file mode 100644 index 0000000000000000000000000000000000000000..d092e7a4c5da0c9ac6481b6f4ddf93db790ea89a --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5/eefb3ff3-8f56-4c9c-9340-e68e5994d8f4.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/xiaomi_mimo-v2.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5", + "id": "xiaomi/mimo-v2.5", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 42.174, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "42.174", + "raw_stderr": "4.568", + "latency": "1580.348", + "cost_per_test": "0.068395", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 4.568, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/xiaomi/mimo-v2.5/f0983d4f-57ef-4339-bbf3-51bae1081275.json b/data/vals_ai/xiaomi/mimo-v2.5/f0983d4f-57ef-4339-bbf3-51bae1081275.json new file mode 100644 index 0000000000000000000000000000000000000000..880940ba1da7469e45e5fbe601e6b8618a41c54a --- /dev/null +++ b/data/vals_ai/xiaomi/mimo-v2.5/f0983d4f-57ef-4339-bbf3-51bae1081275.json @@ -0,0 +1,740 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/xiaomi_mimo-v2.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "mimo-v2.5", + "id": "xiaomi/mimo-v2.5", + "developer": "xiaomi", + "additional_details": { + "vals_model_id": "xiaomi/mimo-v2.5", + "vals_provider": "Xiaomi" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 31.746, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "31.746", + "raw_stderr": "2.451", + "latency": "328.95", + "cost_per_test": "0.088997", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 2.451, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 23.785, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "23.785", + "raw_stderr": "0.783", + "latency": "339.26", + "cost_per_test": "0.08693", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.783, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 31.672, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "31.672", + "raw_stderr": "0.096", + "latency": "417.311", + "cost_per_test": "0.114992", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.096, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 47.346, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "47.346", + "raw_stderr": "4.019", + "latency": "331.617", + "cost_per_test": "0.109825", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 4.019, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 45.006, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "45.006", + "raw_stderr": "1.467", + "latency": "269.15", + "cost_per_test": "0.030265", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.467, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 5.271, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "5.271", + "raw_stderr": "0.598", + "latency": "594.806", + "cost_per_test": "0.091582", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.598, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 47.992, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "47.992", + "raw_stderr": "2.089", + "latency": "278.393", + "cost_per_test": "0.097278", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 2.089, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 55.744, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "55.744", + "raw_stderr": "2.949", + "latency": "304.449", + "cost_per_test": "0.098023", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 2.949, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 48.567, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "48.567", + "raw_stderr": "1.659", + "latency": "246.009", + "cost_per_test": "0.03703", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.659, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:overall:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 36.728, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "36.728", + "raw_stderr": "0.109", + "latency": "339.26", + "cost_per_test": "0.08693", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 0.109, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:xiaomi/mimo-v2.5:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 17.206, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "17.206", + "raw_stderr": "1.229", + "latency": "413.518", + "cost_per_test": "0.114392", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "128000", + "provider": "Xiaomi" + }, + "uncertainty": { + "standard_error": { + "value": 1.229, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-4.5/2c932bba-2163-4a43-a1ba-4549f0255494.json b/data/vals_ai/zai/glm-4.5/2c932bba-2163-4a43-a1ba-4549f0255494.json new file mode 100644 index 0000000000000000000000000000000000000000..1aad003ad2f1d8ac2217185146a8114a29831606 --- /dev/null +++ b/data/vals_ai/zai/glm-4.5/2c932bba-2163-4a43-a1ba-4549f0255494.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/zai_glm-4.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-4.5", + "id": "zai/glm-4.5", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-4.5", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:zai/glm-4.5:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 64.336, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.336", + "raw_stderr": "1.635", + "latency": "57.11", + "cost_per_test": "0.002725", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.635, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:zai/glm-4.5:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 55.478, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "55.478", + "raw_stderr": "1.697", + "latency": "136.325", + "cost_per_test": "0.069957", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.697, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:zai/glm-4.5:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.956, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.956", + "raw_stderr": "0.959", + "latency": "95.244", + "cost_per_test": "0.03783", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.959, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:zai/glm-4.5:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 63.054, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.054", + "raw_stderr": "1.648", + "latency": "92.297", + "cost_per_test": "0.040809", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.648, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-4.5/3ceeae14-ac58-43e0-81f4-df1fce0ec9f1.json b/data/vals_ai/zai/glm-4.5/3ceeae14-ac58-43e0-81f4-df1fce0ec9f1.json new file mode 100644 index 0000000000000000000000000000000000000000..9e79b698de53508c49181ce22715f77d37e527ec --- /dev/null +++ b/data/vals_ai/zai/glm-4.5/3ceeae14-ac58-43e0-81f4-df1fce0ec9f1.json @@ -0,0 +1,432 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/zai_glm-4.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-4.5", + "id": "zai/glm-4.5", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-4.5", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:zai/glm-4.5:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 88.598, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.598", + "raw_stderr": "0.83", + "latency": "11.244", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.83, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:zai/glm-4.5:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 74.17, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "74.17", + "raw_stderr": "0.242", + "latency": "17.594", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.242, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:zai/glm-4.5:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 64.714, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "64.714", + "raw_stderr": "0.901", + "latency": "4.049", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.901, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:zai/glm-4.5:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 75.627, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "75.627", + "raw_stderr": "0.511", + "latency": "16.402", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.511, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:zai/glm-4.5:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.641, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.641", + "raw_stderr": "1.311", + "latency": "6.622", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.311, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:zai/glm-4.5:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 73.01, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "73.01", + "raw_stderr": "1.806", + "latency": "42.5", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.806, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-4.5/67cd1a5f-5865-44f2-a150-0bff350c1205.json b/data/vals_ai/zai/glm-4.5/67cd1a5f-5865-44f2-a150-0bff350c1205.json new file mode 100644 index 0000000000000000000000000000000000000000..5d23be66ee77e3dadfc41d77f2d3c823b17c67b5 --- /dev/null +++ b/data/vals_ai/zai/glm-4.5/67cd1a5f-5865-44f2-a150-0bff350c1205.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/zai_glm-4.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-4.5", + "id": "zai/glm-4.5", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-4.5", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:zai/glm-4.5:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 70.707, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "70.707", + "raw_stderr": "3.25", + "latency": "694.557", + "cost_per_test": "0.020131", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.25, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:zai/glm-4.5:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 73.737, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.737", + "raw_stderr": "3.146", + "latency": "713.892", + "cost_per_test": "0.018777", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.146, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:zai/glm-4.5:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 72.222, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.222", + "raw_stderr": "2.262", + "latency": "704.224", + "cost_per_test": "0.019454", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.262, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-4.5/7bcf8663-e6cc-4ae8-a8f9-48dac0c86014.json b/data/vals_ai/zai/glm-4.5/7bcf8663-e6cc-4ae8-a8f9-48dac0c86014.json new file mode 100644 index 0000000000000000000000000000000000000000..71ee80f00934f1604b2d29c94ed49f644885afcd --- /dev/null +++ b/data/vals_ai/zai/glm-4.5/7bcf8663-e6cc-4ae8-a8f9-48dac0c86014.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/zai_glm-4.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-4.5", + "id": "zai/glm-4.5", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-4.5", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:zai/glm-4.5:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 3.167, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "3.167", + "raw_stderr": "1.535", + "latency": "2172.987", + "cost_per_test": "1.093628", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.535, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:zai/glm-4.5:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 2.667, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "2.667", + "raw_stderr": "1.24", + "latency": "2499.873", + "cost_per_test": "0.790922", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.24, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:zai/glm-4.5:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 2.917, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "2.917", + "raw_stderr": "0.987", + "latency": "2336.43", + "cost_per_test": "0.942275", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.987, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-4.5/9b80cf6b-473a-452b-9c30-bc96311b9116.json b/data/vals_ai/zai/glm-4.5/9b80cf6b-473a-452b-9c30-bc96311b9116.json new file mode 100644 index 0000000000000000000000000000000000000000..1bdc9906195035be8282e63a06c8a0165c609908 --- /dev/null +++ b/data/vals_ai/zai/glm-4.5/9b80cf6b-473a-452b-9c30-bc96311b9116.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/zai_glm-4.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-4.5", + "id": "zai/glm-4.5", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-4.5", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:zai/glm-4.5:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 60.752, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.752", + "raw_stderr": "1.396", + "latency": "129.388", + "cost_per_test": "0.009559", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.396, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:zai/glm-4.5:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 72.404, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "72.404", + "raw_stderr": "0.872", + "latency": "129.388", + "cost_per_test": "0.009559", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.872, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:zai/glm-4.5:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 84.056, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "84.056", + "raw_stderr": "1.047", + "latency": "129.388", + "cost_per_test": "0.009559", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.047, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-4.5/ef0a9f9a-4379-4c4b-8dc1-4ff3cd04aa83.json b/data/vals_ai/zai/glm-4.5/ef0a9f9a-4379-4c4b-8dc1-4ff3cd04aa83.json new file mode 100644 index 0000000000000000000000000000000000000000..25bf336fbe21d696335d7d88e551789e3ced6cf4 --- /dev/null +++ b/data/vals_ai/zai/glm-4.5/ef0a9f9a-4379-4c4b-8dc1-4ff3cd04aa83.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/zai_glm-4.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-4.5", + "id": "zai/glm-4.5", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-4.5", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:zai/glm-4.5:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.447, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.447", + "raw_stderr": "0.689", + "latency": "134.586", + "cost_per_test": "0.006886", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.689, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:zai/glm-4.5:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 30.0, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "30", + "raw_stderr": "2.449", + "latency": "551.224", + "cost_per_test": "0.029016", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.449, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:zai/glm-4.5:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 73.89, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "73.89", + "raw_stderr": "2.244", + "latency": "487.993", + "cost_per_test": "0.018933", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.244, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:zai/glm-4.5:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 67.446, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "67.446", + "raw_stderr": "1.131", + "latency": "391.268", + "cost_per_test": "0.018278", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.131, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-4.5/fe55815e-1e95-461e-9ef9-bf8027376255.json b/data/vals_ai/zai/glm-4.5/fe55815e-1e95-461e-9ef9-bf8027376255.json new file mode 100644 index 0000000000000000000000000000000000000000..d437f511e7be82b9aa6d055bac4ac0202584954c --- /dev/null +++ b/data/vals_ai/zai/glm-4.5/fe55815e-1e95-461e-9ef9-bf8027376255.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/zai_glm-4.5/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-4.5", + "id": "zai/glm-4.5", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-4.5", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:zai/glm-4.5:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.656, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.656", + "raw_stderr": "1.087", + "latency": "80.35", + "cost_per_test": "0.007079", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.087, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:zai/glm-4.5:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.34, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.34", + "raw_stderr": "1.143", + "latency": "219.493", + "cost_per_test": "0.010354", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.143, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:zai/glm-4.5:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.131, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.131", + "raw_stderr": "1.027", + "latency": "350.263", + "cost_per_test": "0.016233", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.027, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:zai/glm-4.5:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.098, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.098", + "raw_stderr": "1.709", + "latency": "162.109", + "cost_per_test": "0.009198", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.709, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:zai/glm-4.5:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.967, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.967", + "raw_stderr": "1.159", + "latency": "76.691", + "cost_per_test": "0.006371", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.159, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:zai/glm-4.5:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 67.389, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "67.389", + "raw_stderr": "1.506", + "latency": "379.722", + "cost_per_test": "0.021658", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.506, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:zai/glm-4.5:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.117, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.117", + "raw_stderr": "1.446", + "latency": "57.607", + "cost_per_test": "0.006468", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.446, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:zai/glm-4.5:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 71.129, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "71.129", + "raw_stderr": "2.322", + "latency": "64.589", + "cost_per_test": "0.00669", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.322, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:zai/glm-4.5:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 65.123, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "65.123", + "raw_stderr": "1.436", + "latency": "98.1", + "cost_per_test": "0.008511", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.436, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:zai/glm-4.5:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.861, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.861", + "raw_stderr": "0.889", + "latency": "115.062", + "cost_per_test": "0.012745", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.889, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:zai/glm-4.5:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.411, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.411", + "raw_stderr": "1.306", + "latency": "69.032", + "cost_per_test": "0.005996", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.306, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:zai/glm-4.5:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.222, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.222", + "raw_stderr": "0.378", + "latency": "136.72", + "cost_per_test": "0.009757", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.378, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:zai/glm-4.5:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.16, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.16", + "raw_stderr": "1.785", + "latency": "74.109", + "cost_per_test": "0.005902", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.785, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:zai/glm-4.5:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.144, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.144", + "raw_stderr": "0.929", + "latency": "96.592", + "cost_per_test": "0.013783", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.929, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:zai/glm-4.5:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.579, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.579", + "raw_stderr": "1.372", + "latency": "70.368", + "cost_per_test": "0.005607", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "81920", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.372, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 81920, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-4.6/10505cf4-bf32-4fdc-bdaa-1578b8509746.json b/data/vals_ai/zai/glm-4.6/10505cf4-bf32-4fdc-bdaa-1578b8509746.json new file mode 100644 index 0000000000000000000000000000000000000000..61094fb7974ee7266648a0226683d1fc50649fa8 --- /dev/null +++ b/data/vals_ai/zai/glm-4.6/10505cf4-bf32-4fdc-bdaa-1578b8509746.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/zai_glm-4.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-4.6", + "id": "zai/glm-4.6", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-4.6", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:zai/glm-4.6:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 73.737, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "73.737", + "raw_stderr": "3.146", + "latency": "494.128", + "cost_per_test": "0.015371", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.146, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:zai/glm-4.6:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 75.253, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "75.253", + "raw_stderr": "3.067", + "latency": "455.539", + "cost_per_test": "0.015448", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.067, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:zai/glm-4.6:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 74.495, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74.495", + "raw_stderr": "2.197", + "latency": "474.834", + "cost_per_test": "0.01541", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.197, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-4.6/178629ad-436a-41c1-8697-70d3409ac7a6.json b/data/vals_ai/zai/glm-4.6/178629ad-436a-41c1-8697-70d3409ac7a6.json new file mode 100644 index 0000000000000000000000000000000000000000..0e54a5eb0e8edb83a8a9a3e9fef2d95dae4631ff --- /dev/null +++ b/data/vals_ai/zai/glm-4.6/178629ad-436a-41c1-8697-70d3409ac7a6.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/zai_glm-4.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-4.6", + "id": "zai/glm-4.6", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-4.6", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:zai/glm-4.6:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 3.09, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "3.09", + "raw_stderr": "1.162", + "latency": "10002.446", + "cost_per_test": "10.845475", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.162, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-4.6/19d4f8cd-c053-4f90-bf65-a38ffd8ce6cd.json b/data/vals_ai/zai/glm-4.6/19d4f8cd-c053-4f90-bf65-a38ffd8ce6cd.json new file mode 100644 index 0000000000000000000000000000000000000000..15289f7b320d280eb18bbd13cfa9f343efe1485b --- /dev/null +++ b/data/vals_ai/zai/glm-4.6/19d4f8cd-c053-4f90-bf65-a38ffd8ce6cd.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/zai_glm-4.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-4.6", + "id": "zai/glm-4.6", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-4.6", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:zai/glm-4.6:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 58.159, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "58.159", + "raw_stderr": "1.684", + "latency": "15.911", + "cost_per_test": "0.003792", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.684, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:zai/glm-4.6:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 49.184, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "49.184", + "raw_stderr": "1.707", + "latency": "155.354", + "cost_per_test": "0.089158", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.707, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:zai/glm-4.6:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 56.838, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.838", + "raw_stderr": "0.97", + "latency": "78.061", + "cost_per_test": "0.04461", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.97, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:zai/glm-4.6:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 63.17, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.17", + "raw_stderr": "1.647", + "latency": "62.917", + "cost_per_test": "0.040881", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.647, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-4.6/51c0f728-e881-4ba8-bbfc-bd36dcd67d11.json b/data/vals_ai/zai/glm-4.6/51c0f728-e881-4ba8-bbfc-bd36dcd67d11.json new file mode 100644 index 0000000000000000000000000000000000000000..7c2d30625682f8eeca68303b6c853bc42688f3d9 --- /dev/null +++ b/data/vals_ai/zai/glm-4.6/51c0f728-e881-4ba8-bbfc-bd36dcd67d11.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/zai_glm-4.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-4.6", + "id": "zai/glm-4.6", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-4.6", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:zai/glm-4.6:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.137, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.137", + "raw_stderr": "0.754", + "latency": "25.863", + "cost_per_test": "0.003968", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.754, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:zai/glm-4.6:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 58.286, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "58.286", + "raw_stderr": "2.636", + "latency": "546.351", + "cost_per_test": "0.055417", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.636, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:zai/glm-4.6:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 86.684, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "86.684", + "raw_stderr": "1.736", + "latency": "136.313", + "cost_per_test": "0.021343", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.736, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:zai/glm-4.6:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 81.036, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "81.036", + "raw_stderr": "1.082", + "latency": "236.176", + "cost_per_test": "0.026909", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.082, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-4.6/715aad0f-7e0d-4342-856b-8d8d10b06c67.json b/data/vals_ai/zai/glm-4.6/715aad0f-7e0d-4342-856b-8d8d10b06c67.json new file mode 100644 index 0000000000000000000000000000000000000000..ce85a91eb04e94c282d8e5e3aac1d29eb2851ec1 --- /dev/null +++ b/data/vals_ai/zai/glm-4.6/715aad0f-7e0d-4342-856b-8d8d10b06c67.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/zai_glm-4.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-4.6", + "id": "zai/glm-4.6", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-4.6", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:zai/glm-4.6:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 51.329, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "51.329", + "raw_stderr": "1.272", + "latency": "64.149", + "cost_per_test": "0.00928", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.272, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:zai/glm-4.6:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 66.235, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.235", + "raw_stderr": "0.808", + "latency": "64.149", + "cost_per_test": "0.00928", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.808, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:zai/glm-4.6:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 81.141, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "81.141", + "raw_stderr": "0.996", + "latency": "64.149", + "cost_per_test": "0.00928", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.996, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-4.6/a006692e-41a7-455f-b847-cad060be8d23.json b/data/vals_ai/zai/glm-4.6/a006692e-41a7-455f-b847-cad060be8d23.json new file mode 100644 index 0000000000000000000000000000000000000000..85c7423d8e1f044d08fc1aca469b395f9304cdcf --- /dev/null +++ b/data/vals_ai/zai/glm-4.6/a006692e-41a7-455f-b847-cad060be8d23.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/zai_glm-4.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-4.6", + "id": "zai/glm-4.6", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-4.6", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:zai/glm-4.6:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 1.0, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "1", + "raw_stderr": "0.577", + "latency": "6110.323", + "cost_per_test": "2.024354", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.577, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:zai/glm-4.6:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 7.667, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "7.667", + "raw_stderr": "4.937", + "latency": "8101.384", + "cost_per_test": "1.00872", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 4.937, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:zai/glm-4.6:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 4.334, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "4.334", + "raw_stderr": "2.485", + "latency": "7105.854", + "cost_per_test": "1.516537", + "temperature": "0.6", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.485, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.6, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-4.6/cb577f20-966d-49fe-a028-57b8e59d352f.json b/data/vals_ai/zai/glm-4.6/cb577f20-966d-49fe-a028-57b8e59d352f.json new file mode 100644 index 0000000000000000000000000000000000000000..2a045656c138386734b8100701768fcaec9c8701 --- /dev/null +++ b/data/vals_ai/zai/glm-4.6/cb577f20-966d-49fe-a028-57b8e59d352f.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/zai_glm-4.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-4.6", + "id": "zai/glm-4.6", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-4.6", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:zai/glm-4.6:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.05, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.05", + "raw_stderr": "1.01", + "latency": "29.233", + "cost_per_test": "0.004832", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.01, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:zai/glm-4.6:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.466, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.466", + "raw_stderr": "1.143", + "latency": "49.227", + "cost_per_test": "0.006347", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.143, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:zai/glm-4.6:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.134, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.134", + "raw_stderr": "0.928", + "latency": "53.391", + "cost_per_test": "0.007878", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.928, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:zai/glm-4.6:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.341, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.341", + "raw_stderr": "1.709", + "latency": "50.687", + "cost_per_test": "0.006772", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.709, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:zai/glm-4.6:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.493, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.493", + "raw_stderr": "1.177", + "latency": "36.579", + "cost_per_test": "0.005003", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.177, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:zai/glm-4.6:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.193, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.193", + "raw_stderr": "1.354", + "latency": "105.568", + "cost_per_test": "0.015562", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.354, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:zai/glm-4.6:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.462, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.462", + "raw_stderr": "1.412", + "latency": "32.202", + "cost_per_test": "0.004877", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.412, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:zai/glm-4.6:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 67.979, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "67.979", + "raw_stderr": "2.39", + "latency": "36.461", + "cost_per_test": "0.005921", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.39, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:zai/glm-4.6:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 60.763, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "60.763", + "raw_stderr": "1.472", + "latency": "47.136", + "cost_per_test": "0.00724", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.472, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:zai/glm-4.6:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.634, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.634", + "raw_stderr": "0.699", + "latency": "51.314", + "cost_per_test": "0.00669", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.699, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:zai/glm-4.6:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 79.545, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "79.545", + "raw_stderr": "1.327", + "latency": "34.003", + "cost_per_test": "0.004223", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.327, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:zai/glm-4.6:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.203, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.203", + "raw_stderr": "0.375", + "latency": "46.996", + "cost_per_test": "0.006572", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.375, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:zai/glm-4.6:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.353, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.353", + "raw_stderr": "1.902", + "latency": "35.981", + "cost_per_test": "0.005123", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.902, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:zai/glm-4.6:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.223, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.223", + "raw_stderr": "0.833", + "latency": "56.515", + "cost_per_test": "0.007233", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.833, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:zai/glm-4.6:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.208, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.208", + "raw_stderr": "1.327", + "latency": "39.645", + "cost_per_test": "0.004307", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.327, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-4.6/e689e815-ff87-4bca-a10c-46780b6a632f.json b/data/vals_ai/zai/glm-4.6/e689e815-ff87-4bca-a10c-46780b6a632f.json new file mode 100644 index 0000000000000000000000000000000000000000..23ae5ce7694f77f766f29a7408dd722978d87d71 --- /dev/null +++ b/data/vals_ai/zai/glm-4.6/e689e815-ff87-4bca-a10c-46780b6a632f.json @@ -0,0 +1,432 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/zai_glm-4.6/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-4.6", + "id": "zai/glm-4.6", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-4.6", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:zai/glm-4.6:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 85.616, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.616", + "raw_stderr": "0.861", + "latency": "32.329", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.861, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:zai/glm-4.6:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.915, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.915", + "raw_stderr": "0.212", + "latency": "29.207", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.212, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:zai/glm-4.6:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 84.529, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.529", + "raw_stderr": "0.681", + "latency": "16.281", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.681, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:zai/glm-4.6:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.608, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.608", + "raw_stderr": "0.522", + "latency": "29.819", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.522, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:zai/glm-4.6:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 77.244, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "77.244", + "raw_stderr": "1.334", + "latency": "31.96", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.334, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:zai/glm-4.6:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 71.734, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "71.734", + "raw_stderr": "1.943", + "latency": "39.317", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "122880", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.943, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 122880, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-4.7/0f09da7b-c3eb-4509-a9b5-8fa8ab39464f.json b/data/vals_ai/zai/glm-4.7/0f09da7b-c3eb-4509-a9b5-8fa8ab39464f.json new file mode 100644 index 0000000000000000000000000000000000000000..ef8369e2a8fb6040229cea9f428ea639cc454dcf --- /dev/null +++ b/data/vals_ai/zai/glm-4.7/0f09da7b-c3eb-4509-a9b5-8fa8ab39464f.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/zai_glm-4.7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-4.7", + "id": "zai/glm-4.7", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-4.7", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:zai/glm-4.7:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.137, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.137", + "raw_stderr": "0.754", + "latency": "39.111", + "cost_per_test": "0.004361", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.754, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:zai/glm-4.7:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 61.098, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "61.098", + "raw_stderr": "2.332", + "latency": "884.647", + "cost_per_test": "0.073162", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.332, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:zai/glm-4.7:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 87.467, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "87.467", + "raw_stderr": "1.692", + "latency": "256.071", + "cost_per_test": "0.028885", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.692, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:zai/glm-4.7:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 82.234, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "82.234", + "raw_stderr": "0.993", + "latency": "393.276", + "cost_per_test": "0.035469", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.993, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-4.7/10bbaa16-f6dc-4029-8cee-6a5ee3490580.json b/data/vals_ai/zai/glm-4.7/10bbaa16-f6dc-4029-8cee-6a5ee3490580.json new file mode 100644 index 0000000000000000000000000000000000000000..15084c9ea6f03fbc76c1cdcd0cde06441023c5fc --- /dev/null +++ b/data/vals_ai/zai/glm-4.7/10bbaa16-f6dc-4029-8cee-6a5ee3490580.json @@ -0,0 +1,371 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/zai_glm-4.7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-4.7", + "id": "zai/glm-4.7", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-4.7", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:zai/glm-4.7:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 35.714, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "35.714", + "raw_stderr": "7.483", + "latency": "470.132", + "cost_per_test": "0.312315", + "temperature": "0.7", + "top_p": "1", + "max_output_tokens": "16384", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 7.483, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:zai/glm-4.7:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 64.368, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "64.368", + "raw_stderr": "2.97", + "latency": "401.044", + "cost_per_test": "0.226941", + "temperature": "0.7", + "top_p": "1", + "max_output_tokens": "16384", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.97, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:zai/glm-4.7:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 84.021, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "84.021", + "raw_stderr": "2.638", + "latency": "290.707", + "cost_per_test": "0.147064", + "temperature": "0.7", + "top_p": "1", + "max_output_tokens": "16384", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.638, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:zai/glm-4.7:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "694.784", + "cost_per_test": "0.322823", + "temperature": "0.7", + "top_p": "1", + "max_output_tokens": "16384", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:zai/glm-4.7:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 69.4, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "69.4", + "raw_stderr": "2.063", + "latency": "365.799", + "cost_per_test": "0.203695", + "temperature": "0.7", + "top_p": "1", + "max_output_tokens": "16384", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.063, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 0.7, + "top_p": 1.0, + "max_tokens": 16384, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-4.7/465e6075-bd18-4405-b9fb-e33670f2efa7.json b/data/vals_ai/zai/glm-4.7/465e6075-bd18-4405-b9fb-e33670f2efa7.json new file mode 100644 index 0000000000000000000000000000000000000000..a6152f80e75ea5d86921991983cb4d5d82ca2171 --- /dev/null +++ b/data/vals_ai/zai/glm-4.7/465e6075-bd18-4405-b9fb-e33670f2efa7.json @@ -0,0 +1,438 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/zai_glm-4.7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-4.7", + "id": "zai/glm-4.7", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-4.7", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:zai/glm-4.7:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 90.08, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "90.08", + "raw_stderr": "0.991", + "latency": "28.536", + "cost_per_test": "0.002008", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.991, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:zai/glm-4.7:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.084, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.084", + "raw_stderr": "0.271", + "latency": "28.613", + "cost_per_test": "0.002465", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.271, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:zai/glm-4.7:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 88.746, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.746", + "raw_stderr": "0.595", + "latency": "26.198", + "cost_per_test": "0.002019", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.595, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:zai/glm-4.7:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 83.358, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.358", + "raw_stderr": "0.473", + "latency": "30.675", + "cost_per_test": "0.002348", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.473, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:zai/glm-4.7:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.7, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.7", + "raw_stderr": "1.228", + "latency": "23.862", + "cost_per_test": "0.002035", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.228, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:zai/glm-4.7:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 79.181, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.181", + "raw_stderr": "1.635", + "latency": "46.168", + "cost_per_test": "0.003212", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.635, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-4.7/4682d219-34fe-46cf-bf4a-a5cda1510f89.json b/data/vals_ai/zai/glm-4.7/4682d219-34fe-46cf-bf4a-a5cda1510f89.json new file mode 100644 index 0000000000000000000000000000000000000000..2cfe318f1020857b9b79afce532d6826216f0f34 --- /dev/null +++ b/data/vals_ai/zai/glm-4.7/4682d219-34fe-46cf-bf4a-a5cda1510f89.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/zai_glm-4.7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-4.7", + "id": "zai/glm-4.7", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-4.7", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:zai/glm-4.7:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 68.629, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "68.629", + "raw_stderr": "2.123", + "latency": "167.479", + "cost_per_test": "0.019082", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "30000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.123, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-4.7/7234a0c8-4353-4195-ab6c-9b112c1a1974.json b/data/vals_ai/zai/glm-4.7/7234a0c8-4353-4195-ab6c-9b112c1a1974.json new file mode 100644 index 0000000000000000000000000000000000000000..a1dc41cbc78c053a4e03f237a8374ba6a72b1d2d --- /dev/null +++ b/data/vals_ai/zai/glm-4.7/7234a0c8-4353-4195-ab6c-9b112c1a1974.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/zai_glm-4.7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-4.7", + "id": "zai/glm-4.7", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-4.7", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:zai/glm-4.7:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.023, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.023", + "raw_stderr": "1.672", + "latency": "39.216", + "cost_per_test": "0.00438", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.672, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:zai/glm-4.7:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 17.949, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "17.949", + "raw_stderr": "1.31", + "latency": "93.697", + "cost_per_test": "0.067835", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.31, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:zai/glm-4.7:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 46.387, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "46.387", + "raw_stderr": "0.899", + "latency": "66.101", + "cost_per_test": "0.039082", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.899, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:zai/glm-4.7:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.189, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.189", + "raw_stderr": "1.664", + "latency": "65.391", + "cost_per_test": "0.045032", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.664, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-4.7/944b83dc-056f-40dd-94f5-d721baf86919.json b/data/vals_ai/zai/glm-4.7/944b83dc-056f-40dd-94f5-d721baf86919.json new file mode 100644 index 0000000000000000000000000000000000000000..6349bfaa23950c87f2700d073144a29b1ea4cd6b --- /dev/null +++ b/data/vals_ai/zai/glm-4.7/944b83dc-056f-40dd-94f5-d721baf86919.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/zai_glm-4.7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-4.7", + "id": "zai/glm-4.7", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-4.7", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:zai/glm-4.7:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 78.283, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.283", + "raw_stderr": "2.93", + "latency": "205.538", + "cost_per_test": "0.024294", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.93, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:zai/glm-4.7:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 81.818, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.818", + "raw_stderr": "2.741", + "latency": "216.62", + "cost_per_test": "0.025437", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.741, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:zai/glm-4.7:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 80.05, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.05", + "raw_stderr": "2.006", + "latency": "211.079", + "cost_per_test": "0.024866", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.006, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-4.7/b6b2b666-a4f4-4689-8dcd-206639a61bd2.json b/data/vals_ai/zai/glm-4.7/b6b2b666-a4f4-4689-8dcd-206639a61bd2.json new file mode 100644 index 0000000000000000000000000000000000000000..51c52e76e553167a9d17f98e7a0ef577292e90ab --- /dev/null +++ b/data/vals_ai/zai/glm-4.7/b6b2b666-a4f4-4689-8dcd-206639a61bd2.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/zai_glm-4.7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-4.7", + "id": "zai/glm-4.7", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-4.7", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:zai/glm-4.7:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 10.5, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "10.5", + "raw_stderr": "5.882", + "latency": "3615.545", + "cost_per_test": "0.384617", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 5.882, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:zai/glm-4.7:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 4.667, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "4.667", + "raw_stderr": "3.429", + "latency": "7018.207", + "cost_per_test": "0.502413", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.429, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:zai/glm-4.7:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 7.584, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "7.584", + "raw_stderr": "3.404", + "latency": "5316.876", + "cost_per_test": "0.443515", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.404, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-4.7/c629d68d-be27-41df-92e8-dd58a49b7be2.json b/data/vals_ai/zai/glm-4.7/c629d68d-be27-41df-92e8-dd58a49b7be2.json new file mode 100644 index 0000000000000000000000000000000000000000..21008e35d44ccfa5927fb5b11d977c63b8fd0495 --- /dev/null +++ b/data/vals_ai/zai/glm-4.7/c629d68d-be27-41df-92e8-dd58a49b7be2.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/zai_glm-4.7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-4.7", + "id": "zai/glm-4.7", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-4.7", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:zai/glm-4.7:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 60.262, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.262", + "raw_stderr": "1.399", + "latency": "187.181", + "cost_per_test": "0.013873", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.399, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:zai/glm-4.7:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 68.766, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.766", + "raw_stderr": "0.921", + "latency": "187.181", + "cost_per_test": "0.013873", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.921, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:zai/glm-4.7:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 77.269, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "77.269", + "raw_stderr": "1.198", + "latency": "187.181", + "cost_per_test": "0.013873", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.198, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-4.7/cbf6909d-cc19-45d7-87a7-3f46614c206c.json b/data/vals_ai/zai/glm-4.7/cbf6909d-cc19-45d7-87a7-3f46614c206c.json new file mode 100644 index 0000000000000000000000000000000000000000..9c550e675c44e141db8b3f581265589b3aa6d3ad --- /dev/null +++ b/data/vals_ai/zai/glm-4.7/cbf6909d-cc19-45d7-87a7-3f46614c206c.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/zai_glm-4.7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-4.7", + "id": "zai/glm-4.7", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-4.7", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:zai/glm-4.7:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 6.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "6", + "raw_stderr": "2.375", + "latency": "3364.789", + "cost_per_test": "0.14404", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.375, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-4.7/fe8fbaeb-123e-4980-9eec-3d39fd9445ab.json b/data/vals_ai/zai/glm-4.7/fe8fbaeb-123e-4980-9eec-3d39fd9445ab.json new file mode 100644 index 0000000000000000000000000000000000000000..98c3250fb4baa57a16634516c2c398696f0a098d --- /dev/null +++ b/data/vals_ai/zai/glm-4.7/fe8fbaeb-123e-4980-9eec-3d39fd9445ab.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/zai_glm-4.7/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-4.7", + "id": "zai/glm-4.7", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-4.7", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:zai/glm-4.7:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.377, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.377", + "raw_stderr": "1.101", + "latency": "64.727", + "cost_per_test": "0.005043", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.101, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:zai/glm-4.7:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.199, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.199", + "raw_stderr": "1.189", + "latency": "86.161", + "cost_per_test": "0.0068", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.189, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:zai/glm-4.7:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.223, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.223", + "raw_stderr": "0.922", + "latency": "160.189", + "cost_per_test": "0.013335", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.922, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:zai/glm-4.7:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.585, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.585", + "raw_stderr": "1.696", + "latency": "103.442", + "cost_per_test": "0.007504", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.696, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:zai/glm-4.7:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.848, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.848", + "raw_stderr": "1.163", + "latency": "64.194", + "cost_per_test": "0.00463", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.163, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:zai/glm-4.7:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.638, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.638", + "raw_stderr": "1.317", + "latency": "291.489", + "cost_per_test": "0.020993", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.317, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:zai/glm-4.7:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.628, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.628", + "raw_stderr": "1.457", + "latency": "63.083", + "cost_per_test": "0.004571", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.457, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:zai/glm-4.7:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 72.703, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.703", + "raw_stderr": "2.282", + "latency": "61.031", + "cost_per_test": "0.005314", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.282, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:zai/glm-4.7:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 64.759, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "64.759", + "raw_stderr": "1.44", + "latency": "121.12", + "cost_per_test": "0.00951", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.44, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:zai/glm-4.7:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.968, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.968", + "raw_stderr": "0.696", + "latency": "96.951", + "cost_per_test": "0.008694", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.696, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:zai/glm-4.7:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.303, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.303", + "raw_stderr": "1.308", + "latency": "72.709", + "cost_per_test": "0.004031", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.308, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:zai/glm-4.7:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 82.736, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.736", + "raw_stderr": "0.371", + "latency": "104.451", + "cost_per_test": "0.00798", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.371, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:zai/glm-4.7:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.956, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.956", + "raw_stderr": "1.862", + "latency": "77.484", + "cost_per_test": "0.005365", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.862, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:zai/glm-4.7:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.53, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.53", + "raw_stderr": "0.852", + "latency": "135.407", + "cost_per_test": "0.012007", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.852, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:zai/glm-4.7:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.584, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.584", + "raw_stderr": "1.315", + "latency": "64.321", + "cost_per_test": "0.003926", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "128000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.315, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 128000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5-thinking/3b042de5-b7d8-4d58-aa39-96f79826c419.json b/data/vals_ai/zai/glm-5-thinking/3b042de5-b7d8-4d58-aa39-96f79826c419.json new file mode 100644 index 0000000000000000000000000000000000000000..58d81c226bfa087e890f6f813dd3bb7d1b55bf37 --- /dev/null +++ b/data/vals_ai/zai/glm-5-thinking/3b042de5-b7d8-4d58-aa39-96f79826c419.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/zai_glm-5-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5-thinking", + "id": "zai/glm-5-thinking", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5-thinking", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 60.14, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.14", + "raw_stderr": "1.672", + "latency": "31.137", + "cost_per_test": "0.006603", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.672, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.538, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.538", + "raw_stderr": "1.661", + "latency": "104.077", + "cost_per_test": "0.123859", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.661, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 62.898, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.898", + "raw_stderr": "0.95", + "latency": "68.577", + "cost_per_test": "0.067926", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.95, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 67.016, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "67.016", + "raw_stderr": "1.605", + "latency": "70.517", + "cost_per_test": "0.073316", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.605, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5-thinking/5ddc9dfb-3682-4cbf-98b2-9a895effdaaf.json b/data/vals_ai/zai/glm-5-thinking/5ddc9dfb-3682-4cbf-98b2-9a895effdaaf.json new file mode 100644 index 0000000000000000000000000000000000000000..82da3fbd699c3e4ae52b1ad27c6e414cc7040a22 --- /dev/null +++ b/data/vals_ai/zai/glm-5-thinking/5ddc9dfb-3682-4cbf-98b2-9a895effdaaf.json @@ -0,0 +1,438 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/zai_glm-5-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5-thinking", + "id": "zai/glm-5-thinking", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5-thinking", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 88.837, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.837", + "raw_stderr": "0.723", + "latency": "12.629", + "cost_per_test": "0.001675", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.723, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.827, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.827", + "raw_stderr": "0.229", + "latency": "19.112", + "cost_per_test": "0.002786", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.229, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 87.755, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.755", + "raw_stderr": "0.613", + "latency": "8.909", + "cost_per_test": "0.00167", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.613, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 84.059, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.059", + "raw_stderr": "0.411", + "latency": "16.056", + "cost_per_test": "0.002311", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.411, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.649, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.649", + "raw_stderr": "0.918", + "latency": "14.362", + "cost_per_test": "0.002135", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.918, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.228, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.228", + "raw_stderr": "1.558", + "latency": "25.269", + "cost_per_test": "0.00329", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.558, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5-thinking/6bba7a79-e3cc-47f0-bfeb-1800ef146eee.json b/data/vals_ai/zai/glm-5-thinking/6bba7a79-e3cc-47f0-bfeb-1800ef146eee.json new file mode 100644 index 0000000000000000000000000000000000000000..f8fd32ecc1e588f0bdae7a67463cb503cf65ac88 --- /dev/null +++ b/data/vals_ai/zai/glm-5-thinking/6bba7a79-e3cc-47f0-bfeb-1800ef146eee.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/zai_glm-5-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5-thinking", + "id": "zai/glm-5-thinking", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5-thinking", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 59.199, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "59.199", + "raw_stderr": "1.405", + "latency": "149.298", + "cost_per_test": "0.021913", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.405, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 70.033, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "70.033", + "raw_stderr": "0.9", + "latency": "149.298", + "cost_per_test": "0.021913", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.9, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 80.867, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "80.867", + "raw_stderr": "1.125", + "latency": "149.298", + "cost_per_test": "0.021913", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.125, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5-thinking/726ddc73-3e77-41d6-8937-a6f506295f68.json b/data/vals_ai/zai/glm-5-thinking/726ddc73-3e77-41d6-8937-a6f506295f68.json new file mode 100644 index 0000000000000000000000000000000000000000..a3d6feb7f0e26aa5a999d7e13fe38ce9138aa68c --- /dev/null +++ b/data/vals_ai/zai/glm-5-thinking/726ddc73-3e77-41d6-8937-a6f506295f68.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/zai_glm-5-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5-thinking", + "id": "zai/glm-5-thinking", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5-thinking", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 23.359, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "23.359", + "raw_stderr": "4.034", + "latency": "13455.794", + "cost_per_test": "40.270258", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 4.034, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5-thinking/77fa925b-3dad-43ef-92e6-df8c510ee5a2.json b/data/vals_ai/zai/glm-5-thinking/77fa925b-3dad-43ef-92e6-df8c510ee5a2.json new file mode 100644 index 0000000000000000000000000000000000000000..0168b1e79046dc0ab5195c58b2d5eeebfdca5113 --- /dev/null +++ b/data/vals_ai/zai/glm-5-thinking/77fa925b-3dad-43ef-92e6-df8c510ee5a2.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/ioi/zai_glm-5-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - IOI", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5-thinking", + "id": "zai/glm-5-thinking", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5-thinking", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "ioi:ioi2024:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.ioi.ioi2024", + "source_data": { + "dataset_name": "IOI - IOI 2024", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2024", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2024).", + "metric_id": "vals_ai.ioi.ioi2024.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 23.667, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2024", + "task_name": "IOI 2024", + "dataset_type": "public", + "industry": "coding", + "raw_score": "23.667", + "raw_stderr": "11.716", + "latency": "7231.432", + "cost_per_test": "2.864289", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 11.716, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:ioi2025:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.ioi.ioi2025", + "source_data": { + "dataset_name": "IOI - IOI 2025", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "ioi2025", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (IOI 2025).", + "metric_id": "vals_ai.ioi.ioi2025.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 20.333, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "ioi2025", + "task_name": "IOI 2025", + "dataset_type": "public", + "industry": "coding", + "raw_score": "20.333", + "raw_stderr": "10.397", + "latency": "6382.444", + "cost_per_test": "3.68306", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 10.397, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "ioi:overall:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.ioi.overall", + "source_data": { + "dataset_name": "IOI - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/ioi" + ], + "additional_details": { + "benchmark_slug": "ioi", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for IOI (Overall).", + "metric_id": "vals_ai.ioi.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/ioi" + } + }, + "score_details": { + "score": 22.0, + "details": { + "benchmark_slug": "ioi", + "benchmark_name": "IOI", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "22", + "raw_stderr": "7.832", + "latency": "6806.938", + "cost_per_test": "3.273674", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 7.832, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5-thinking/a0ca6709-5c3c-46fe-b827-c40a149569f8.json b/data/vals_ai/zai/glm-5-thinking/a0ca6709-5c3c-46fe-b827-c40a149569f8.json new file mode 100644 index 0000000000000000000000000000000000000000..318aff748f9046e7b8189e0c3849cb8dade867b2 --- /dev/null +++ b/data/vals_ai/zai/glm-5-thinking/a0ca6709-5c3c-46fe-b827-c40a149569f8.json @@ -0,0 +1,371 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/zai_glm-5-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5-thinking", + "id": "zai/glm-5-thinking", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5-thinking", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 35.714, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "35.714", + "raw_stderr": "7.483", + "latency": "748.809", + "cost_per_test": "0.749215", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 7.483, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 68.582, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "68.582", + "raw_stderr": "2.879", + "latency": "564.558", + "cost_per_test": "0.498572", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.879, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 83.505, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "83.505", + "raw_stderr": "2.671", + "latency": "391.99", + "cost_per_test": "0.31426", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.671, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "691.637", + "cost_per_test": "0.631472", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 71.4, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "71.4", + "raw_stderr": "2.023", + "latency": "513.842", + "cost_per_test": "0.448911", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.023, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5-thinking/ad61e394-f850-4bc4-b88e-0239a281e935.json b/data/vals_ai/zai/glm-5-thinking/ad61e394-f850-4bc4-b88e-0239a281e935.json new file mode 100644 index 0000000000000000000000000000000000000000..dddfdabe1f8b6452b98ab7ce5ed4f47cbf4dc1e3 --- /dev/null +++ b/data/vals_ai/zai/glm-5-thinking/ad61e394-f850-4bc4-b88e-0239a281e935.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/zai_glm-5-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5-thinking", + "id": "zai/glm-5-thinking", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5-thinking", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.282, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.282", + "raw_stderr": "0.867", + "latency": "57.433", + "cost_per_test": "0.007637", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.867, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.748, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.748", + "raw_stderr": "1.032", + "latency": "91.325", + "cost_per_test": "0.011059", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.032, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.813, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.813", + "raw_stderr": "0.859", + "latency": "115.706", + "cost_per_test": "0.017438", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.859, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.561, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.561", + "raw_stderr": "1.63", + "latency": "75.826", + "cost_per_test": "0.010569", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.63, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.915, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.915", + "raw_stderr": "1.122", + "latency": "71.636", + "cost_per_test": "0.007021", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.122, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.179, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.179", + "raw_stderr": "1.205", + "latency": "198.893", + "cost_per_test": "0.030891", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.205, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 80.196, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "80.196", + "raw_stderr": "1.393", + "latency": "82.493", + "cost_per_test": "0.008348", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.393, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 77.428, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "77.428", + "raw_stderr": "2.142", + "latency": "57.957", + "cost_per_test": "0.007885", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.142, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 72.843, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.843", + "raw_stderr": "1.34", + "latency": "97.959", + "cost_per_test": "0.012339", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.34, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.819, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.819", + "raw_stderr": "0.603", + "latency": "79.691", + "cost_per_test": "0.012724", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.603, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.199, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.199", + "raw_stderr": "1.2", + "latency": "53.559", + "cost_per_test": "0.006713", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.2, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.03, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.03", + "raw_stderr": "0.343", + "latency": "86.889", + "cost_per_test": "0.011542", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.343, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.166, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.166", + "raw_stderr": "1.683", + "latency": "73.251", + "cost_per_test": "0.00732", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.683, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.686, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.686", + "raw_stderr": "0.772", + "latency": "109.966", + "cost_per_test": "0.01562", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.772, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 85.589, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.589", + "raw_stderr": "1.243", + "latency": "50.749", + "cost_per_test": "0.006019", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.243, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5-thinking/f83e935d-d360-4faf-ae57-f751321a1804.json b/data/vals_ai/zai/glm-5-thinking/f83e935d-d360-4faf-ae57-f751321a1804.json new file mode 100644 index 0000000000000000000000000000000000000000..08b1c434b09ba7720756187b0d7db781a461e6a0 --- /dev/null +++ b/data/vals_ai/zai/glm-5-thinking/f83e935d-d360-4faf-ae57-f751321a1804.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/zai_glm-5-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5-thinking", + "id": "zai/glm-5-thinking", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5-thinking", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 82.828, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "82.828", + "raw_stderr": "2.68", + "latency": "234.216", + "cost_per_test": "0.031964", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.68, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 83.838, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.838", + "raw_stderr": "2.616", + "latency": "250.83", + "cost_per_test": "0.036218", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.616, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 83.333, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.333", + "raw_stderr": "1.873", + "latency": "242.523", + "cost_per_test": "0.034091", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.873, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5-thinking/fe16a269-2597-481f-be87-d6c7cacbba2b.json b/data/vals_ai/zai/glm-5-thinking/fe16a269-2597-481f-be87-d6c7cacbba2b.json new file mode 100644 index 0000000000000000000000000000000000000000..c023d7973628fba093452a5d90921fb1c614d562 --- /dev/null +++ b/data/vals_ai/zai/glm-5-thinking/fe16a269-2597-481f-be87-d6c7cacbba2b.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/zai_glm-5-thinking/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5-thinking", + "id": "zai/glm-5-thinking", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5-thinking", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.758, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.758", + "raw_stderr": "0.617", + "latency": "56.68", + "cost_per_test": "0.008151", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.617, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 58.857, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "58.857", + "raw_stderr": "2.63", + "latency": "565.964", + "cost_per_test": "0.084257", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.63, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 87.99, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "87.99", + "raw_stderr": "1.661", + "latency": "246.613", + "cost_per_test": "0.039346", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.661, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:zai/glm-5-thinking:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 81.868, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "81.868", + "raw_stderr": "1.057", + "latency": "289.752", + "cost_per_test": "0.043918", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.057, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.1/07fd66d5-1261-478b-a564-40f5a7cd34d8.json b/data/vals_ai/zai/glm-5.1/07fd66d5-1261-478b-a564-40f5a7cd34d8.json new file mode 100644 index 0000000000000000000000000000000000000000..02665a8cfa70b97c84483aed488dc1d8452766a0 --- /dev/null +++ b/data/vals_ai/zai/glm-5.1/07fd66d5-1261-478b-a564-40f5a7cd34d8.json @@ -0,0 +1,228 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/zai_glm-5.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.1", + "id": "zai/glm-5.1", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.1", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:zai/glm-5.1:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 61.538, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.538", + "raw_stderr": "1.392", + "latency": "84.072", + "cost_per_test": "0.016958", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.392, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:zai/glm-5.1:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 71.194, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "71.194", + "raw_stderr": "0.895", + "latency": "84.072", + "cost_per_test": "0.016958", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.895, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:zai/glm-5.1:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 80.851, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "80.851", + "raw_stderr": "1.126", + "latency": "84.072", + "cost_per_test": "0.016958", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.126, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.1/16e81ab5-50cd-462f-acc2-a7da7406dc80.json b/data/vals_ai/zai/glm-5.1/16e81ab5-50cd-462f-acc2-a7da7406dc80.json new file mode 100644 index 0000000000000000000000000000000000000000..84306fd76e26d512d3f04baf84a8a29b1b28ed5d --- /dev/null +++ b/data/vals_ai/zai/glm-5.1/16e81ab5-50cd-462f-acc2-a7da7406dc80.json @@ -0,0 +1,1041 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/zai_glm-5.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.1", + "id": "zai/glm-5.1", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.1", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:zai/glm-5.1:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 93.724, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "93.724", + "raw_stderr": "0.906", + "latency": "29.038", + "cost_per_test": "0.005623", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.906, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:zai/glm-5.1:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.128, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.128", + "raw_stderr": "1.012", + "latency": "66.948", + "cost_per_test": "0.012184", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.012, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:zai/glm-5.1:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.844, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.844", + "raw_stderr": "0.815", + "latency": "126.749", + "cost_per_test": "0.02417", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.815, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:zai/glm-5.1:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 89.487, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "89.487", + "raw_stderr": "1.517", + "latency": "44.147", + "cost_per_test": "0.007687", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.517, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:zai/glm-5.1:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 88.981, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "88.981", + "raw_stderr": "1.078", + "latency": "28.009", + "cost_per_test": "0.00488", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.078, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:zai/glm-5.1:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.029, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.029", + "raw_stderr": "1.184", + "latency": "266.126", + "cost_per_test": "0.049265", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.184, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:zai/glm-5.1:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.518, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.518", + "raw_stderr": "1.358", + "latency": "42.893", + "cost_per_test": "0.006832", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.358, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:zai/glm-5.1:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 78.421, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "78.421", + "raw_stderr": "2.11", + "latency": "32.724", + "cost_per_test": "0.006012", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.11, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:zai/glm-5.1:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 74.0, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "74", + "raw_stderr": "1.323", + "latency": "57.748", + "cost_per_test": "0.010337", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.323, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:zai/glm-5.1:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 95.485, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "95.485", + "raw_stderr": "0.565", + "latency": "62.824", + "cost_per_test": "0.01127", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.565, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:zai/glm-5.1:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.524, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.524", + "raw_stderr": "1.19", + "latency": "36.629", + "cost_per_test": "0.005805", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.19, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:zai/glm-5.1:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.9, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.9", + "raw_stderr": "0.334", + "latency": "69.318", + "cost_per_test": "0.012383", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.334, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:zai/glm-5.1:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.936, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.936", + "raw_stderr": "1.645", + "latency": "32.773", + "cost_per_test": "0.006152", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.645, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:zai/glm-5.1:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.302, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.302", + "raw_stderr": "0.74", + "latency": "117.665", + "cost_per_test": "0.018672", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.74, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:zai/glm-5.1:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 87.218, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.218", + "raw_stderr": "1.182", + "latency": "26.174", + "cost_per_test": "0.004468", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.182, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.1/19486c9b-1558-424a-bce1-3f91676c8c36.json b/data/vals_ai/zai/glm-5.1/19486c9b-1558-424a-bce1-3f91676c8c36.json new file mode 100644 index 0000000000000000000000000000000000000000..891400556fceddfe8a210773a41161caf3b90dce --- /dev/null +++ b/data/vals_ai/zai/glm-5.1/19486c9b-1558-424a-bce1-3f91676c8c36.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/public-benefits-bench/zai_glm-5.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Public Benefits Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "social mobility", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.1", + "id": "zai/glm-5.1", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.1", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "public-benefits-bench:overall:zai/glm-5.1:score", + "evaluation_name": "vals_ai.public-benefits-bench.overall", + "source_data": { + "dataset_name": "Public Benefits Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "public-benefits-bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Public Benefits Bench v1.1 (Overall).", + "metric_id": "vals_ai.public-benefits-bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/public-benefits-bench" + } + }, + "score_details": { + "score": 61.84, + "details": { + "benchmark_slug": "public-benefits-bench", + "benchmark_name": "Public Benefits Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "social mobility", + "raw_score": "61.84", + "raw_stderr": "1.264", + "latency": "725.11", + "cost_per_test": "0.340666", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.264, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.1/24d1ec90-c03f-49e4-8f2c-e91f16010a66.json b/data/vals_ai/zai/glm-5.1/24d1ec90-c03f-49e4-8f2c-e91f16010a66.json new file mode 100644 index 0000000000000000000000000000000000000000..a07d459920902042fd05defc013f9f1e01942eb6 --- /dev/null +++ b/data/vals_ai/zai/glm-5.1/24d1ec90-c03f-49e4-8f2c-e91f16010a66.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/zai_glm-5.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.1", + "id": "zai/glm-5.1", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.1", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:zai/glm-5.1:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 61.772, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.772", + "raw_stderr": "1.659", + "latency": "18.86", + "cost_per_test": "0.005086", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.659, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:zai/glm-5.1:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 66.317, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.317", + "raw_stderr": "1.614", + "latency": "139.915", + "cost_per_test": "0.15991", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.614, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:zai/glm-5.1:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 64.452, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "64.452", + "raw_stderr": "0.943", + "latency": "68.17", + "cost_per_test": "0.078922", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.943, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:zai/glm-5.1:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 65.268, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.268", + "raw_stderr": "1.625", + "latency": "45.736", + "cost_per_test": "0.071771", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.625, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.1/2ea4f548-1bec-4428-9f5b-ee54d459e8bd.json b/data/vals_ai/zai/glm-5.1/2ea4f548-1bec-4428-9f5b-ee54d459e8bd.json new file mode 100644 index 0000000000000000000000000000000000000000..e7ff58276585334bf1887cd471a0bf808ec206cb --- /dev/null +++ b/data/vals_ai/zai/glm-5.1/2ea4f548-1bec-4428-9f5b-ee54d459e8bd.json @@ -0,0 +1,408 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/zai_glm-5.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.1", + "id": "zai/glm-5.1", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.1", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:zai/glm-5.1:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 65.268, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "65.268", + "raw_stderr": "1.625", + "latency": "45.736", + "cost_per_test": "0.071771", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.625, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:zai/glm-5.1:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 44.79, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "44.79", + "raw_stderr": "0.347", + "latency": "897.798", + "cost_per_test": "0.584284", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.347, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:overall:zai/glm-5.1:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 52.452, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "52.452", + "raw_stderr": "1.607", + "latency": "899.497", + "cost_per_test": "0.863578", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.607, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:zai/glm-5.1:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 72.549, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "72.549", + "raw_stderr": "4.419", + "latency": "598.277", + "cost_per_test": "0.534539" + }, + "uncertainty": { + "standard_error": { + "value": 4.419, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:zai/glm-5.1:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 56.929, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "56.929", + "raw_stderr": "1.35", + "latency": "940.943", + "cost_per_test": "0.232386", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.35, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:zai/glm-5.1:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 32.804, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "32.804", + "raw_stderr": "7.069", + "latency": "2014.732", + "cost_per_test": "2.894911", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 7.069, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.1/3c6362a6-b10c-453a-b564-1a69680d5418.json b/data/vals_ai/zai/glm-5.1/3c6362a6-b10c-453a-b564-1a69680d5418.json new file mode 100644 index 0000000000000000000000000000000000000000..d55836a133ef52631d1d6d2b1bd016dc58e6ecc9 --- /dev/null +++ b/data/vals_ai/zai/glm-5.1/3c6362a6-b10c-453a-b564-1a69680d5418.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/zai_glm-5.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.1", + "id": "zai/glm-5.1", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.1", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:zai/glm-5.1:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 98.754, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "98.754", + "raw_stderr": "0.619", + "latency": "44.075", + "cost_per_test": "0.008931", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.619, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:hard:zai/glm-5.1:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 57.429, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "57.429", + "raw_stderr": "2.643", + "latency": "433.76", + "cost_per_test": "0.086606", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.643, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:medium:zai/glm-5.1:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 87.958, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "87.958", + "raw_stderr": "1.665", + "latency": "201.438", + "cost_per_test": "0.042412", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.665, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "lcb:overall:zai/glm-5.1:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 81.38, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "81.38", + "raw_stderr": "1.061", + "latency": "226.424", + "cost_per_test": "0.045983", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.061, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.1/4bc94baf-2340-4163-8cae-1e90efcacaed.json b/data/vals_ai/zai/glm-5.1/4bc94baf-2340-4163-8cae-1e90efcacaed.json new file mode 100644 index 0000000000000000000000000000000000000000..dcab1ce1eccf809e3a674c5655dbaede32acec5c --- /dev/null +++ b/data/vals_ai/zai/glm-5.1/4bc94baf-2340-4163-8cae-1e90efcacaed.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/zai_glm-5.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.1", + "id": "zai/glm-5.1", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.1", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:zai/glm-5.1:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 72.27, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "72.27", + "raw_stderr": "2.064", + "latency": "95.701", + "cost_per_test": "0.023717", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.064, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.1/54c1b81f-aba2-4462-a2b4-0bc51b458dd7.json b/data/vals_ai/zai/glm-5.1/54c1b81f-aba2-4462-a2b4-0bc51b458dd7.json new file mode 100644 index 0000000000000000000000000000000000000000..a05cab2b9a8a91621c8422b88e6614bfcd9b81d6 --- /dev/null +++ b/data/vals_ai/zai/glm-5.1/54c1b81f-aba2-4462-a2b4-0bc51b458dd7.json @@ -0,0 +1,292 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/code-migration/zai_glm-5.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Code Migration", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.1", + "id": "zai/glm-5.1", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.1", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "code-migration:cli:zai/glm-5.1:score", + "evaluation_name": "vals_ai.code-migration.cli", + "source_data": { + "dataset_name": "Code Migration - CLI", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cli", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (CLI).", + "metric_id": "vals_ai.code-migration.cli.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 18.833, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cli", + "task_name": "CLI", + "dataset_type": "private", + "industry": "coding", + "raw_score": "18.833", + "raw_stderr": "1.569", + "latency": "3837.721", + "cost_per_test": "6.652074", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.569, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "code-migration:cobol:zai/glm-5.1:score", + "evaluation_name": "vals_ai.code-migration.cobol", + "source_data": { + "dataset_name": "Code Migration - COBOL", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cobol", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (COBOL).", + "metric_id": "vals_ai.code-migration.cobol.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 46.571, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cobol", + "task_name": "COBOL", + "dataset_type": "private", + "industry": "coding", + "raw_score": "46.571", + "raw_stderr": "15.655", + "latency": "2496.289", + "cost_per_test": "2.84903", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 15.655, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "code-migration:code_quality:zai/glm-5.1:score", + "evaluation_name": "vals_ai.code-migration.code_quality", + "source_data": { + "dataset_name": "Code Migration - Code Quality", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "code_quality", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Code Quality).", + "metric_id": "vals_ai.code-migration.code_quality.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 3.544, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "code_quality", + "task_name": "Code Quality", + "dataset_type": "private", + "industry": "coding", + "raw_score": "3.544", + "raw_stderr": "0.061", + "latency": "3502.363", + "cost_per_test": "5.701313", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.061, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "code-migration:overall:zai/glm-5.1:score", + "evaluation_name": "vals_ai.code-migration.overall", + "source_data": { + "dataset_name": "Code Migration - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Overall).", + "metric_id": "vals_ai.code-migration.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 25.768, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "25.768", + "raw_stderr": "4.087", + "latency": "3502.363", + "cost_per_test": "5.701313", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 4.087, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.1/6484f067-ae1a-4bc3-8325-230d754f8125.json b/data/vals_ai/zai/glm-5.1/6484f067-ae1a-4bc3-8325-230d754f8125.json new file mode 100644 index 0000000000000000000000000000000000000000..d98ac9f07aeaa009284637aaf5188a8c205c9d37 --- /dev/null +++ b/data/vals_ai/zai/glm-5.1/6484f067-ae1a-4bc3-8325-230d754f8125.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/zai_glm-5.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.1", + "id": "zai/glm-5.1", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.1", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:zai/glm-5.1:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 22.222, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "22.222", + "raw_stderr": "4.178", + "latency": "4755.746", + "cost_per_test": "0.584577", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 4.178, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.1/673eb047-181d-4f39-94f0-be6523c6172a.json b/data/vals_ai/zai/glm-5.1/673eb047-181d-4f39-94f0-be6523c6172a.json new file mode 100644 index 0000000000000000000000000000000000000000..51062b8f64520cdff9af656925d462d0d12c8e8a --- /dev/null +++ b/data/vals_ai/zai/glm-5.1/673eb047-181d-4f39-94f0-be6523c6172a.json @@ -0,0 +1,438 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/zai_glm-5.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.1", + "id": "zai/glm-5.1", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.1", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:zai/glm-5.1:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 89.575, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.575", + "raw_stderr": "0.723", + "latency": "11.322", + "cost_per_test": "0.001422", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.723, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:zai/glm-5.1:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.709, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.709", + "raw_stderr": "0.212", + "latency": "14.32", + "cost_per_test": "0.002309", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.212, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:zai/glm-5.1:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 89.02, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.02", + "raw_stderr": "0.588", + "latency": "10.961", + "cost_per_test": "0.001949", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.588, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:zai/glm-5.1:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 84.394, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.394", + "raw_stderr": "0.429", + "latency": "13.749", + "cost_per_test": "0.002183", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.429, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:zai/glm-5.1:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.108, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.108", + "raw_stderr": "0.889", + "latency": "12.873", + "cost_per_test": "0.001795", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.889, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:zai/glm-5.1:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 80.56, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.56", + "raw_stderr": "1.7", + "latency": "19.27", + "cost_per_test": "0.003438", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.7, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.1/7130a848-0e9e-44b3-b2c0-820d50a2deb6.json b/data/vals_ai/zai/glm-5.1/7130a848-0e9e-44b3-b2c0-820d50a2deb6.json new file mode 100644 index 0000000000000000000000000000000000000000..06be47aed825b3788a9683d0967e56c47ab92c11 --- /dev/null +++ b/data/vals_ai/zai/glm-5.1/7130a848-0e9e-44b3-b2c0-820d50a2deb6.json @@ -0,0 +1,3386 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/hlab/zai_glm-5.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Harvey's Legal Agent Benchmark", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.1", + "id": "zai/glm-5.1", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.1", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "hlab:antitrust_competition:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "525.641", + "cost_per_test": "0.679439", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:arbitration_international_dispute_resolution:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "429.529", + "cost_per_test": "0.563167", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:banking_finance:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "412.219", + "cost_per_test": "0.507355", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:bankruptcy_restructuring:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "452.897", + "cost_per_test": "0.62914", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:capital_markets:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "392.204", + "cost_per_test": "0.443784", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:corporate_governance:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "463.866", + "cost_per_test": "0.55015", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:corporate_ma:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "498.818", + "cost_per_test": "0.580534", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Criteria Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Criteria Pass Rate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 71.718, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate", + "task_name": "Criteria Pass Rate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "71.718", + "raw_stderr": "0.542", + "latency": "496.63", + "cost_per_test": "0.555636", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.542, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__antitrust_competition:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 71.377, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "71.377", + "raw_stderr": "2.726", + "latency": "525.641", + "cost_per_test": "0.679439", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.726, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__arbitration_international_dispute_resolution:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 53.405, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "53.405", + "raw_stderr": "2.992", + "latency": "429.529", + "cost_per_test": "0.563167", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.992, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__banking_finance:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 71.004, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "71.004", + "raw_stderr": "2.772", + "latency": "412.219", + "cost_per_test": "0.507355", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.772, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__bankruptcy_restructuring:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 75.432, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "75.432", + "raw_stderr": "2.537", + "latency": "452.897", + "cost_per_test": "0.62914", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.537, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__capital_markets:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 71.765, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "71.765", + "raw_stderr": "2.824", + "latency": "392.204", + "cost_per_test": "0.443784", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.824, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_governance:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 72.945, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "72.945", + "raw_stderr": "2.604", + "latency": "463.866", + "cost_per_test": "0.55015", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.604, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_ma:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 85.897, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.897", + "raw_stderr": "1.974", + "latency": "498.818", + "cost_per_test": "0.580534", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.974, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__data_privacy_cybersecurity:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 88.693, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.693", + "raw_stderr": "1.886", + "latency": "560.689", + "cost_per_test": "0.628345", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.886, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__emerging_companies_venture_capital:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 70.307, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "70.307", + "raw_stderr": "2.674", + "latency": "453.289", + "cost_per_test": "0.487189", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.674, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__employment_labor:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 71.91, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "71.91", + "raw_stderr": "2.756", + "latency": "562.235", + "cost_per_test": "0.552987", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.756, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__energy_natural_resources:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 55.446, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "55.446", + "raw_stderr": "2.86", + "latency": "335.583", + "cost_per_test": "0.495202", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.86, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__environmental_esg:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 78.495, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.495", + "raw_stderr": "2.464", + "latency": "612.022", + "cost_per_test": "0.620051", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.464, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__funds_asset_management:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 79.883, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.883", + "raw_stderr": "2.168", + "latency": "683.729", + "cost_per_test": "0.718616", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.168, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__healthcare_life_sciences:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 67.987, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "67.987", + "raw_stderr": "2.685", + "latency": "616.135", + "cost_per_test": "0.759264", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.685, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__immigration:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 85.454, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.454", + "raw_stderr": "2.13", + "latency": "537.588", + "cost_per_test": "0.513525", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.13, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__insurance:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 80.586, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.586", + "raw_stderr": "2.398", + "latency": "526.817", + "cost_per_test": "0.601521", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.398, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__intellectual_property:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 86.219, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.219", + "raw_stderr": "2.053", + "latency": "524.275", + "cost_per_test": "0.593024", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.053, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__international_trade_sanctions:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 80.159, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "80.159", + "raw_stderr": "2.517", + "latency": "632.28", + "cost_per_test": "0.653116", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.517, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__litigation_dispute_resolution:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 69.288, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "69.288", + "raw_stderr": "2.828", + "latency": "485.681", + "cost_per_test": "0.447406", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.828, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__real_estate:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 49.854, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "49.854", + "raw_stderr": "2.704", + "latency": "387.328", + "cost_per_test": "0.461787", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.704, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__structured_finance_securitization:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 53.016, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "53.016", + "raw_stderr": "2.817", + "latency": "382.798", + "cost_per_test": "0.396449", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.817, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__tax:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 67.893, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "67.893", + "raw_stderr": "2.705", + "latency": "454.745", + "cost_per_test": "0.530065", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.705, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__trusts_estates_private_client:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 74.046, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "74.046", + "raw_stderr": "2.714", + "latency": "397.802", + "cost_per_test": "0.400358", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.714, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__white_collar_defense_investigations:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 67.003, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "67.003", + "raw_stderr": "2.733", + "latency": "590.945", + "cost_per_test": "0.522794", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.733, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:data_privacy_cybersecurity:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "560.689", + "cost_per_test": "0.628345", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:emerging_companies_venture_capital:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "453.289", + "cost_per_test": "0.487189", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:employment_labor:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "562.235", + "cost_per_test": "0.552987", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:energy_natural_resources:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "335.583", + "cost_per_test": "0.495202", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:environmental_esg:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "612.022", + "cost_per_test": "0.620051", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:funds_asset_management:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "683.729", + "cost_per_test": "0.718616", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:healthcare_life_sciences:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "616.135", + "cost_per_test": "0.759264", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:immigration:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "537.588", + "cost_per_test": "0.513525", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:insurance:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "526.817", + "cost_per_test": "0.601521", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:intellectual_property:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "524.275", + "cost_per_test": "0.593024", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:international_trade_sanctions:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "632.28", + "cost_per_test": "0.653116", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:litigation_dispute_resolution:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "485.681", + "cost_per_test": "0.447406", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:overall:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.overall", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Overall).", + "metric_id": "vals_ai.hlab.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "496.63", + "cost_per_test": "0.555636", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:real_estate:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "387.328", + "cost_per_test": "0.461787", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:structured_finance_securitization:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "382.798", + "cost_per_test": "0.396449", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:tax:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "454.745", + "cost_per_test": "0.530065", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:trusts_estates_private_client:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "397.802", + "cost_per_test": "0.400358", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "hlab:white_collar_defense_investigations:zai/glm-5.1:score", + "evaluation_name": "vals_ai.hlab.white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "590.945", + "cost_per_test": "0.522794", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.1/71dcf01e-1747-47d1-8999-5662e52c6f72.json b/data/vals_ai/zai/glm-5.1/71dcf01e-1747-47d1-8999-5662e52c6f72.json new file mode 100644 index 0000000000000000000000000000000000000000..7622a2dedbcfc9990c0fa92b9b968f1e9daf646d --- /dev/null +++ b/data/vals_ai/zai/glm-5.1/71dcf01e-1747-47d1-8999-5662e52c6f72.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/programbench/zai_glm-5.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProgramBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.1", + "id": "zai/glm-5.1", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.1", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "programbench:almost:zai/glm-5.1:score", + "evaluation_name": "vals_ai.programbench.almost", + "source_data": { + "dataset_name": "ProgramBench - Almost Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "almost", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Almost Resolved).", + "metric_id": "vals_ai.programbench.almost.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 1.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "almost", + "task_name": "Almost Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "1", + "raw_stderr": "0.705", + "latency": "3931.735", + "cost_per_test": "5.598531", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.705, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "programbench:overall:zai/glm-5.1:score", + "evaluation_name": "vals_ai.programbench.overall", + "source_data": { + "dataset_name": "ProgramBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Overall).", + "metric_id": "vals_ai.programbench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "3931.735", + "cost_per_test": "5.598531", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "programbench:partial:zai/glm-5.1:score", + "evaluation_name": "vals_ai.programbench.partial", + "source_data": { + "dataset_name": "ProgramBench - Raw Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "partial", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Raw Pass Rate).", + "metric_id": "vals_ai.programbench.partial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 50.879, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "partial", + "task_name": "Raw Pass Rate", + "dataset_type": "public", + "industry": "coding", + "raw_score": "50.879", + "raw_stderr": "1.697", + "latency": "3931.735", + "cost_per_test": "5.598531", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.697, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "programbench:strict:zai/glm-5.1:score", + "evaluation_name": "vals_ai.programbench.strict", + "source_data": { + "dataset_name": "ProgramBench - Fully Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "strict", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Fully Resolved).", + "metric_id": "vals_ai.programbench.strict.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "strict", + "task_name": "Fully Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0", + "raw_stderr": "0", + "latency": "3931.735", + "cost_per_test": "5.598531", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.1/9db96cb1-7e49-4c56-86e1-dd9eae6275b6.json b/data/vals_ai/zai/glm-5.1/9db96cb1-7e49-4c56-86e1-dd9eae6275b6.json new file mode 100644 index 0000000000000000000000000000000000000000..73d66f41642b16f0b045879df1baa435e73f4a64 --- /dev/null +++ b/data/vals_ai/zai/glm-5.1/9db96cb1-7e49-4c56-86e1-dd9eae6275b6.json @@ -0,0 +1,304 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/zai_glm-5.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.1", + "id": "zai/glm-5.1", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.1", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:zai/glm-5.1:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 83.333, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "83.333", + "raw_stderr": "8.333", + "latency": "20935.973", + "cost_per_test": "5.170582", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 8.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:zai/glm-5.1:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 37.778, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "37.778", + "raw_stderr": "1.111", + "latency": "2791.463", + "cost_per_test": "0.689411", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.111, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:zai/glm-5.1:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 65.455, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "65.455", + "raw_stderr": "1.05", + "latency": "1522.616", + "cost_per_test": "0.376042", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.05, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:zai/glm-5.1:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 56.929, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "56.929", + "raw_stderr": "1.35", + "latency": "940.943", + "cost_per_test": "0.232386", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.35, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.1/a13c5d09-c494-4402-a561-7ba53a96aed9.json b/data/vals_ai/zai/glm-5.1/a13c5d09-c494-4402-a561-7ba53a96aed9.json new file mode 100644 index 0000000000000000000000000000000000000000..f17442d702f2280dc7e9a8ea13489f03884fea4b --- /dev/null +++ b/data/vals_ai/zai/glm-5.1/a13c5d09-c494-4402-a561-7ba53a96aed9.json @@ -0,0 +1,1188 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_research/zai_glm-5.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Legal Research Bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.1", + "id": "zai/glm-5.1", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.1", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_research:administrative_regulatory:zai/glm-5.1:score", + "evaluation_name": "vals_ai.legal_research.administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 31.818, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "31.818", + "raw_stderr": "5.777", + "latency": "307.232", + "cost_per_test": "0.323431", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 5.777, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:business_commercial:zai/glm-5.1:score", + "evaluation_name": "vals_ai.legal_research.business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 23.256, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "23.256", + "raw_stderr": "6.519", + "latency": "336.248", + "cost_per_test": "0.399162", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 6.519, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:civil_litigation:zai/glm-5.1:score", + "evaluation_name": "vals_ai.legal_research.civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 22.222, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "22.222", + "raw_stderr": "8.153", + "latency": "361.704", + "cost_per_test": "0.473442", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 8.153, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:constitutional_civil_rights:zai/glm-5.1:score", + "evaluation_name": "vals_ai.legal_research.constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 34.211, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "34.211", + "raw_stderr": "7.799", + "latency": "355.395", + "cost_per_test": "0.362114", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 7.799, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:criminal:zai/glm-5.1:score", + "evaluation_name": "vals_ai.legal_research.criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 29.825, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "29.825", + "raw_stderr": "6.113", + "latency": "331.685", + "cost_per_test": "0.405872", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 6.113, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:family:zai/glm-5.1:score", + "evaluation_name": "vals_ai.legal_research.family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 9.091, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "9.091", + "raw_stderr": "6.273", + "latency": "342.126", + "cost_per_test": "0.41232", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 6.273, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:health:zai/glm-5.1:score", + "evaluation_name": "vals_ai.legal_research.health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 36.667, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "36.667", + "raw_stderr": "8.949", + "latency": "322.342", + "cost_per_test": "0.321248", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 8.949, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:immigration:zai/glm-5.1:score", + "evaluation_name": "vals_ai.legal_research.immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 18.182, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "18.182", + "raw_stderr": "12.197", + "latency": "389.141", + "cost_per_test": "0.560419", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 12.197, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:overall:zai/glm-5.1:score", + "evaluation_name": "vals_ai.legal_research.overall", + "source_data": { + "dataset_name": "Legal Research Bench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Overall).", + "metric_id": "vals_ai.legal_research.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 27.885, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "legal", + "raw_score": "27.885", + "raw_stderr": "3.117", + "latency": "325.824", + "cost_per_test": "0.385737", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.117, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass:zai/glm-5.1:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass", + "source_data": { + "dataset_name": "Legal Research Bench - Weighted Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Weighted Pass).", + "metric_id": "vals_ai.legal_research.weighted_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 77.808, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass", + "task_name": "Weighted Pass", + "dataset_type": "private", + "industry": "legal", + "raw_score": "77.808", + "raw_stderr": "1.696", + "latency": "325.824", + "cost_per_test": "0.385737", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.696, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__administrative_regulatory:zai/glm-5.1:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.weighted_pass__administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 80.005, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "80.005", + "raw_stderr": "3.081", + "latency": "307.232", + "cost_per_test": "0.323431", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.081, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__business_commercial:zai/glm-5.1:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.weighted_pass__business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 81.655, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "81.655", + "raw_stderr": "2.98", + "latency": "336.248", + "cost_per_test": "0.399162", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.98, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__civil_litigation:zai/glm-5.1:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.weighted_pass__civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 83.513, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "83.513", + "raw_stderr": "3.846", + "latency": "361.704", + "cost_per_test": "0.473442", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.846, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__constitutional_civil_rights:zai/glm-5.1:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 84.883, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "84.883", + "raw_stderr": "2.734", + "latency": "355.395", + "cost_per_test": "0.362114", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.734, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__criminal:zai/glm-5.1:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.weighted_pass__criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 78.143, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "78.143", + "raw_stderr": "3.194", + "latency": "331.685", + "cost_per_test": "0.405872", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.194, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__family:zai/glm-5.1:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.weighted_pass__family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 58.154, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "58.154", + "raw_stderr": "5.699", + "latency": "342.126", + "cost_per_test": "0.41232", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 5.699, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__health:zai/glm-5.1:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.weighted_pass__health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 82.755, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "82.755", + "raw_stderr": "3.924", + "latency": "322.342", + "cost_per_test": "0.321248", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.924, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__immigration:zai/glm-5.1:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.weighted_pass__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 80.706, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "80.706", + "raw_stderr": "7.246", + "latency": "389.141", + "cost_per_test": "0.560419", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 7.246, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.1/aef64909-c8eb-4f64-bf6d-59b8d06d82b4.json b/data/vals_ai/zai/glm-5.1/aef64909-c8eb-4f64-bf6d-59b8d06d82b4.json new file mode 100644 index 0000000000000000000000000000000000000000..9072f5661b69c0102bb4e115650d546493962628 --- /dev/null +++ b/data/vals_ai/zai/glm-5.1/aef64909-c8eb-4f64-bf6d-59b8d06d82b4.json @@ -0,0 +1,371 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/zai_glm-5.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.1", + "id": "zai/glm-5.1", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.1", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:zai/glm-5.1:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 45.238, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "45.238", + "raw_stderr": "7.773", + "latency": "891.993", + "cost_per_test": "0.895768", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 7.773, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:zai/glm-5.1:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 72.797, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "72.797", + "raw_stderr": "2.76", + "latency": "576.031", + "cost_per_test": "0.503938", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.76, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:zai/glm-5.1:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 88.66, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "88.66", + "raw_stderr": "2.282", + "latency": "374.733", + "cost_per_test": "0.293252", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.282, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:zai/glm-5.1:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 33.333, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "33.333", + "raw_stderr": "33.333", + "latency": "1009.772", + "cost_per_test": "0.75351", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "swebench:overall:zai/glm-5.1:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 76.4, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "76.4", + "raw_stderr": "1.901", + "latency": "527.071", + "cost_per_test": "0.456603", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.901, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.1/b44690b9-d9ce-4ec3-9918-335f478ec218.json b/data/vals_ai/zai/glm-5.1/b44690b9-d9ce-4ec3-9918-335f478ec218.json new file mode 100644 index 0000000000000000000000000000000000000000..ad9a7981e53d98544c00546d1c85972863603d63 --- /dev/null +++ b/data/vals_ai/zai/glm-5.1/b44690b9-d9ce-4ec3-9918-335f478ec218.json @@ -0,0 +1,237 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/zai_glm-5.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.1", + "id": "zai/glm-5.1", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.1", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:zai/glm-5.1:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 85.787, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.787", + "raw_stderr": "2.488", + "latency": "197.756", + "cost_per_test": "0.034962", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.488, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:zai/glm-5.1:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 83.249, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.249", + "raw_stderr": "2.661", + "latency": "301.653", + "cost_per_test": "0.054832", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.661, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "gpqa:overall:zai/glm-5.1:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 84.518, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.518", + "raw_stderr": "1.821", + "latency": "249.704", + "cost_per_test": "0.044897", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.821, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.1/dc557553-fa00-499f-a076-666c72b97c60.json b/data/vals_ai/zai/glm-5.1/dc557553-fa00-499f-a076-666c72b97c60.json new file mode 100644 index 0000000000000000000000000000000000000000..695e284fec0ae43fc6590fed181e2d7fb74bbc51 --- /dev/null +++ b/data/vals_ai/zai/glm-5.1/dc557553-fa00-499f-a076-666c72b97c60.json @@ -0,0 +1,740 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/zai_glm-5.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.1", + "id": "zai/glm-5.1", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.1", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:zai/glm-5.1:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 43.276, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "43.276", + "raw_stderr": "3.152", + "latency": "936.369", + "cost_per_test": "0.586714", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.152, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:zai/glm-5.1:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 32.68, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "32.68", + "raw_stderr": "0.707", + "latency": "924.078", + "cost_per_test": "0.618185", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.707, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:zai/glm-5.1:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 41.557, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "41.557", + "raw_stderr": "0.234", + "latency": "1058.07", + "cost_per_test": "0.750335", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.234, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:zai/glm-5.1:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 52.713, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.713", + "raw_stderr": "2.288", + "latency": "955.188", + "cost_per_test": "0.690792", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.288, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:zai/glm-5.1:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 56.958, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "56.958", + "raw_stderr": "0.613", + "latency": "526.957", + "cost_per_test": "0.319196", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.613, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:zai/glm-5.1:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 13.466, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "13.466", + "raw_stderr": "1.012", + "latency": "1345.298", + "cost_per_test": "0.848763", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.012, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:zai/glm-5.1:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 63.665, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.665", + "raw_stderr": "0.193", + "latency": "761.661", + "cost_per_test": "0.494021", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.193, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:zai/glm-5.1:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 61.87, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.87", + "raw_stderr": "1.774", + "latency": "698.813", + "cost_per_test": "0.556569", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.774, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:zai/glm-5.1:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 51.861, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "51.861", + "raw_stderr": "1.156", + "latency": "484.552", + "cost_per_test": "0.352587", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.156, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:overall:zai/glm-5.1:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 44.792, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "44.792", + "raw_stderr": "0.606", + "latency": "924.078", + "cost_per_test": "0.618185", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.606, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:zai/glm-5.1:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 17.764, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "17.764", + "raw_stderr": "2.425", + "latency": "1623.858", + "cost_per_test": "0.997596", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.425, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.1/ee50175e-77ec-473b-b0e2-97d5b2864604.json b/data/vals_ai/zai/glm-5.1/ee50175e-77ec-473b-b0e2-97d5b2864604.json new file mode 100644 index 0000000000000000000000000000000000000000..ec0cb531764932b67d3b378b55111c644a394515 --- /dev/null +++ b/data/vals_ai/zai/glm-5.1/ee50175e-77ec-473b-b0e2-97d5b2864604.json @@ -0,0 +1,100 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/zai_glm-5.1/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.1", + "id": "zai/glm-5.1", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.1", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:zai/glm-5.1:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 31.456, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "31.456", + "raw_stderr": "4.553", + "latency": "2014.732", + "cost_per_test": "2.894911", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 4.553, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.2/0dfd7c62-624c-4cf2-af08-81835547284a.json b/data/vals_ai/zai/glm-5.2/0dfd7c62-624c-4cf2-af08-81835547284a.json new file mode 100644 index 0000000000000000000000000000000000000000..048f013e232874d75daf92972119983130c60cce --- /dev/null +++ b/data/vals_ai/zai/glm-5.2/0dfd7c62-624c-4cf2-af08-81835547284a.json @@ -0,0 +1,240 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/tax_eval_v2/zai_glm-5.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - TaxEval v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.2", + "id": "zai/glm-5.2", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.2", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "tax_eval_v2:correctness:zai/glm-5.2:score", + "evaluation_name": "vals_ai.tax_eval_v2.correctness", + "source_data": { + "dataset_name": "TaxEval v2 - Correctness", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "correctness", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Correctness).", + "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 62.061, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "correctness", + "task_name": "Correctness", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.061", + "raw_stderr": "1.388", + "latency": "64.942", + "cost_per_test": "0.015431", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.388, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:overall:zai/glm-5.2:score", + "evaluation_name": "vals_ai.tax_eval_v2.overall", + "source_data": { + "dataset_name": "TaxEval v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Overall).", + "metric_id": "vals_ai.tax_eval_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 73.344, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "73.344", + "raw_stderr": "0.865", + "latency": "64.942", + "cost_per_test": "0.015431", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.865, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:zai/glm-5.2:score", + "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning", + "source_data": { + "dataset_name": "TaxEval v2 - Stepwise Reasoning", + "source_type": "other", + "additional_details": { + "benchmark_slug": "tax_eval_v2", + "task_key": "stepwise_reasoning", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for TaxEval v2 (Stepwise Reasoning).", + "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2" + } + }, + "score_details": { + "score": 84.628, + "details": { + "benchmark_slug": "tax_eval_v2", + "benchmark_name": "TaxEval v2", + "benchmark_updated": "2026-07-09", + "task_key": "stepwise_reasoning", + "task_name": "Stepwise Reasoning", + "dataset_type": "private", + "industry": "finance", + "raw_score": "84.628", + "raw_stderr": "1.031", + "latency": "64.942", + "cost_per_test": "0.015431", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.031, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.2/17a1d7a1-2ff2-452c-8719-c5c92487dbab.json b/data/vals_ai/zai/glm-5.2/17a1d7a1-2ff2-452c-8719-c5c92487dbab.json new file mode 100644 index 0000000000000000000000000000000000000000..5eba991c6570e11627decc00d4fff6e6ba8c2bda --- /dev/null +++ b/data/vals_ai/zai/glm-5.2/17a1d7a1-2ff2-452c-8719-c5c92487dbab.json @@ -0,0 +1,391 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/swebench/zai_glm-5.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - SWE-bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.2", + "id": "zai/glm-5.2", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.2", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "swebench:1-4 hours:zai/glm-5.2:score", + "evaluation_name": "vals_ai.swebench.1-4 hours", + "source_data": { + "dataset_name": "SWE-bench - 1-4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "1-4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (1-4 hours).", + "metric_id": "vals_ai.swebench.1-4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "1-4 hours", + "task_name": "1-4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66.667", + "raw_stderr": "7.362", + "latency": "1151.401", + "cost_per_test": "1.213786", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 7.362, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:15 min - 1 hour:zai/glm-5.2:score", + "evaluation_name": "vals_ai.swebench.15 min - 1 hour", + "source_data": { + "dataset_name": "SWE-bench - 15 min - 1 hour", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "15 min - 1 hour", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (15 min - 1 hour).", + "metric_id": "vals_ai.swebench.15 min - 1 hour.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 81.609, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "15 min - 1 hour", + "task_name": "15 min - 1 hour", + "dataset_type": "public", + "industry": "coding", + "raw_score": "81.609", + "raw_stderr": "2.403", + "latency": "734.948", + "cost_per_test": "0.627143", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.403, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:<15 min fix:zai/glm-5.2:score", + "evaluation_name": "vals_ai.swebench.<15 min fix", + "source_data": { + "dataset_name": "SWE-bench - <15 min fix", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "<15 min fix", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (<15 min fix).", + "metric_id": "vals_ai.swebench.<15 min fix.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 88.144, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "<15 min fix", + "task_name": "<15 min fix", + "dataset_type": "public", + "industry": "coding", + "raw_score": "88.144", + "raw_stderr": "2.327", + "latency": "462.996", + "cost_per_test": "0.36378", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.327, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:>4 hours:zai/glm-5.2:score", + "evaluation_name": "vals_ai.swebench.>4 hours", + "source_data": { + "dataset_name": "SWE-bench - >4 hours", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": ">4 hours", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (>4 hours).", + "metric_id": "vals_ai.swebench.>4 hours.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 66.667, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": ">4 hours", + "task_name": ">4 hours", + "dataset_type": "public", + "industry": "coding", + "raw_score": "66.667", + "raw_stderr": "33.333", + "latency": "1610.617", + "cost_per_test": "1.341924", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 33.333, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "swebench:overall:zai/glm-5.2:score", + "evaluation_name": "vals_ai.swebench.overall", + "source_data": { + "dataset_name": "SWE-bench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/swebench" + ], + "additional_details": { + "benchmark_slug": "swebench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for SWE-bench (Overall).", + "metric_id": "vals_ai.swebench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/swebench" + } + }, + "score_details": { + "score": 82.8, + "details": { + "benchmark_slug": "swebench", + "benchmark_name": "SWE-bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "82.8", + "raw_stderr": "1.689", + "latency": "669.667", + "cost_per_test": "0.710002", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.689, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.2/2200baff-3120-4c82-98df-28fecb0eac91.json b/data/vals_ai/zai/glm-5.2/2200baff-3120-4c82-98df-28fecb0eac91.json new file mode 100644 index 0000000000000000000000000000000000000000..5efe6aea89270c305be0decfee07f3b1a73ca8be --- /dev/null +++ b/data/vals_ai/zai/glm-5.2/2200baff-3120-4c82-98df-28fecb0eac91.json @@ -0,0 +1,195 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/cyber/zai_glm-5.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CyberBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "beta", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.2", + "id": "zai/glm-5.2", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.2", + "vals_provider": "Z.ai" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "cyber:overall:zai/glm-5.2:score", + "evaluation_name": "vals_ai.cyber.overall", + "source_data": { + "dataset_name": "CyberBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (Overall).", + "metric_id": "vals_ai.cyber.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 77.359, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "beta", + "raw_score": "77.359", + "raw_stderr": "5.49", + "latency": "1349.142", + "cost_per_test": "1.206778", + "provider": "Z.ai" + }, + "uncertainty": { + "standard_error": { + "value": 5.49, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "cyber:patch:zai/glm-5.2:score", + "evaluation_name": "vals_ai.cyber.patch", + "source_data": { + "dataset_name": "CyberBench - Patch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "patch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (Patch).", + "metric_id": "vals_ai.cyber.patch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 81.034, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "patch", + "task_name": "Patch", + "dataset_type": "private", + "industry": "beta", + "raw_score": "81.034", + "raw_stderr": "5.148", + "latency": "1549.48", + "cost_per_test": "0.849708", + "provider": "Z.ai" + }, + "uncertainty": { + "standard_error": { + "value": 5.148, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "cyber:poc:zai/glm-5.2:score", + "evaluation_name": "vals_ai.cyber.poc", + "source_data": { + "dataset_name": "CyberBench - PoC", + "source_type": "other", + "additional_details": { + "benchmark_slug": "cyber", + "task_key": "poc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CyberBench (PoC).", + "metric_id": "vals_ai.cyber.poc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/cyber" + } + }, + "score_details": { + "score": 73.684, + "details": { + "benchmark_slug": "cyber", + "benchmark_name": "CyberBench", + "benchmark_updated": "2026-07-09", + "task_key": "poc", + "task_name": "PoC", + "dataset_type": "private", + "industry": "beta", + "raw_score": "73.684", + "raw_stderr": "5.833", + "latency": "1148.804", + "cost_per_test": "1.563847", + "provider": "Z.ai" + }, + "uncertainty": { + "standard_error": { + "value": 5.833, + "method": "vals_reported" + } + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.2/31af49cf-0edd-4a16-87af-f6eddd693366.json b/data/vals_ai/zai/glm-5.2/31af49cf-0edd-4a16-87af-f6eddd693366.json new file mode 100644 index 0000000000000000000000000000000000000000..2ae7557361bded312357d28d16a07bf97e360acd --- /dev/null +++ b/data/vals_ai/zai/glm-5.2/31af49cf-0edd-4a16-87af-f6eddd693366.json @@ -0,0 +1,316 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/programbench/zai_glm-5.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProgramBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.2", + "id": "zai/glm-5.2", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.2", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "programbench:almost:zai/glm-5.2:score", + "evaluation_name": "vals_ai.programbench.almost", + "source_data": { + "dataset_name": "ProgramBench - Almost Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "almost", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Almost Resolved).", + "metric_id": "vals_ai.programbench.almost.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 9.5, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "almost", + "task_name": "Almost Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "9.5", + "raw_stderr": "2.079", + "latency": "7588.932", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "64000", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.079, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "programbench:overall:zai/glm-5.2:score", + "evaluation_name": "vals_ai.programbench.overall", + "source_data": { + "dataset_name": "ProgramBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Overall).", + "metric_id": "vals_ai.programbench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.5, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0.5", + "raw_stderr": "0.5", + "latency": "7588.932", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "64000", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.5, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "programbench:partial:zai/glm-5.2:score", + "evaluation_name": "vals_ai.programbench.partial", + "source_data": { + "dataset_name": "ProgramBench - Raw Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "partial", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Raw Pass Rate).", + "metric_id": "vals_ai.programbench.partial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 62.605, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "partial", + "task_name": "Raw Pass Rate", + "dataset_type": "public", + "industry": "coding", + "raw_score": "62.605", + "raw_stderr": "2.026", + "latency": "7588.932", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "64000", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.026, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "programbench:strict:zai/glm-5.2:score", + "evaluation_name": "vals_ai.programbench.strict", + "source_data": { + "dataset_name": "ProgramBench - Fully Resolved", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/programbench" + ], + "additional_details": { + "benchmark_slug": "programbench", + "task_key": "strict", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProgramBench (Fully Resolved).", + "metric_id": "vals_ai.programbench.strict.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/programbench" + } + }, + "score_details": { + "score": 0.5, + "details": { + "benchmark_slug": "programbench", + "benchmark_name": "ProgramBench", + "benchmark_updated": "2026-07-09", + "task_key": "strict", + "task_name": "Fully Resolved", + "dataset_type": "public", + "industry": "coding", + "raw_score": "0.5", + "raw_stderr": "0.5", + "latency": "7588.932", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "64000", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.5, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 64000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.2/3ddf1b2c-ba69-40d6-9e0f-4e3fda73c46e.json b/data/vals_ai/zai/glm-5.2/3ddf1b2c-ba69-40d6-9e0f-4e3fda73c46e.json new file mode 100644 index 0000000000000000000000000000000000000000..68a62457e591577518d24cfff00d8930d6b7fda4 --- /dev/null +++ b/data/vals_ai/zai/glm-5.2/3ddf1b2c-ba69-40d6-9e0f-4e3fda73c46e.json @@ -0,0 +1,784 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/fabv2/zai_glm-5.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Finance Agent (v2)", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.2", + "id": "zai/glm-5.2", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.2", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "fabv2:adjustments:zai/glm-5.2:score", + "evaluation_name": "vals_ai.fabv2.adjustments", + "source_data": { + "dataset_name": "Finance Agent (v2) - Adjustments", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "adjustments", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Adjustments).", + "metric_id": "vals_ai.fabv2.adjustments.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 46.305, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "adjustments", + "task_name": "Adjustments", + "dataset_type": "private", + "industry": "finance", + "raw_score": "46.305", + "raw_stderr": "2.981", + "latency": "410.892", + "cost_per_test": "0.619233", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.981, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:all_pass:zai/glm-5.2:score", + "evaluation_name": "vals_ai.fabv2.all_pass", + "source_data": { + "dataset_name": "Finance Agent (v2) - All-Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "all_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (All-Pass).", + "metric_id": "vals_ai.fabv2.all_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 38.006, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "all_pass", + "task_name": "All-Pass", + "dataset_type": "private", + "industry": "finance", + "raw_score": "38.006", + "raw_stderr": "0.786", + "latency": "473.158", + "cost_per_test": "0.713981", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.786, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:comparables:zai/glm-5.2:score", + "evaluation_name": "vals_ai.fabv2.comparables", + "source_data": { + "dataset_name": "Finance Agent (v2) - Comparables", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "comparables", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Comparables).", + "metric_id": "vals_ai.fabv2.comparables.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 44.78, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "comparables", + "task_name": "Comparables", + "dataset_type": "private", + "industry": "finance", + "raw_score": "44.78", + "raw_stderr": "0.922", + "latency": "648.959", + "cost_per_test": "1.029011", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.922, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:disclosure_analysis:zai/glm-5.2:score", + "evaluation_name": "vals_ai.fabv2.disclosure_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Disclosure Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "disclosure_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Disclosure Analysis).", + "metric_id": "vals_ai.fabv2.disclosure_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 61.588, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "disclosure_analysis", + "task_name": "Disclosure Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.588", + "raw_stderr": "1.686", + "latency": "372.664", + "cost_per_test": "0.714998", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.686, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:earnings_analysis:zai/glm-5.2:score", + "evaluation_name": "vals_ai.fabv2.earnings_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Earnings Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "earnings_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Earnings Analysis).", + "metric_id": "vals_ai.fabv2.earnings_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 66.596, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "earnings_analysis", + "task_name": "Earnings Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.596", + "raw_stderr": "2.878", + "latency": "289.351", + "cost_per_test": "0.286337", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.878, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:financial_modeling:zai/glm-5.2:score", + "evaluation_name": "vals_ai.fabv2.financial_modeling", + "source_data": { + "dataset_name": "Finance Agent (v2) - Financial Modeling", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "financial_modeling", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Financial Modeling).", + "metric_id": "vals_ai.fabv2.financial_modeling.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 12.885, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "financial_modeling", + "task_name": "Financial Modeling", + "dataset_type": "private", + "industry": "finance", + "raw_score": "12.885", + "raw_stderr": "1.787", + "latency": "777.094", + "cost_per_test": "0.926811", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.787, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:general_qualitative_analysis:zai/glm-5.2:score", + "evaluation_name": "vals_ai.fabv2.general_qualitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Qualitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_qualitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Qualitative Analysis).", + "metric_id": "vals_ai.fabv2.general_qualitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 68.73, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_qualitative_analysis", + "task_name": "General Qualitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.73", + "raw_stderr": "3.079", + "latency": "390.229", + "cost_per_test": "0.64217", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.079, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:general_quantitative_analysis:zai/glm-5.2:score", + "evaluation_name": "vals_ai.fabv2.general_quantitative_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - General Quantitative Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "general_quantitative_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (General Quantitative Analysis).", + "metric_id": "vals_ai.fabv2.general_quantitative_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 65.524, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "general_quantitative_analysis", + "task_name": "General Quantitative Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "65.524", + "raw_stderr": "0.325", + "latency": "318.272", + "cost_per_test": "0.625755", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.325, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:market_analysis:zai/glm-5.2:score", + "evaluation_name": "vals_ai.fabv2.market_analysis", + "source_data": { + "dataset_name": "Finance Agent (v2) - Market Analysis", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "market_analysis", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Market Analysis).", + "metric_id": "vals_ai.fabv2.market_analysis.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 61.501, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "market_analysis", + "task_name": "Market Analysis", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.501", + "raw_stderr": "1.829", + "latency": "405.12", + "cost_per_test": "0.512704", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.829, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:overall:zai/glm-5.2:score", + "evaluation_name": "vals_ai.fabv2.overall", + "source_data": { + "dataset_name": "Finance Agent (v2) - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Overall).", + "metric_id": "vals_ai.fabv2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 49.699, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "49.699", + "raw_stderr": "0.878", + "latency": "473.158", + "cost_per_test": "0.713981", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.878, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "fabv2:precedents:zai/glm-5.2:score", + "evaluation_name": "vals_ai.fabv2.precedents", + "source_data": { + "dataset_name": "Finance Agent (v2) - Precedents", + "source_type": "other", + "additional_details": { + "benchmark_slug": "fabv2", + "task_key": "precedents", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Finance Agent (v2) (Precedents).", + "metric_id": "vals_ai.fabv2.precedents.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/fabv2" + } + }, + "score_details": { + "score": 19.379, + "details": { + "benchmark_slug": "fabv2", + "benchmark_name": "Finance Agent (v2)", + "benchmark_updated": "2026-07-09", + "task_key": "precedents", + "task_name": "Precedents", + "dataset_type": "private", + "industry": "finance", + "raw_score": "19.379", + "raw_stderr": "1.262", + "latency": "724.685", + "cost_per_test": "1.122176", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.262, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.2/709ad24d-0e63-4c6a-b297-92a17492e36c.json b/data/vals_ai/zai/glm-5.2/709ad24d-0e63-4c6a-b297-92a17492e36c.json new file mode 100644 index 0000000000000000000000000000000000000000..839fc10f58f6d1bf63b1c09d14c397feea1debcf --- /dev/null +++ b/data/vals_ai/zai/glm-5.2/709ad24d-0e63-4c6a-b297-92a17492e36c.json @@ -0,0 +1,320 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/terminal-bench-2-1/zai_glm-5.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Terminal-Bench 2.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.2", + "id": "zai/glm-5.2", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.2", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "terminal-bench-2-1:easy:zai/glm-5.2:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.easy", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Easy).", + "metric_id": "vals_ai.terminal-bench-2-1.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 100.0, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "100", + "raw_stderr": "0", + "latency": "18256.253", + "cost_per_test": "9.460336", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "48000", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 48000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:hard:zai/glm-5.2:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.hard", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Hard).", + "metric_id": "vals_ai.terminal-bench-2-1.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 51.111, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "51.111", + "raw_stderr": "2.222", + "latency": "2434.167", + "cost_per_test": "1.261378", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "48000", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.222, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 48000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:medium:zai/glm-5.2:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.medium", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Medium).", + "metric_id": "vals_ai.terminal-bench-2-1.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 74.545, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "74.545", + "raw_stderr": "1.05", + "latency": "1327.727", + "cost_per_test": "0.688024", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "48000", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.05, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 48000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "terminal-bench-2-1:overall:zai/glm-5.2:score", + "evaluation_name": "vals_ai.terminal-bench-2-1.overall", + "source_data": { + "dataset_name": "Terminal-Bench 2.1 - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/terminal-bench-2-1" + ], + "additional_details": { + "benchmark_slug": "terminal-bench-2-1", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Terminal-Bench 2.1 (Overall).", + "metric_id": "vals_ai.terminal-bench-2-1.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/terminal-bench-2-1" + } + }, + "score_details": { + "score": 67.79, + "details": { + "benchmark_slug": "terminal-bench-2-1", + "benchmark_name": "Terminal-Bench 2.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "67.79", + "raw_stderr": "0.991", + "latency": "820.506", + "cost_per_test": "0.425184", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "48000", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.991, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 48000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.2/7cd9a7d6-19e5-44dd-8cc3-2ca993a82ddd.json b/data/vals_ai/zai/glm-5.2/7cd9a7d6-19e5-44dd-8cc3-2ca993a82ddd.json new file mode 100644 index 0000000000000000000000000000000000000000..7ff31da8aba34d7a0101a159ad412e6dee1282fe --- /dev/null +++ b/data/vals_ai/zai/glm-5.2/7cd9a7d6-19e5-44dd-8cc3-2ca993a82ddd.json @@ -0,0 +1,320 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/lcb/zai_glm-5.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LiveCodeBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.2", + "id": "zai/glm-5.2", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.2", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "lcb:easy:zai/glm-5.2:score", + "evaluation_name": "vals_ai.lcb.easy", + "source_data": { + "dataset_name": "LiveCodeBench - Easy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "easy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).", + "metric_id": "vals_ai.lcb.easy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 97.205, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "easy", + "task_name": "Easy", + "dataset_type": "public", + "industry": "coding", + "raw_score": "97.205", + "raw_stderr": "0.919", + "latency": "57.872", + "cost_per_test": "0.014984", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.919, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "lcb:hard:zai/glm-5.2:score", + "evaluation_name": "vals_ai.lcb.hard", + "source_data": { + "dataset_name": "LiveCodeBench - Hard", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "hard", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).", + "metric_id": "vals_ai.lcb.hard.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 37.143, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "hard", + "task_name": "Hard", + "dataset_type": "public", + "industry": "coding", + "raw_score": "37.143", + "raw_stderr": "2.583", + "latency": "1020.063", + "cost_per_test": "0.239479", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.583, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "lcb:medium:zai/glm-5.2:score", + "evaluation_name": "vals_ai.lcb.medium", + "source_data": { + "dataset_name": "LiveCodeBench - Medium", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "medium", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).", + "metric_id": "vals_ai.lcb.medium.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 74.151, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "medium", + "task_name": "Medium", + "dataset_type": "public", + "industry": "coding", + "raw_score": "74.151", + "raw_stderr": "2.237", + "latency": "518.732", + "cost_per_test": "0.133461", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.237, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "lcb:overall:zai/glm-5.2:score", + "evaluation_name": "vals_ai.lcb.overall", + "source_data": { + "dataset_name": "LiveCodeBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/lcb" + ], + "additional_details": { + "benchmark_slug": "lcb", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).", + "metric_id": "vals_ai.lcb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb" + } + }, + "score_details": { + "score": 69.5, + "details": { + "benchmark_slug": "lcb", + "benchmark_name": "LiveCodeBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "coding", + "raw_score": "69.5", + "raw_stderr": "1.179", + "latency": "532.222", + "cost_per_test": "0.129308", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.179, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.2/7e3f8dec-1b44-4ca5-9b29-b46d18c611b5.json b/data/vals_ai/zai/glm-5.2/7e3f8dec-1b44-4ca5-9b29-b46d18c611b5.json new file mode 100644 index 0000000000000000000000000000000000000000..93836acc79f2aafca4f847d8077bae12c026d56c --- /dev/null +++ b/data/vals_ai/zai/glm-5.2/7e3f8dec-1b44-4ca5-9b29-b46d18c611b5.json @@ -0,0 +1,249 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/gpqa/zai_glm-5.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - GPQA Diamond", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.2", + "id": "zai/glm-5.2", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.2", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "gpqa:diamond_few_shot_cot:zai/glm-5.2:score", + "evaluation_name": "vals_ai.gpqa.diamond_few_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Few-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_few_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Few-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_few_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 83.333, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_few_shot_cot", + "task_name": "Few-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.333", + "raw_stderr": "2.649", + "latency": "178.184", + "cost_per_test": "0.049779", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.649, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "gpqa:diamond_zero_shot_cot:zai/glm-5.2:score", + "evaluation_name": "vals_ai.gpqa.diamond_zero_shot_cot", + "source_data": { + "dataset_name": "GPQA Diamond - Zero-Shot CoT", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "diamond_zero_shot_cot", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Zero-Shot CoT).", + "metric_id": "vals_ai.gpqa.diamond_zero_shot_cot.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 87.879, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "diamond_zero_shot_cot", + "task_name": "Zero-Shot CoT", + "dataset_type": "public", + "industry": "academic", + "raw_score": "87.879", + "raw_stderr": "2.36", + "latency": "271.144", + "cost_per_test": "0.06444", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.36, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "gpqa:overall:zai/glm-5.2:score", + "evaluation_name": "vals_ai.gpqa.overall", + "source_data": { + "dataset_name": "GPQA Diamond - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/gpqa" + ], + "additional_details": { + "benchmark_slug": "gpqa", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for GPQA Diamond (Overall).", + "metric_id": "vals_ai.gpqa.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/gpqa" + } + }, + "score_details": { + "score": 85.606, + "details": { + "benchmark_slug": "gpqa", + "benchmark_name": "GPQA Diamond", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "85.606", + "raw_stderr": "1.774", + "latency": "224.664", + "cost_per_test": "0.05711", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.774, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.2/7e6ae2de-a966-4d8b-9cec-c629606d0ca1.json b/data/vals_ai/zai/glm-5.2/7e6ae2de-a966-4d8b-9cec-c629606d0ca1.json new file mode 100644 index 0000000000000000000000000000000000000000..b6deb077ae2188924b93d21143c8d19a6f21e156 --- /dev/null +++ b/data/vals_ai/zai/glm-5.2/7e6ae2de-a966-4d8b-9cec-c629606d0ca1.json @@ -0,0 +1,308 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/corp_fin_v2/zai_glm-5.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - CorpFin v2", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.2", + "id": "zai/glm-5.2", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.2", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "corp_fin_v2:exact_pages:zai/glm-5.2:score", + "evaluation_name": "vals_ai.corp_fin_v2.exact_pages", + "source_data": { + "dataset_name": "CorpFin v2 - Exact Pages", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "exact_pages", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Exact Pages).", + "metric_id": "vals_ai.corp_fin_v2.exact_pages.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 63.636, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "exact_pages", + "task_name": "Exact Pages", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.636", + "raw_stderr": "1.642", + "latency": "12.744", + "cost_per_test": "0.005283", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.642, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:max_fitting_context:zai/glm-5.2:score", + "evaluation_name": "vals_ai.corp_fin_v2.max_fitting_context", + "source_data": { + "dataset_name": "CorpFin v2 - Max Fitting Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "max_fitting_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Max Fitting Context).", + "metric_id": "vals_ai.corp_fin_v2.max_fitting_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 68.415, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "max_fitting_context", + "task_name": "Max Fitting Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "68.415", + "raw_stderr": "1.587", + "latency": "34.665", + "cost_per_test": "0.262089", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.587, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:overall:zai/glm-5.2:score", + "evaluation_name": "vals_ai.corp_fin_v2.overall", + "source_data": { + "dataset_name": "CorpFin v2 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Overall).", + "metric_id": "vals_ai.corp_fin_v2.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 66.123, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.123", + "raw_stderr": "0.932", + "latency": "24.586", + "cost_per_test": "0.121857", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.932, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "corp_fin_v2:shared_max_context:zai/glm-5.2:score", + "evaluation_name": "vals_ai.corp_fin_v2.shared_max_context", + "source_data": { + "dataset_name": "CorpFin v2 - Shared Max Context", + "source_type": "other", + "additional_details": { + "benchmark_slug": "corp_fin_v2", + "task_key": "shared_max_context", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for CorpFin v2 (Shared Max Context).", + "metric_id": "vals_ai.corp_fin_v2.shared_max_context.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/corp_fin_v2" + } + }, + "score_details": { + "score": 66.317, + "details": { + "benchmark_slug": "corp_fin_v2", + "benchmark_name": "CorpFin v2", + "benchmark_updated": "2026-07-09", + "task_key": "shared_max_context", + "task_name": "Shared Max Context", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.317", + "raw_stderr": "1.614", + "latency": "26.348", + "cost_per_test": "0.0982", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.614, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.2/7eb8abc8-c96e-4907-a41d-34275035cc92.json b/data/vals_ai/zai/glm-5.2/7eb8abc8-c96e-4907-a41d-34275035cc92.json new file mode 100644 index 0000000000000000000000000000000000000000..06bd36a4ece743b435d93605bfef20dadd4650bc --- /dev/null +++ b/data/vals_ai/zai/glm-5.2/7eb8abc8-c96e-4907-a41d-34275035cc92.json @@ -0,0 +1,3586 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/hlab/zai_glm-5.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Harvey's Legal Agent Benchmark", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.2", + "id": "zai/glm-5.2", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.2", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "hlab:antitrust_competition:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1101.626", + "cost_per_test": "2.027261", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:arbitration_international_dispute_resolution:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1346.451", + "cost_per_test": "2.719674", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:banking_finance:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1264.311", + "cost_per_test": "2.104294", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:bankruptcy_restructuring:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1531.126", + "cost_per_test": "2.87069", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:capital_markets:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1055.512", + "cost_per_test": "1.763195", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:corporate_governance:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1380.377", + "cost_per_test": "2.394515", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:corporate_ma:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "1002.311", + "cost_per_test": "1.650999", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Criteria Pass Rate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Criteria Pass Rate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 85.598, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate", + "task_name": "Criteria Pass Rate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.598", + "raw_stderr": "0.422", + "latency": "1250.96", + "cost_per_test": "2.057915", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.422, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__antitrust_competition:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__antitrust_competition", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Antitrust Competition", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__antitrust_competition", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Antitrust Competition).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__antitrust_competition.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 87.319, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__antitrust_competition", + "task_name": "Antitrust Competition", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.319", + "raw_stderr": "2.007", + "latency": "1101.626", + "cost_per_test": "2.027261", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.007, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__arbitration_international_dispute_resolution:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Arbitration International Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Arbitration International Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__arbitration_international_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 91.756, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__arbitration_international_dispute_resolution", + "task_name": "Arbitration International Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "91.756", + "raw_stderr": "1.65", + "latency": "1346.451", + "cost_per_test": "2.719674", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.65, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__banking_finance:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__banking_finance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Banking Finance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__banking_finance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Banking Finance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__banking_finance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 85.874, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__banking_finance", + "task_name": "Banking Finance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.874", + "raw_stderr": "2.128", + "latency": "1264.311", + "cost_per_test": "2.104294", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.128, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__bankruptcy_restructuring:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Bankruptcy Restructuring", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Bankruptcy Restructuring).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__bankruptcy_restructuring.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 74.394, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__bankruptcy_restructuring", + "task_name": "Bankruptcy Restructuring", + "dataset_type": "public", + "industry": "legal", + "raw_score": "74.394", + "raw_stderr": "2.572", + "latency": "1531.126", + "cost_per_test": "2.87069", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.572, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__capital_markets:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__capital_markets", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Capital Markets", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__capital_markets", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Capital Markets).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__capital_markets.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 87.059, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__capital_markets", + "task_name": "Capital Markets", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.059", + "raw_stderr": "2.106", + "latency": "1055.512", + "cost_per_test": "1.763195", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.106, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_governance:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_governance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Governance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_governance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Governance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_governance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 93.151, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_governance", + "task_name": "Corporate Governance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "93.151", + "raw_stderr": "1.481", + "latency": "1380.377", + "cost_per_test": "2.394515", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.481, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__corporate_ma:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__corporate_ma", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Corporate Ma", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__corporate_ma", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Corporate Ma).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__corporate_ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 78.846, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__corporate_ma", + "task_name": "Corporate Ma", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.846", + "raw_stderr": "2.316", + "latency": "1002.311", + "cost_per_test": "1.650999", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.316, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__data_privacy_cybersecurity:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 95.053, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "95.053", + "raw_stderr": "1.291", + "latency": "1354.476", + "cost_per_test": "2.842591", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.291, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__emerging_companies_venture_capital:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 87.372, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.372", + "raw_stderr": "1.944", + "latency": "1326.424", + "cost_per_test": "2.290803", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.944, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__employment_labor:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 81.273, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.273", + "raw_stderr": "2.392", + "latency": "1191.52", + "cost_per_test": "1.421054", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.392, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__energy_natural_resources:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 79.208, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "79.208", + "raw_stderr": "2.335", + "latency": "853.216", + "cost_per_test": "2.004243", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.335, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__environmental_esg:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 87.814, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.814", + "raw_stderr": "1.962", + "latency": "1423.765", + "cost_per_test": "1.886054", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.962, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__funds_asset_management:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 88.338, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.338", + "raw_stderr": "1.736", + "latency": "1389.994", + "cost_per_test": "2.133545", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.736, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__healthcare_life_sciences:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 91.419, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "91.419", + "raw_stderr": "1.612", + "latency": "1360.684", + "cost_per_test": "2.190988", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.612, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__immigration:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 88.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88", + "raw_stderr": "1.963", + "latency": "1326.439", + "cost_per_test": "1.8212", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.963, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__insurance:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 87.18, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "87.18", + "raw_stderr": "2.027", + "latency": "1287.204", + "cost_per_test": "1.917632", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.027, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__intellectual_property:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 76.325, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "76.325", + "raw_stderr": "2.531", + "latency": "1072.968", + "cost_per_test": "1.749857", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.531, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__international_trade_sanctions:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 71.032, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "71.032", + "raw_stderr": "2.863", + "latency": "1147.416", + "cost_per_test": "1.320403", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.863, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__litigation_dispute_resolution:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 86.517, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "86.517", + "raw_stderr": "2.094", + "latency": "1443.4", + "cost_per_test": "2.117782", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.094, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__real_estate:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 91.837, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "91.837", + "raw_stderr": "1.481", + "latency": "1153.206", + "cost_per_test": "1.899249", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.481, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__structured_finance_securitization:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 83.492, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "83.492", + "raw_stderr": "2.095", + "latency": "1144.86", + "cost_per_test": "1.925613", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.095, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__tax:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 85.619, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "85.619", + "raw_stderr": "2.033", + "latency": "1637.675", + "cost_per_test": "2.610141", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.033, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__trusts_estates_private_client:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 88.55, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.55", + "raw_stderr": "1.971", + "latency": "1109.499", + "cost_per_test": "2.162434", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.971, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:criteria_pass_rate__white_collar_defense_investigations:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.criteria_pass_rate__white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 84.848, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "criteria_pass_rate__white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.848", + "raw_stderr": "2.084", + "latency": "1118.586", + "cost_per_test": "1.565734", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.084, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:data_privacy_cybersecurity:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.data_privacy_cybersecurity", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Data Privacy Cybersecurity", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "data_privacy_cybersecurity", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Data Privacy Cybersecurity).", + "metric_id": "vals_ai.hlab.data_privacy_cybersecurity.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "data_privacy_cybersecurity", + "task_name": "Data Privacy Cybersecurity", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "1354.476", + "cost_per_test": "2.842591", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:emerging_companies_venture_capital:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.emerging_companies_venture_capital", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Emerging Companies Venture Capital", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "emerging_companies_venture_capital", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Emerging Companies Venture Capital).", + "metric_id": "vals_ai.hlab.emerging_companies_venture_capital.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "emerging_companies_venture_capital", + "task_name": "Emerging Companies Venture Capital", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1326.424", + "cost_per_test": "2.290803", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:employment_labor:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.employment_labor", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Employment Labor", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "employment_labor", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Employment Labor).", + "metric_id": "vals_ai.hlab.employment_labor.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "employment_labor", + "task_name": "Employment Labor", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1191.52", + "cost_per_test": "1.421054", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:energy_natural_resources:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.energy_natural_resources", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Energy Natural Resources", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "energy_natural_resources", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Energy Natural Resources).", + "metric_id": "vals_ai.hlab.energy_natural_resources.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "energy_natural_resources", + "task_name": "Energy Natural Resources", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "853.216", + "cost_per_test": "2.004243", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:environmental_esg:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.environmental_esg", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Environmental Esg", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "environmental_esg", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Environmental Esg).", + "metric_id": "vals_ai.hlab.environmental_esg.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "environmental_esg", + "task_name": "Environmental Esg", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1423.765", + "cost_per_test": "1.886054", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:funds_asset_management:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.funds_asset_management", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Funds Asset Management", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "funds_asset_management", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Funds Asset Management).", + "metric_id": "vals_ai.hlab.funds_asset_management.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "funds_asset_management", + "task_name": "Funds Asset Management", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1389.994", + "cost_per_test": "2.133545", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:healthcare_life_sciences:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.healthcare_life_sciences", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Healthcare Life Sciences", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "healthcare_life_sciences", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Healthcare Life Sciences).", + "metric_id": "vals_ai.hlab.healthcare_life_sciences.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "healthcare_life_sciences", + "task_name": "Healthcare Life Sciences", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "1360.684", + "cost_per_test": "2.190988", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:immigration:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.immigration", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Immigration", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "immigration", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Immigration).", + "metric_id": "vals_ai.hlab.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1326.439", + "cost_per_test": "1.8212", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:insurance:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.insurance", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Insurance", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "insurance", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Insurance).", + "metric_id": "vals_ai.hlab.insurance.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "insurance", + "task_name": "Insurance", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1287.204", + "cost_per_test": "1.917632", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:intellectual_property:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.intellectual_property", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Intellectual Property", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "intellectual_property", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Intellectual Property).", + "metric_id": "vals_ai.hlab.intellectual_property.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "intellectual_property", + "task_name": "Intellectual Property", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1072.968", + "cost_per_test": "1.749857", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:international_trade_sanctions:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.international_trade_sanctions", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - International Trade Sanctions", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "international_trade_sanctions", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (International Trade Sanctions).", + "metric_id": "vals_ai.hlab.international_trade_sanctions.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "international_trade_sanctions", + "task_name": "International Trade Sanctions", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1147.416", + "cost_per_test": "1.320403", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:litigation_dispute_resolution:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.litigation_dispute_resolution", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Litigation Dispute Resolution", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "litigation_dispute_resolution", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Litigation Dispute Resolution).", + "metric_id": "vals_ai.hlab.litigation_dispute_resolution.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "litigation_dispute_resolution", + "task_name": "Litigation Dispute Resolution", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1443.4", + "cost_per_test": "2.117782", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:overall:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.overall", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Overall).", + "metric_id": "vals_ai.hlab.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 7.083, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "7.083", + "raw_stderr": "1.998", + "latency": "1250.96", + "cost_per_test": "2.057915", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.998, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:real_estate:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.real_estate", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Real Estate", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "real_estate", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Real Estate).", + "metric_id": "vals_ai.hlab.real_estate.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "real_estate", + "task_name": "Real Estate", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "1153.206", + "cost_per_test": "1.899249", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:structured_finance_securitization:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.structured_finance_securitization", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Structured Finance Securitization", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "structured_finance_securitization", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Structured Finance Securitization).", + "metric_id": "vals_ai.hlab.structured_finance_securitization.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 20.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "structured_finance_securitization", + "task_name": "Structured Finance Securitization", + "dataset_type": "public", + "industry": "legal", + "raw_score": "20", + "raw_stderr": "20", + "latency": "1144.86", + "cost_per_test": "1.925613", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 20.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:tax:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.tax", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Tax", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "tax", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Tax).", + "metric_id": "vals_ai.hlab.tax.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "tax", + "task_name": "Tax", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1637.675", + "cost_per_test": "2.610141", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:trusts_estates_private_client:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.trusts_estates_private_client", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - Trusts Estates Private Client", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "trusts_estates_private_client", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (Trusts Estates Private Client).", + "metric_id": "vals_ai.hlab.trusts_estates_private_client.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "trusts_estates_private_client", + "task_name": "Trusts Estates Private Client", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1109.499", + "cost_per_test": "2.162434", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "hlab:white_collar_defense_investigations:zai/glm-5.2:score", + "evaluation_name": "vals_ai.hlab.white_collar_defense_investigations", + "source_data": { + "dataset_name": "Harvey's Legal Agent Benchmark - White Collar Defense Investigations", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/hlab" + ], + "additional_details": { + "benchmark_slug": "hlab", + "task_key": "white_collar_defense_investigations", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Harvey's Legal Agent Benchmark (White Collar Defense Investigations).", + "metric_id": "vals_ai.hlab.white_collar_defense_investigations.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/hlab" + } + }, + "score_details": { + "score": 0.0, + "details": { + "benchmark_slug": "hlab", + "benchmark_name": "Harvey's Legal Agent Benchmark", + "benchmark_updated": "2026-07-09", + "task_key": "white_collar_defense_investigations", + "task_name": "White Collar Defense Investigations", + "dataset_type": "public", + "industry": "legal", + "raw_score": "0", + "raw_stderr": "0", + "latency": "1118.586", + "cost_per_test": "1.565734", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.0, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.2/8f2fd8b2-4c34-4e29-95e8-45bbe2fd5098.json b/data/vals_ai/zai/glm-5.2/8f2fd8b2-4c34-4e29-95e8-45bbe2fd5098.json new file mode 100644 index 0000000000000000000000000000000000000000..a605d969050701fd2d6f20bfbcf86e31b2a1422c --- /dev/null +++ b/data/vals_ai/zai/glm-5.2/8f2fd8b2-4c34-4e29-95e8-45bbe2fd5098.json @@ -0,0 +1,104 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vibe-code/zai_glm-5.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vibe Code Bench v1.1", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.2", + "id": "zai/glm-5.2", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.2", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vibe-code:overall:zai/glm-5.2:score", + "evaluation_name": "vals_ai.vibe-code.overall", + "source_data": { + "dataset_name": "Vibe Code Bench v1.1 - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vibe-code", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vibe Code Bench v1.1 (Overall).", + "metric_id": "vals_ai.vibe-code.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vibe-code" + } + }, + "score_details": { + "score": 63.961, + "details": { + "benchmark_slug": "vibe-code", + "benchmark_name": "Vibe Code Bench v1.1", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "63.961", + "raw_stderr": "4.795", + "latency": "3748.167", + "cost_per_test": "8.488514", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 4.795, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.2/94387f52-3333-40e4-8598-afb23cafe4da.json b/data/vals_ai/zai/glm-5.2/94387f52-3333-40e4-8598-afb23cafe4da.json new file mode 100644 index 0000000000000000000000000000000000000000..f888b76e87f1a87784b11a7f7e581abb276f5874 --- /dev/null +++ b/data/vals_ai/zai/glm-5.2/94387f52-3333-40e4-8598-afb23cafe4da.json @@ -0,0 +1,462 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_bench/zai_glm-5.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - LegalBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.2", + "id": "zai/glm-5.2", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.2", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_bench:conclusion_tasks:zai/glm-5.2:score", + "evaluation_name": "vals_ai.legal_bench.conclusion_tasks", + "source_data": { + "dataset_name": "LegalBench - Conclusion Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "conclusion_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Conclusion Tasks).", + "metric_id": "vals_ai.legal_bench.conclusion_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 88.587, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "conclusion_tasks", + "task_name": "Conclusion Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "88.587", + "raw_stderr": "0.735", + "latency": "4.646", + "cost_per_test": "0.000935", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.735, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:interpretation_tasks:zai/glm-5.2:score", + "evaluation_name": "vals_ai.legal_bench.interpretation_tasks", + "source_data": { + "dataset_name": "LegalBench - Interpretation Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "interpretation_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Interpretation Tasks).", + "metric_id": "vals_ai.legal_bench.interpretation_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 82.17, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "interpretation_tasks", + "task_name": "Interpretation Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "82.17", + "raw_stderr": "0.194", + "latency": "3.657", + "cost_per_test": "0.001164", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.194, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:issue_tasks:zai/glm-5.2:score", + "evaluation_name": "vals_ai.legal_bench.issue_tasks", + "source_data": { + "dataset_name": "LegalBench - Issue Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "issue_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Issue Tasks).", + "metric_id": "vals_ai.legal_bench.issue_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 89.506, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "issue_tasks", + "task_name": "Issue Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "89.506", + "raw_stderr": "0.575", + "latency": "2.862", + "cost_per_test": "0.001761", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.575, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:overall:zai/glm-5.2:score", + "evaluation_name": "vals_ai.legal_bench.overall", + "source_data": { + "dataset_name": "LegalBench - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Overall).", + "metric_id": "vals_ai.legal_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 84.073, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "legal", + "raw_score": "84.073", + "raw_stderr": "0.452", + "latency": "5.766", + "cost_per_test": "0.001596", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.452, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:rhetoric_tasks:zai/glm-5.2:score", + "evaluation_name": "vals_ai.legal_bench.rhetoric_tasks", + "source_data": { + "dataset_name": "LegalBench - Rhetoric Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rhetoric_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rhetoric Tasks).", + "metric_id": "vals_ai.legal_bench.rhetoric_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 81.263, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rhetoric_tasks", + "task_name": "Rhetoric Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "81.263", + "raw_stderr": "0.99", + "latency": "3.041", + "cost_per_test": "0.001005", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.99, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_bench:rule_tasks:zai/glm-5.2:score", + "evaluation_name": "vals_ai.legal_bench.rule_tasks", + "source_data": { + "dataset_name": "LegalBench - Rule Tasks", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/legal_bench" + ], + "additional_details": { + "benchmark_slug": "legal_bench", + "task_key": "rule_tasks", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for LegalBench (Rule Tasks).", + "metric_id": "vals_ai.legal_bench.rule_tasks.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_bench" + } + }, + "score_details": { + "score": 78.84, + "details": { + "benchmark_slug": "legal_bench", + "benchmark_name": "LegalBench", + "benchmark_updated": "2026-07-09", + "task_key": "rule_tasks", + "task_name": "Rule Tasks", + "dataset_type": "public", + "industry": "legal", + "raw_score": "78.84", + "raw_stderr": "1.795", + "latency": "14.626", + "cost_per_test": "0.003114", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.795, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.2/ba0a141f-7a40-49bb-8b4a-733037e3c378.json b/data/vals_ai/zai/glm-5.2/ba0a141f-7a40-49bb-8b4a-733037e3c378.json new file mode 100644 index 0000000000000000000000000000000000000000..d67c1eeb7cc48833beabe235d1ebe420b40fc4a8 --- /dev/null +++ b/data/vals_ai/zai/glm-5.2/ba0a141f-7a40-49bb-8b4a-733037e3c378.json @@ -0,0 +1,716 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/emb/zai_glm-5.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - EMB", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "finance", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.2", + "id": "zai/glm-5.2", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.2", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "emb:comps:zai/glm-5.2:score", + "evaluation_name": "vals_ai.emb.comps", + "source_data": { + "dataset_name": "EMB - Comps", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "comps", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Comps).", + "metric_id": "vals_ai.emb.comps.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 54.922, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "comps", + "task_name": "Comps", + "dataset_type": "private", + "industry": "finance", + "raw_score": "54.922", + "raw_stderr": "7.703", + "latency": "3252.76", + "cost_per_test": "4.832217", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 7.703, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:dataroom:zai/glm-5.2:score", + "evaluation_name": "vals_ai.emb.dataroom", + "source_data": { + "dataset_name": "EMB - Dataroom", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dataroom", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Dataroom).", + "metric_id": "vals_ai.emb.dataroom.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 83.444, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dataroom", + "task_name": "Dataroom", + "dataset_type": "private", + "industry": "finance", + "raw_score": "83.444", + "raw_stderr": "3.94", + "latency": "2702.431", + "cost_per_test": "3.41684", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.94, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:dcf:zai/glm-5.2:score", + "evaluation_name": "vals_ai.emb.dcf", + "source_data": { + "dataset_name": "EMB - DCF", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "dcf", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (DCF).", + "metric_id": "vals_ai.emb.dcf.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 52.968, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "dcf", + "task_name": "DCF", + "dataset_type": "private", + "industry": "finance", + "raw_score": "52.968", + "raw_stderr": "8.572", + "latency": "3370.007", + "cost_per_test": "4.152392", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 8.572, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:lbo:zai/glm-5.2:score", + "evaluation_name": "vals_ai.emb.lbo", + "source_data": { + "dataset_name": "EMB - LBO", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "lbo", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (LBO).", + "metric_id": "vals_ai.emb.lbo.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 51.815, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "lbo", + "task_name": "LBO", + "dataset_type": "private", + "industry": "finance", + "raw_score": "51.815", + "raw_stderr": "5.52", + "latency": "4821.481", + "cost_per_test": "6.512343", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 5.52, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:ma:zai/glm-5.2:score", + "evaluation_name": "vals_ai.emb.ma", + "source_data": { + "dataset_name": "EMB - M&A", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "ma", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (M&A).", + "metric_id": "vals_ai.emb.ma.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 63.913, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "ma", + "task_name": "M&A", + "dataset_type": "private", + "industry": "finance", + "raw_score": "63.913", + "raw_stderr": "8.842", + "latency": "3655.024", + "cost_per_test": "4.390075", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 8.842, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:misc:zai/glm-5.2:score", + "evaluation_name": "vals_ai.emb.misc", + "source_data": { + "dataset_name": "EMB - Misc", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "misc", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Misc).", + "metric_id": "vals_ai.emb.misc.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 60.775, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "misc", + "task_name": "Misc", + "dataset_type": "private", + "industry": "finance", + "raw_score": "60.775", + "raw_stderr": "6.49", + "latency": "3601.77", + "cost_per_test": "4.199757", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 6.49, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:operating:zai/glm-5.2:score", + "evaluation_name": "vals_ai.emb.operating", + "source_data": { + "dataset_name": "EMB - Operating", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "operating", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Operating).", + "metric_id": "vals_ai.emb.operating.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 62.888, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "operating", + "task_name": "Operating", + "dataset_type": "private", + "industry": "finance", + "raw_score": "62.888", + "raw_stderr": "6.441", + "latency": "3231.995", + "cost_per_test": "3.50696", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 6.441, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:overall:zai/glm-5.2:score", + "evaluation_name": "vals_ai.emb.overall", + "source_data": { + "dataset_name": "EMB - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Overall).", + "metric_id": "vals_ai.emb.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 61.532, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "finance", + "raw_score": "61.532", + "raw_stderr": "2.879", + "latency": "3532.407", + "cost_per_test": "4.465213", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.879, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:scratch:zai/glm-5.2:score", + "evaluation_name": "vals_ai.emb.scratch", + "source_data": { + "dataset_name": "EMB - Scratch", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "scratch", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Scratch).", + "metric_id": "vals_ai.emb.scratch.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 66.693, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "scratch", + "task_name": "Scratch", + "dataset_type": "private", + "industry": "finance", + "raw_score": "66.693", + "raw_stderr": "2.545", + "latency": "3508.741", + "cost_per_test": "4.028214", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.545, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "emb:template:zai/glm-5.2:score", + "evaluation_name": "vals_ai.emb.template", + "source_data": { + "dataset_name": "EMB - Template", + "source_type": "other", + "additional_details": { + "benchmark_slug": "emb", + "task_key": "template", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for EMB (Template).", + "metric_id": "vals_ai.emb.template.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/emb" + } + }, + "score_details": { + "score": 55.814, + "details": { + "benchmark_slug": "emb", + "benchmark_name": "EMB", + "benchmark_updated": "2026-07-09", + "task_key": "template", + "task_name": "Template", + "dataset_type": "private", + "industry": "finance", + "raw_score": "55.814", + "raw_stderr": "4.771", + "latency": "3556.073", + "cost_per_test": "4.902211", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 4.771, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.2/c25f6cc9-1f9f-42ce-a7fd-9769fc2855f4.json b/data/vals_ai/zai/glm-5.2/c25f6cc9-1f9f-42ce-a7fd-9769fc2855f4.json new file mode 100644 index 0000000000000000000000000000000000000000..39aea19f94052530819e5261d18c6af3674cf97c --- /dev/null +++ b/data/vals_ai/zai/glm-5.2/c25f6cc9-1f9f-42ce-a7fd-9769fc2855f4.json @@ -0,0 +1,308 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/code-migration/zai_glm-5.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Code Migration", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "dataset_type": "private", + "industry": "coding", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.2", + "id": "zai/glm-5.2", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.2", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "code-migration:cli:zai/glm-5.2:score", + "evaluation_name": "vals_ai.code-migration.cli", + "source_data": { + "dataset_name": "Code Migration - CLI", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cli", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (CLI).", + "metric_id": "vals_ai.code-migration.cli.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 27.16, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cli", + "task_name": "CLI", + "dataset_type": "private", + "industry": "coding", + "raw_score": "27.16", + "raw_stderr": "2.143", + "latency": "7597.777", + "cost_per_test": "15.070562", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.143, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "code-migration:cobol:zai/glm-5.2:score", + "evaluation_name": "vals_ai.code-migration.cobol", + "source_data": { + "dataset_name": "Code Migration - COBOL", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "cobol", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (COBOL).", + "metric_id": "vals_ai.code-migration.cobol.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 70.0, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "cobol", + "task_name": "COBOL", + "dataset_type": "private", + "industry": "coding", + "raw_score": "70", + "raw_stderr": "15.275", + "latency": "4632.139", + "cost_per_test": "5.331586", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 15.275, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "code-migration:code_quality:zai/glm-5.2:score", + "evaluation_name": "vals_ai.code-migration.code_quality", + "source_data": { + "dataset_name": "Code Migration - Code Quality", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "code_quality", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Code Quality).", + "metric_id": "vals_ai.code-migration.code_quality.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 3.981, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "code_quality", + "task_name": "Code Quality", + "dataset_type": "private", + "industry": "coding", + "raw_score": "3.981", + "raw_stderr": "0.052", + "latency": "6856.367", + "cost_per_test": "12.635818", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.052, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "code-migration:overall:zai/glm-5.2:score", + "evaluation_name": "vals_ai.code-migration.overall", + "source_data": { + "dataset_name": "Code Migration - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "code-migration", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Code Migration (Overall).", + "metric_id": "vals_ai.code-migration.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/code-migration" + } + }, + "score_details": { + "score": 37.87, + "details": { + "benchmark_slug": "code-migration", + "benchmark_name": "Code Migration", + "benchmark_updated": "2026-07-10", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "coding", + "raw_score": "37.87", + "raw_stderr": "4.143", + "latency": "6856.367", + "cost_per_test": "12.635818", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 4.143, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.2/ceafa51f-5260-4105-8342-c3c6a7d76f61.json b/data/vals_ai/zai/glm-5.2/ceafa51f-5260-4105-8342-c3c6a7d76f61.json new file mode 100644 index 0000000000000000000000000000000000000000..dbdf42f2662c4d903a05223e259a3e2e61d2d974 --- /dev/null +++ b/data/vals_ai/zai/glm-5.2/ceafa51f-5260-4105-8342-c3c6a7d76f61.json @@ -0,0 +1,104 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/medscribe/zai_glm-5.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MedScribe", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "healthcare", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.2", + "id": "zai/glm-5.2", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.2", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "medscribe:overall:zai/glm-5.2:score", + "evaluation_name": "vals_ai.medscribe.overall", + "source_data": { + "dataset_name": "MedScribe - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "medscribe", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MedScribe (Overall).", + "metric_id": "vals_ai.medscribe.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/medscribe" + } + }, + "score_details": { + "score": 83.534, + "details": { + "benchmark_slug": "medscribe", + "benchmark_name": "MedScribe", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "healthcare", + "raw_score": "83.534", + "raw_stderr": "2.002", + "latency": "138.372", + "cost_per_test": "0.044912", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "30000", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.002, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 30000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.2/e72f16ae-92ec-4202-9fcf-72d3cabc4c4a.json b/data/vals_ai/zai/glm-5.2/e72f16ae-92ec-4202-9fcf-72d3cabc4c4a.json new file mode 100644 index 0000000000000000000000000000000000000000..a86cb1d43bf2f809d9e8a9a2d77dc6e818f234f7 --- /dev/null +++ b/data/vals_ai/zai/glm-5.2/e72f16ae-92ec-4202-9fcf-72d3cabc4c4a.json @@ -0,0 +1,1260 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/legal_research/zai_glm-5.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Legal Research Bench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "legal", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.2", + "id": "zai/glm-5.2", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.2", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "legal_research:administrative_regulatory:zai/glm-5.2:score", + "evaluation_name": "vals_ai.legal_research.administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 39.394, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "39.394", + "raw_stderr": "6.061", + "latency": "836.031", + "cost_per_test": "0.808378", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 6.061, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:business_commercial:zai/glm-5.2:score", + "evaluation_name": "vals_ai.legal_research.business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 27.907, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "27.907", + "raw_stderr": "6.921", + "latency": "1106.364", + "cost_per_test": "0.920113", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 6.921, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:civil_litigation:zai/glm-5.2:score", + "evaluation_name": "vals_ai.legal_research.civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 25.926, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "25.926", + "raw_stderr": "8.594", + "latency": "885.063", + "cost_per_test": "0.857222", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 8.594, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:constitutional_civil_rights:zai/glm-5.2:score", + "evaluation_name": "vals_ai.legal_research.constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 28.947, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "28.947", + "raw_stderr": "7.456", + "latency": "558.667", + "cost_per_test": "0.395873", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 7.456, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:criminal:zai/glm-5.2:score", + "evaluation_name": "vals_ai.legal_research.criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 22.807, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "22.807", + "raw_stderr": "5.607", + "latency": "1034.653", + "cost_per_test": "0.846458", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 5.607, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:family:zai/glm-5.2:score", + "evaluation_name": "vals_ai.legal_research.family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 22.727, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "22.727", + "raw_stderr": "9.145", + "latency": "1806.747", + "cost_per_test": "1.50639", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 9.145, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:health:zai/glm-5.2:score", + "evaluation_name": "vals_ai.legal_research.health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 50.0, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "50", + "raw_stderr": "9.285", + "latency": "931.988", + "cost_per_test": "0.931196", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 9.285, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:immigration:zai/glm-5.2:score", + "evaluation_name": "vals_ai.legal_research.immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 27.273, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "27.273", + "raw_stderr": "14.084", + "latency": "1192", + "cost_per_test": "1.003117", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 14.084, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:overall:zai/glm-5.2:score", + "evaluation_name": "vals_ai.legal_research.overall", + "source_data": { + "dataset_name": "Legal Research Bench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Overall).", + "metric_id": "vals_ai.legal_research.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 31.25, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "legal", + "raw_score": "31.25", + "raw_stderr": "3.222", + "latency": "1023.468", + "cost_per_test": "0.889209", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.222, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass:zai/glm-5.2:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass", + "source_data": { + "dataset_name": "Legal Research Bench - Weighted Pass", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Weighted Pass).", + "metric_id": "vals_ai.legal_research.weighted_pass.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 82.871, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass", + "task_name": "Weighted Pass", + "dataset_type": "private", + "industry": "legal", + "raw_score": "82.871", + "raw_stderr": "1.478", + "latency": "1023.468", + "cost_per_test": "0.889209", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.478, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__administrative_regulatory:zai/glm-5.2:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__administrative_regulatory", + "source_data": { + "dataset_name": "Legal Research Bench - Administrative / Regulatory", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__administrative_regulatory", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Administrative / Regulatory).", + "metric_id": "vals_ai.legal_research.weighted_pass__administrative_regulatory.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 85.95, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__administrative_regulatory", + "task_name": "Administrative / Regulatory", + "dataset_type": "private", + "industry": "legal", + "raw_score": "85.95", + "raw_stderr": "2.455", + "latency": "836.031", + "cost_per_test": "0.808378", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.455, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__business_commercial:zai/glm-5.2:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__business_commercial", + "source_data": { + "dataset_name": "Legal Research Bench - Business & Commercial", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__business_commercial", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Business & Commercial).", + "metric_id": "vals_ai.legal_research.weighted_pass__business_commercial.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 87.466, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__business_commercial", + "task_name": "Business & Commercial", + "dataset_type": "private", + "industry": "legal", + "raw_score": "87.466", + "raw_stderr": "2.623", + "latency": "1106.364", + "cost_per_test": "0.920113", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.623, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__civil_litigation:zai/glm-5.2:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__civil_litigation", + "source_data": { + "dataset_name": "Legal Research Bench - Civil Litigation", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__civil_litigation", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Civil Litigation).", + "metric_id": "vals_ai.legal_research.weighted_pass__civil_litigation.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 86.955, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__civil_litigation", + "task_name": "Civil Litigation", + "dataset_type": "private", + "industry": "legal", + "raw_score": "86.955", + "raw_stderr": "3.446", + "latency": "885.063", + "cost_per_test": "0.857222", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 3.446, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__constitutional_civil_rights:zai/glm-5.2:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights", + "source_data": { + "dataset_name": "Legal Research Bench - Constitutional / Civil Rights", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__constitutional_civil_rights", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Constitutional / Civil Rights).", + "metric_id": "vals_ai.legal_research.weighted_pass__constitutional_civil_rights.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 86.095, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__constitutional_civil_rights", + "task_name": "Constitutional / Civil Rights", + "dataset_type": "private", + "industry": "legal", + "raw_score": "86.095", + "raw_stderr": "2.502", + "latency": "558.667", + "cost_per_test": "0.395873", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.502, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__criminal:zai/glm-5.2:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__criminal", + "source_data": { + "dataset_name": "Legal Research Bench - Criminal", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__criminal", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Criminal).", + "metric_id": "vals_ai.legal_research.weighted_pass__criminal.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 80.073, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__criminal", + "task_name": "Criminal", + "dataset_type": "private", + "industry": "legal", + "raw_score": "80.073", + "raw_stderr": "2.733", + "latency": "1034.653", + "cost_per_test": "0.846458", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.733, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__family:zai/glm-5.2:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__family", + "source_data": { + "dataset_name": "Legal Research Bench - Family", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__family", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Family).", + "metric_id": "vals_ai.legal_research.weighted_pass__family.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 69.315, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__family", + "task_name": "Family", + "dataset_type": "private", + "industry": "legal", + "raw_score": "69.315", + "raw_stderr": "6.228", + "latency": "1806.747", + "cost_per_test": "1.50639", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 6.228, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__health:zai/glm-5.2:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__health", + "source_data": { + "dataset_name": "Legal Research Bench - Health", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__health", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Health).", + "metric_id": "vals_ai.legal_research.weighted_pass__health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 91.257, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__health", + "task_name": "Health", + "dataset_type": "private", + "industry": "legal", + "raw_score": "91.257", + "raw_stderr": "2.374", + "latency": "931.988", + "cost_per_test": "0.931196", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.374, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "legal_research:weighted_pass__immigration:zai/glm-5.2:score", + "evaluation_name": "vals_ai.legal_research.weighted_pass__immigration", + "source_data": { + "dataset_name": "Legal Research Bench - Immigration", + "source_type": "other", + "additional_details": { + "benchmark_slug": "legal_research", + "task_key": "weighted_pass__immigration", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Legal Research Bench (Immigration).", + "metric_id": "vals_ai.legal_research.weighted_pass__immigration.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/legal_research" + } + }, + "score_details": { + "score": 78.068, + "details": { + "benchmark_slug": "legal_research", + "benchmark_name": "Legal Research Bench", + "benchmark_updated": "2026-07-09", + "task_key": "weighted_pass__immigration", + "task_name": "Immigration", + "dataset_type": "private", + "industry": "legal", + "raw_score": "78.068", + "raw_stderr": "8.926", + "latency": "1192", + "cost_per_test": "1.003117", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 8.926, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.2/ebc25de5-9185-400d-b91f-bd34ebaa12e5.json b/data/vals_ai/zai/glm-5.2/ebc25de5-9185-400d-b91f-bd34ebaa12e5.json new file mode 100644 index 0000000000000000000000000000000000000000..bd62918302694936a56e2094100a35b3fa8ebaaf --- /dev/null +++ b/data/vals_ai/zai/glm-5.2/ebc25de5-9185-400d-b91f-bd34ebaa12e5.json @@ -0,0 +1,1101 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/mmlu_pro/zai_glm-5.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - MMLU Pro", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "dataset_type": "public", + "industry": "academic", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.2", + "id": "zai/glm-5.2", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.2", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "mmlu_pro:biology:zai/glm-5.2:score", + "evaluation_name": "vals_ai.mmlu_pro.biology", + "source_data": { + "dataset_name": "MMLU Pro - Biology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "biology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).", + "metric_id": "vals_ai.mmlu_pro.biology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 94.003, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "biology", + "task_name": "Biology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "94.003", + "raw_stderr": "0.887", + "latency": "34.071", + "cost_per_test": "0.007903", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.887, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:business:zai/glm-5.2:score", + "evaluation_name": "vals_ai.mmlu_pro.business", + "source_data": { + "dataset_name": "MMLU Pro - Business", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "business", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).", + "metric_id": "vals_ai.mmlu_pro.business.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.888, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "business", + "task_name": "Business", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.888", + "raw_stderr": "0.972", + "latency": "47.4", + "cost_per_test": "0.012567", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.972, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:chemistry:zai/glm-5.2:score", + "evaluation_name": "vals_ai.mmlu_pro.chemistry", + "source_data": { + "dataset_name": "MMLU Pro - Chemistry", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "chemistry", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).", + "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 91.608, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "chemistry", + "task_name": "Chemistry", + "dataset_type": "public", + "industry": "academic", + "raw_score": "91.608", + "raw_stderr": "0.824", + "latency": "75.947", + "cost_per_test": "0.020916", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.824, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:computer_science:zai/glm-5.2:score", + "evaluation_name": "vals_ai.mmlu_pro.computer_science", + "source_data": { + "dataset_name": "MMLU Pro - Computer Science", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "computer_science", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).", + "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.244, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "computer_science", + "task_name": "Computer Science", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.244", + "raw_stderr": "1.465", + "latency": "43.156", + "cost_per_test": "0.012207", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.465, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:economics:zai/glm-5.2:score", + "evaluation_name": "vals_ai.mmlu_pro.economics", + "source_data": { + "dataset_name": "MMLU Pro - Economics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "economics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).", + "metric_id": "vals_ai.mmlu_pro.economics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 90.403, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "economics", + "task_name": "Economics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "90.403", + "raw_stderr": "1.014", + "latency": "27", + "cost_per_test": "0.007413", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.014, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:engineering:zai/glm-5.2:score", + "evaluation_name": "vals_ai.mmlu_pro.engineering", + "source_data": { + "dataset_name": "MMLU Pro - Engineering", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "engineering", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).", + "metric_id": "vals_ai.mmlu_pro.engineering.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.385, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "engineering", + "task_name": "Engineering", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.385", + "raw_stderr": "1.196", + "latency": "129.896", + "cost_per_test": "0.037616", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.196, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:health:zai/glm-5.2:score", + "evaluation_name": "vals_ai.mmlu_pro.health", + "source_data": { + "dataset_name": "MMLU Pro - Health", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "health", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).", + "metric_id": "vals_ai.mmlu_pro.health.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 81.418, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "health", + "task_name": "Health", + "dataset_type": "public", + "industry": "academic", + "raw_score": "81.418", + "raw_stderr": "1.36", + "latency": "41.625", + "cost_per_test": "0.010657", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.36, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:history:zai/glm-5.2:score", + "evaluation_name": "vals_ai.mmlu_pro.history", + "source_data": { + "dataset_name": "MMLU Pro - History", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "history", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).", + "metric_id": "vals_ai.mmlu_pro.history.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 76.903, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "history", + "task_name": "History", + "dataset_type": "public", + "industry": "academic", + "raw_score": "76.903", + "raw_stderr": "2.159", + "latency": "38.522", + "cost_per_test": "0.010321", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 2.159, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:law:zai/glm-5.2:score", + "evaluation_name": "vals_ai.mmlu_pro.law", + "source_data": { + "dataset_name": "MMLU Pro - Law", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "law", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).", + "metric_id": "vals_ai.mmlu_pro.law.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 72.025, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "law", + "task_name": "Law", + "dataset_type": "public", + "industry": "academic", + "raw_score": "72.025", + "raw_stderr": "1.353", + "latency": "72.715", + "cost_per_test": "0.018016", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.353, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:math:zai/glm-5.2:score", + "evaluation_name": "vals_ai.mmlu_pro.math", + "source_data": { + "dataset_name": "MMLU Pro - Math", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "math", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).", + "metric_id": "vals_ai.mmlu_pro.math.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 95.855, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "math", + "task_name": "Math", + "dataset_type": "public", + "industry": "academic", + "raw_score": "95.855", + "raw_stderr": "0.542", + "latency": "40.377", + "cost_per_test": "0.012165", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.542, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:other:zai/glm-5.2:score", + "evaluation_name": "vals_ai.mmlu_pro.other", + "source_data": { + "dataset_name": "MMLU Pro - Others", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "other", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).", + "metric_id": "vals_ai.mmlu_pro.other.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 83.658, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "other", + "task_name": "Others", + "dataset_type": "public", + "industry": "academic", + "raw_score": "83.658", + "raw_stderr": "1.216", + "latency": "34.143", + "cost_per_test": "0.008226", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.216, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:overall:zai/glm-5.2:score", + "evaluation_name": "vals_ai.mmlu_pro.overall", + "source_data": { + "dataset_name": "MMLU Pro - Overall", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "overall", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).", + "metric_id": "vals_ai.mmlu_pro.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.714, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.714", + "raw_stderr": "0.334", + "latency": "51.64", + "cost_per_test": "0.013838", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.334, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:philosophy:zai/glm-5.2:score", + "evaluation_name": "vals_ai.mmlu_pro.philosophy", + "source_data": { + "dataset_name": "MMLU Pro - Philosophy", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "philosophy", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).", + "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 84.168, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "philosophy", + "task_name": "Philosophy", + "dataset_type": "public", + "industry": "academic", + "raw_score": "84.168", + "raw_stderr": "1.634", + "latency": "43.916", + "cost_per_test": "0.011096", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.634, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:physics:zai/glm-5.2:score", + "evaluation_name": "vals_ai.mmlu_pro.physics", + "source_data": { + "dataset_name": "MMLU Pro - Physics", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "physics", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).", + "metric_id": "vals_ai.mmlu_pro.physics.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 92.225, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "physics", + "task_name": "Physics", + "dataset_type": "public", + "industry": "academic", + "raw_score": "92.225", + "raw_stderr": "0.743", + "latency": "63.91", + "cost_per_test": "0.017374", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.743, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "mmlu_pro:psychology:zai/glm-5.2:score", + "evaluation_name": "vals_ai.mmlu_pro.psychology", + "source_data": { + "dataset_name": "MMLU Pro - Psychology", + "source_type": "url", + "url": [ + "https://www.vals.ai/benchmarks/mmlu_pro" + ], + "additional_details": { + "benchmark_slug": "mmlu_pro", + "task_key": "psychology", + "dataset_type": "public", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).", + "metric_id": "vals_ai.mmlu_pro.psychology.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro" + } + }, + "score_details": { + "score": 86.216, + "details": { + "benchmark_slug": "mmlu_pro", + "benchmark_name": "MMLU Pro", + "benchmark_updated": "2026-07-09", + "task_key": "psychology", + "task_name": "Psychology", + "dataset_type": "public", + "industry": "academic", + "raw_score": "86.216", + "raw_stderr": "1.22", + "latency": "30.286", + "cost_per_test": "0.007257", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.22, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.2/ee91b445-6989-4368-bb43-f11ac97528d8.json b/data/vals_ai/zai/glm-5.2/ee91b445-6989-4368-bb43-f11ac97528d8.json new file mode 100644 index 0000000000000000000000000000000000000000..63d8c47905cc749390c3133709453a032ec0e948 --- /dev/null +++ b/data/vals_ai/zai/glm-5.2/ee91b445-6989-4368-bb43-f11ac97528d8.json @@ -0,0 +1,428 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/vals_index/zai_glm-5.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - Vals Index", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "index", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.2", + "id": "zai/glm-5.2", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.2", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "vals_index:corp_fin_v2:zai/glm-5.2:score", + "evaluation_name": "vals_ai.vals_index.corp_fin_v2", + "source_data": { + "dataset_name": "Vals Index - CorpFin v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "corp_fin_v2", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (CorpFin v2).", + "metric_id": "vals_ai.vals_index.corp_fin_v2.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 66.317, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "corp_fin_v2", + "task_name": "CorpFin v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "66.317", + "raw_stderr": "1.614", + "latency": "26.348", + "cost_per_test": "0.0982", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.614, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_index:finance_agent:zai/glm-5.2:score", + "evaluation_name": "vals_ai.vals_index.finance_agent", + "source_data": { + "dataset_name": "Vals Index - Finance Agent v2", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "finance_agent", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Finance Agent v2).", + "metric_id": "vals_ai.vals_index.finance_agent.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 50.83, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "finance_agent", + "task_name": "Finance Agent v2", + "dataset_type": "private", + "industry": "index", + "raw_score": "50.83", + "raw_stderr": "0.203", + "latency": "366.437", + "cost_per_test": "0.662554", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.203, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_index:overall:zai/glm-5.2:score", + "evaluation_name": "vals_ai.vals_index.overall", + "source_data": { + "dataset_name": "Vals Index - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Overall).", + "metric_id": "vals_ai.vals_index.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 65.017, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "index", + "raw_score": "65.017", + "raw_stderr": "1.422", + "latency": "1122.395", + "cost_per_test": "2.076891", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 1.422, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_index:swebench:zai/glm-5.2:score", + "evaluation_name": "vals_ai.vals_index.swebench", + "source_data": { + "dataset_name": "Vals Index - SWE-bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "swebench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (SWE-bench).", + "metric_id": "vals_ai.vals_index.swebench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 83.333, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "swebench", + "task_name": "SWE-bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "83.333", + "raw_stderr": "3.69", + "latency": "650.517", + "cost_per_test": "0.710002" + }, + "uncertainty": { + "standard_error": { + "value": 3.69, + "method": "vals_reported" + } + } + } + }, + { + "evaluation_result_id": "vals_index:terminal_bench_2_1:zai/glm-5.2:score", + "evaluation_name": "vals_ai.vals_index.terminal_bench_2_1", + "source_data": { + "dataset_name": "Vals Index - Terminal-Bench 2.1", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "terminal_bench_2_1", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Terminal-Bench 2.1).", + "metric_id": "vals_ai.vals_index.terminal_bench_2_1.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 67.79, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "terminal_bench_2_1", + "task_name": "Terminal-Bench 2.1", + "dataset_type": "private", + "industry": "index", + "raw_score": "67.79", + "raw_stderr": "0.991", + "latency": "820.506", + "cost_per_test": "0.425184", + "temperature": "1", + "top_p": "1", + "max_output_tokens": "48000", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 0.991, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 1.0, + "max_tokens": 48000, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + }, + { + "evaluation_result_id": "vals_index:vibe_code_bench:zai/glm-5.2:score", + "evaluation_name": "vals_ai.vals_index.vibe_code_bench", + "source_data": { + "dataset_name": "Vals Index - Vibe Code Bench", + "source_type": "other", + "additional_details": { + "benchmark_slug": "vals_index", + "task_key": "vibe_code_bench", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for Vals Index (Vibe Code Bench).", + "metric_id": "vals_ai.vals_index.vibe_code_bench.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/vals_index" + } + }, + "score_details": { + "score": 72.881, + "details": { + "benchmark_slug": "vals_index", + "benchmark_name": "Vals Index", + "benchmark_updated": "2026-07-09", + "task_key": "vibe_code_bench", + "task_name": "Vibe Code Bench", + "dataset_type": "private", + "industry": "index", + "raw_score": "72.881", + "raw_stderr": "6.217", + "latency": "3748.167", + "cost_per_test": "8.488514", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 6.217, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file diff --git a/data/vals_ai/zai/glm-5.2/f631849c-3657-49f4-b6b8-48fd5bcda17c.json b/data/vals_ai/zai/glm-5.2/f631849c-3657-49f4-b6b8-48fd5bcda17c.json new file mode 100644 index 0000000000000000000000000000000000000000..4bab799dd9addd9f9e40a665f8182970ea1c0605 --- /dev/null +++ b/data/vals_ai/zai/glm-5.2/f631849c-3657-49f4-b6b8-48fd5bcda17c.json @@ -0,0 +1,104 @@ +{ + "schema_version": "0.2.2", + "evaluation_id": "vals-ai/proof_bench/zai_glm-5.2/1783734275.1656754", + "retrieved_timestamp": "1783734275.1656754", + "source_metadata": { + "source_name": "Vals.ai Leaderboard - ProofBench", + "source_type": "documentation", + "source_organization_name": "Vals.ai", + "source_organization_url": "https://www.vals.ai", + "evaluator_relationship": "third_party", + "additional_details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "dataset_type": "private", + "industry": "math", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench", + "extraction_method": "static_astro_benchmark_view_props" + } + }, + "eval_library": { + "name": "Vals.ai", + "version": "unknown" + }, + "model_info": { + "name": "glm-5.2", + "id": "zai/glm-5.2", + "developer": "zai", + "additional_details": { + "vals_model_id": "zai/glm-5.2", + "vals_provider": "Zhipu AI" + } + }, + "evaluation_results": [ + { + "evaluation_result_id": "proof_bench:overall:zai/glm-5.2:score", + "evaluation_name": "vals_ai.proof_bench.overall", + "source_data": { + "dataset_name": "ProofBench - Overall", + "source_type": "other", + "additional_details": { + "benchmark_slug": "proof_bench", + "task_key": "overall", + "dataset_type": "private", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "metric_config": { + "evaluation_description": "Accuracy reported by Vals.ai for ProofBench (Overall).", + "metric_id": "vals_ai.proof_bench.overall.accuracy", + "metric_name": "Accuracy", + "metric_kind": "accuracy", + "metric_unit": "percent", + "lower_is_better": false, + "score_type": "continuous", + "min_score": 0.0, + "max_score": 100.0, + "additional_details": { + "score_scale": "percent_0_to_100", + "max_score_source": "fixed_percentage_bound", + "leaderboard_page_url": "https://www.vals.ai/benchmarks/proof_bench" + } + }, + "score_details": { + "score": 35.0, + "details": { + "benchmark_slug": "proof_bench", + "benchmark_name": "ProofBench", + "benchmark_updated": "2026-07-09", + "task_key": "overall", + "task_name": "Overall", + "dataset_type": "private", + "industry": "math", + "raw_score": "35", + "raw_stderr": "4.794", + "latency": "0", + "cost_per_test": "0", + "temperature": "1", + "top_p": "0.95", + "max_output_tokens": "131072", + "reasoning_effort": "max", + "provider": "Zhipu AI" + }, + "uncertainty": { + "standard_error": { + "value": 4.794, + "method": "vals_reported" + } + } + }, + "generation_config": { + "generation_args": { + "temperature": 1.0, + "top_p": 0.95, + "max_tokens": 131072, + "max_attempts": 1 + }, + "additional_details": { + "reasoning_effort": "max" + } + } + } + ] +} \ No newline at end of file